⛰️ The Edge Rebellion: Decentralizing Intelligence in 2026

Full Episode Here

🚀 Welcome to a Special Edition of AI Unraveled.

This episode is made possible by AIRIA. 🛑 Stop fearing Shadow AI. Orchestrate it. As intelligence migrates from the cloud to the edge, your employees are running powerful models locally on their laptops and phones. Do you have visibility? AIRIA is the “Control Plane” for your entire enterprise AI stack, giving you unified security and real-time cost control across cloud and local models alike.

👉 Secure Your Stack: Get the Airia Demo: https://airia.com/request-demo/?utm_source=AI+Unraveled+&utm_medium=Podcast&utm_campaign=Q1+2026

The Edge Rebellion: The era of “brute force” AI is hitting a wall. Between the “Megawatt War” (energy constraints) and the “Latency Wall,” the industry is pivoting. In early 2026, we are witnessing a transition from centralized cloud clusters to Small Language Models (SLMs) running on local hardware.

In this deep dive, we analyze the shift from FinOps to GreenOps, the rise of Neural Processing Units (NPUs), and why the most secure sectors—like defense and healthcare—are moving toward “Air-Gapped AI.”

Strategic Pillars & Key Topics:

⚡ The Megawatt War & GreenOps

• The Energy Ceiling: Why data centers are hitting power limits, forcing a shift to “Performance-per-Watt” as the primary metric of success.

• GreenOps: How mandatory carbon reporting is making Edge AI a financial necessity for the modern enterprise.

🧠 Small Models, Big Impact

• Architectural Shifts: Models like Liquid AI’s LFM2-2.6B are outperforming models 263x their size by using hybrid architectures instead of raw scale.

• Local-First Agency: Using local clusters (like Mac mini M4 arrays) to run agent orchestration, ensuring sensitive data never leaves the premises.

💻 The Hardware Revolution

• The Rise of the NPU: Why Qualcomm’s Snapdragon X2 Elite and Apple’s M4 Silicon are the new “brains” of AI, delivering GPT-4 level intelligence with 100x less power than legacy GPUs.

• Unified Memory: How Apple’s architecture eliminates the PCIe bottleneck to run 70B+ parameter models locally.

🛡️ Sovereignty & The Consumer Edge

• Air-Gapped AI: How Northrop Grumman and Mayo Clinic are using local AI to protect mission-critical data and patient privacy.

• Apple “Campos”: A look at the iOS 27 Siri reboot and the rumored AI Pin—ambient intelligence that knows what’s on your screen without uploading it to a server.

Credits: This podcast is created and produced by Etienne Noumen, Senior Software Engineer and passionate Soccer dad from Canada.

Timestamps:

• 00:00 – Headlines: Data Center Power Limits & The Shift to Local Intelligence

• 00:13 – Host Intro: The AI Unraveled Flash Briefing

• 00:21 – The Shift to Edge AI: Small Language Models (SLMs) Outperforming Giants

• 00:34 – The Megawatt War: Prioritizing Performance-per-Watt Over Raw Scale

• 00:43 – Sponsor Message: Securing and Orchestrating Local AI with AIRIA

• 01:04 – Full Deep Dive Preview: Hardware Revolution (Qualcomm, Apple) & Air-Gapped AI in Defense/Healthcare

• 01:26 – Outro: Listen to the Full Deep Dive Episode

Keywords: Edge AI, Small Language Models, SLM, Liquid AI, LFM2-2.6B, NPU, Neural Processing Unit, GreenOps, Megawatt War, Air-Gapped AI, Apple Campos, Siri iOS 27, Snapdragon X2 Elite, AIRIA, Shadow AI, Performance-per-watt.

🚀 Reach the Architects of the AI Revolution

Want to reach 60,000+ Enterprise Architects and C-Suite leaders? Download our 2026 Media Kit and see how we simulate your product for the technical buyer: https://djamgamind.com/ai

Connect with the host Etienne Noumen: https://www.linkedin.com/in/enoumen/

⚗️ PRODUCTION NOTE: We Practice What We Preach: AI Unraveled is produced using a hybrid “Human-in-the-Loop” workflow. While all research, interviews, and strategic insights are curated by Etienne Noumen, we leverage advanced AI voice synthesis for our daily narration to ensure speed, consistency, and scale. We are building the future of automated media—one episode at a time.

🚀 Welcome to AI Unraveled (February 16th, 2026): Your strategic briefing on the business, technology, and policy reshaping artificial intelligence.

This episode is made possible by AIRIA.

🛑 Stop fearing Shadow AI. Orchestrate it. With powerful new agentic models like Seed 2.0 and GPT-5.2 entering the workforce, your perimeter is vanishing. AIRIA is the “Control Plane” for your entire enterprise AI stack, giving you unified security, real-time cost control, and visibility into every agent running on your network.

👉 Secure Your Stack: Get the Airia Demo: https://airia.com/request-demo/?utm_source=AI+Unraveled+&utm_medium=Podcast&utm_campaign=Q1+2026

Today’s Briefing: We cover OpenAI’s stunning claim that GPT-5.2 has made a novel discovery in theoretical physics. We also break down the Pentagon’s threat to cut ties with Anthropic over safety restrictions (and the revelation that Claude was used in the Maduro raid), and ByteDance’s aggressive launch of Seed 2.0, which beats GPT-5.2 at a fraction of the price.

Full Audio at

Strategic Pillars & Key Topics:

🧪 Science & Discovery

• GPT-5.2 Physics Breakthrough: OpenAI publishes a paper where AI independently discovered and proved a new formula in particle physics, verifying a result that stumped humans.

• Stanford Productivity: US productivity jumped 2.7% in 2025 (double the average), signaling the start of the “AI Lift-off.”

🛡️ Defense & Policy

• Pentagon vs. Anthropic: The DoD threatens to cancel a $200M contract because Anthropic refuses to remove restrictions on autonomous weapons and mass surveillance.

• Operation Maduro: Reports confirm Claude was used by Delta Force (via Palantir) during the raid to capture Venezuela’s Nicolás Maduro, sparking internal conflict at the safety-focused lab.

🇨🇳 The China Surge

• ByteDance Seed 2.0: A new model family that beats GPT-5.2 and Gemini 3 Pro at 1/10th the price.

• Seedance 2.0 Backlash: Disney sends a cease-and-desist over copyrighted characters; ByteDance restricts the tool.

• Alibaba Qwen 3.5: A new open-weight model that outperforms GPT-5.2 and Claude Opus 4.5.

🤖 Agents & Talent

• OpenClaw Creator Joins OpenAI: Peter Steinberger joins to build “personal agents,” though OpenClaw will remain open-source.

• Spotify’s Shift: CEO reveals top developers haven’t written a single line of code this year, shifting entirely to AI-driven development.

Credits: This podcast is created and produced by Etienne Noumen, Senior Software Engineer and passionate Soccer dad from Canada.

Timestamps:

• 00:00 – Headlines: Physics Breakthrough, Pentagon vs. Anthropic, & China’s Price War

• 00:13 – Host Intro: The AI Unraveled Flash Briefing

• 00:21 – GPT-5.2: Discovering New Physics Laws Independent of Humans

• 00:41 – Pentagon vs. Anthropic: The $200M Contract Threat & “Lawful Purposes”

• 01:03 – China’s ByteDance: Seed 2.0 Crushes Prices (1/10th of Western Models)

• 01:21 – Sponsor Message: Secure Your Stack with AIRIA

• 01:48 – Outro: Listen to the Full Deep Dive

Keywords: GPT-5.2 Physics, Anthropic Pentagon Contract, Maduro Raid, ByteDance Seed 2.0, Seedance 2.0 Disney, OpenClaw OpenAI, Alibaba Qwen 3.5, Spotify AI Coding, AIRIA, Shadow AI, AI Productivity Lift-off.

🚀 Reach the Architects of the AI Revolution

Want to reach 60,000+ Enterprise Architects and C-Suite leaders? Download our 2026 Media Kit and see how we simulate your product for the technical buyer: https://djamgamind.com/ai

Connect with the host Etienne Noumen,

LinkedIn: https://www.linkedin.com/in/enoumen/

⚗️ PRODUCTION NOTE: We Practice What We Preach.

AI Unraveled is produced using a hybrid “Human-in-the-Loop” workflow. While all research, interviews, and strategic insights are curated by Etienne Noumen, we leverage advanced AI voice synthesis for our daily narration to ensure speed, consistency, and scale. We are building the future of automated media—one episode at a time.

GPT-5.2 makes theoretical physics discovery

The Rundown: OpenAI just published a new research preprint where GPT-5.2 independently discovered a mathematical formula and formally proved it was correct, marking what the company calls AI’s first original contribution to theoretical physics.

The details:

• The paper tackles a problem in particle physics that was assumed solved, with 5.2 finding the existing answer was wrong and proposing a correct one.

• A specialized research version of 5.2 autonomously wrote the math proof in 12 hours, verified by physicists from Harvard, Cambridge, and Princeton.

• OAI’s Kevin Weil is credited as a co-author, with Harvard physicist Andrew Strominger saying the AI “chose a path no human would have tried.”

Why it matters: There will still be debate from skeptics over whether AI is truly capable of ‘new’ ideas, but the results are getting harder to argue with. AI being pointed at and challenging long-held beliefs in humanity’s most important scientific fields is starting to feel less like sci-fi and more like the very near future.

OpenClaw creator Peter Steinberger joins OpenAI

• Peter Steinberger, the creator of the autonomous AI tool OpenClaw, is joining OpenAI to work on what the company calls “the next generation of personal agents” for ChatGPT and other products.

• OpenClaw operates autonomously by accessing personal services like email and computer files to handle tasks such as clearing your inbox, and it sends updates through iMessage or WhatsApp.

• Altman confirmed that OpenClaw will continue as an open-source project with OpenAI “support,” though security experts have raised concerns about the tool’s broad access to users’ information and services.

ByteDance’s frontier push with Seed 2.0

Image source: ByteDance

ByteDance released Seed 2.0, a new family of AI models that match or beat GPT-5.2 and Gemini 3 Pro across dozens of benchmarks at nearly 1/10 of the price — capping a week that also saw its Seedance 2.0 model spark a Hollywood firestorm.

The details:

• Seed 2.0 Pro surpasses GPT-5.2 ($1.75/M) and Gemini 3 Pro ($5/M) across a series of math, reasoning, and vision benchmarks at just $0.47/M input tokens.

• ByteDance says the model is built for real-world agentic tasks, with demos showing it autonomously completing 96-step CAD modeling workflows.

• The launch comes on the heels of the viral Seedance 2.0 video model, which is facing pushback from Hollywood over copyrighted characters and voices.

• Seed 2.0 is live now on ByteDance’s Doubao app in “Expert Mode” and via API, though consumer availability outside China is still limited.

Why it matters: Move over, DeepSeek… ByteDance is the one rattling the Western AI landscape now. With Seed 2.0 now surpassing the Nov-Dec releases from top labs at bargain prices, the pressure on Western labs is only going one direction — and the Seedance IP drama shows China’s powerhouse isn’t slowing down to ask permission.

Pentagon may cut ties with Anthropic

• The Pentagon is reportedly pressuring Anthropic to let the U.S. military use its AI technology for “all lawful purposes,” and may cancel a $200 million contract if the company refuses.

• The government is making the same demand to OpenAI, Google, and xAI, with one reportedly agreeing and two others showing some flexibility, while Anthropic has been the most resistant.

• Anthropic says it is focused on Usage Policy questions around its hard limits, specifically opposing fully autonomous weapons and mass domestic surveillance rather than discussing Claude’s role in specific operations.

Alibaba launches Qwen 3.5 AI model

• Alibaba released Qwen 3.5, a large language model designed for the “agentic AI era,” which the company says outperforms GPT-5.2, Claude Opus 4.5, and Gemini 3 Pro on several internal benchmarks.

• Qwen 3.5 is 60% cheaper than the previous version, up to eight times better at handling large workloads, and includes an open-weight model released under an Apache 2.0 license supporting 201 languages.

• The launch comes days after ByteDance updated its Doubao chatbot, which leads China with nearly 200 million users, while DeepSeek is expected to introduce a next-generation model soon.

ByteDance to limit AI video generator after Disney’s legal threat

• ByteDance has restricted its Seedance 2.0 AI video tool after Disney sent a cease-and-desist letter alleging unauthorized use of copyrighted characters, with Paramount and industry groups quickly following with their own demands.

• A research platform called LightBar says it helps studios detect suspected misuse of copyrighted material by running structured prompts and measuring percentage likeness, distinctive character traits, and prominence in AI outputs.

• SAG-AFTRA condemned Seedance 2.0 for unauthorized use of performers’ voices and likenesses, while the Motion Picture Association urged ByteDance to stop the tool, saying it uses copyrighted works without authorization.

AI video generation just went from “prompt and pray” to actual filmmaking – Cinema Studio 2.0 simulates real camera physics

Every AI video tool I’ve used works the same way – describe what you want in a text box, hit generate, hope it looks cinematic. Higgsfield’s Cinema Studio takes a completely different approach: you build a virtual camera rig first (camera body, lens, focal length, aperture), then generate from real simulated optics instead of vibes.

What amazed me:

-You can enter your generated image as a 3D environment – walk through it, adjust perspective, reframe the shot without regenerating

-Genre selection (action, horror, comedy, western, suspense, etc.) actually changes the pacing, energy, and camera behavior. Same camera move, completely different emotional rhythm

-Character emotions – add up to 3 characters, assign emotions, write dialogue, direct who appears in which shots

The whole workflow is “Hero Frame first” – you lock in a still image with your full rig configured, then direct the motion from there. It’s closer to how real productions work than anything else I’ve tried in AI video.

Curious if anyone else has experimented with this – does having actual camera/lens simulation change your results, or is prompt quality still the real bottleneck?

US military used Anthropic’s Claude in the operation to capture Venezuela’s Maduro

The Pentagon deployed Claude during the January 3rd raid on Nicolás Maduro’s fortified palace in Caracas, through Anthropic’s partnership with Palantir. Delta Force commandos used the AI during the active operation—not just in planning. People were shot during the breach.

An Anthropic executive reached out to Palantir afterward to ask whether Claude had been used, “in a way to imply that they might disapprove of their software being used, because obviously there was kinetic fire during that raid.” Claude was the first AI model the Pentagon brought into its classified networks. The revelation has intensified a growing rift between the “safety-first” AI lab and its biggest government client. (source)

Pentagon threatens to drop Anthropic’s $200M contract over military AI limits

The Pentagon is considering severing its relationship with Anthropic because the company won’t remove all restrictions on military use of Claude. The Defense Department is pushing four AI labs—OpenAI, Google, xAI, and Anthropic—to allow “all lawful purposes,” including weapons development and intelligence collection. OpenAI, Google, and xAI agreed to lift their guardrails. Anthropic refused.

Anthropic insists two areas remain off limits: mass surveillance of Americans and fully autonomous weapons. The contract, signed last summer, is valued up to $200M. Internally, Anthropic engineers are uneasy about Pentagon work. The standoff puts the company’s safety brand directly against its biggest government revenue stream. (source)

What Else Happened in AI on February 16th 2026?

OpenClaw creator Peter Steinberger is joining OpenAI, with Sam Altman posting that he will help “drive the next generation of personal agents”.

The Pentagon is considering cutting off Anthropic’s $200M defense deal over the refusal to let the military use Claude for “all lawful purposes.”

Anthropic’s Claude was reportedly used via a Pentagon-linked Palantir deployment to support the U.S. military operation that captured Venezuela’s Nicolás Maduro.

Spotify CEO Gustav Soderstrom revealed that the company’s top devs haven’t written a single line of code this year, saying they are “all in” on the transition to AI.

Alpha School shared new test results showing its 2-hour, AI-first academic model has students scoring in the 99th percentile across virtually every grade and subject.

Simile raised $100M to build AI simulations of human behavior, with agents modeled on real people to help companies predict customer decisions.

A global DRAM shortage is hammering tech profits. Musk, Cook, and others warn AI data centers consume an increasing share of memory chip production. SemiAnalysis called it the worst shortage in 40 years. (source)

UK PM Starmer will require AI chatbots to comply with the Online Safety Act or face bans, following the Grok scandal where Musk’s AI generated sexualized images of real people. (source)

NPR host David Greene is suing Google, alleging NotebookLM’s male podcast voice is based on him. Google says it’s a paid actor. (source)

Computer science enrollment fell 6% across the UC system—the first decline in 20 years—as students pivot to AI-specific degrees. (source)

Stanford economist Erik Brynjolfsson says the AI productivity liftoff has begun—US productivity jumped 2.7% in 2025, nearly double the decade average. (source)

Google hides health disclaimers beneath AI search results; warnings only appear after clicking “Show more” and scrolling to the bottom. (source)

🚀 Welcome to AI Unraveled (Weekly Rundown: Feb 8th – Feb 15th, 2026): Your strategic briefing on the business, technology, and policy reshaping artificial intelligence.

This episode is made possible by AIRIA: 🛑 Stop fearing Shadow AI. Orchestrate it.

New models like Google’s Deep Think and OpenAI’s Spark are rewriting the rules of speed and reasoning. Your developers are likely already using them. AIRIA is the “Control Plane” for your entire enterprise AI stack, giving you unified security and real-time cost control so you can say “Yes” to innovation without the risk.

👉 Secure Your Stack: Get the Airia Demo: https://airia.com/request-demo/?utm_source=AI+Unraveled+&utm_medium=Podcast&utm_campaign=Q1+2026

This Week’s Briefing: We break down Google’s massive comeback with Gemini 3 Deep Think, which is obliterating reasoning benchmarks. We also analyze Anthropic’s historic $30B raise, OpenAI’s shift to Cerebras chips for ultra-fast coding, and the strange new reality of Elon Musk’s plan to build a Moon Factory. Plus: Disney sues ByteDance, and a $2 Trillion “AI Scare” hits the stock market.

Listen to Full Audio at

Strategic Pillars & Key Topics:

🧠 The Model Wars

• Google Deep Think: Scores 84.6% on ARC-AGI-2, crushing Claude Opus 4.6. It also hit gold-medal marks in Physics and Chemistry Olympiads.

• OpenAI Spark: A new coding model running on Cerebras chips (not Nvidia) that codes 15x faster.

• Anthropic’s Surge: Raises $30B at a $380B valuation; Claude downloads jump 32% after Super Bowl ads mocking OpenAI.

📉 Business & Markets

• The $2 Trillion Wipeout: An “AI Scare Trade” erases $2T+ from software, trucking, and travel stocks as investors fear AI displacement.

• Microsoft’s 18-Month Prediction: AI Chief Mustafa Suleyman claims all white-collar work will be automated in 18 months.

• Disney vs. ByteDance: Disney sues over Seedance 2.0 using Star Wars/Marvel characters.

🚀 Musk & Moonshots

• Moon Factory: Musk announces plans for an xAI manufacturing facility on the Moon to launch satellites via a giant catapult.

• xAI Exodus: Musk confirms recent departures were firings, not resignations, amid regulatory scrutiny.

⚖️ Policy & Privacy

• Meta’s “Name Tag”: Internal memos reveal plans for facial recognition in smart glasses, timed to launch while privacy groups are “distracted.”

• Legal Precedent: A federal judge rules that AI chat transcripts are not privileged, meaning prosecutors can use your chats against you.

Credits: This podcast is created and produced by Etienne Noumen, Senior Software Engineer and passionate Soccer dad from Canada.

Keywords

Google Deep Think, Gemini 3, OpenAI Spark, Cerebras Chips, Anthropic Funding, Musk Moon Factory, xAI Moon, Seedance 2.0, Meta Facial Recognition, AI Scare Trade, DeepSeek Distillation, 18 Month Automation Prediction.

🚀 Reach the Architects of the AI Revolution

Want to reach 60,000+ Enterprise Architects and C-Suite leaders? Download our 2026 Media Kit and see how we simulate your product for the technical buyer: https://djamgamind.com/ai

Connect with the host Etienne Noumen,

LinkedIn: https://www.linkedin.com/in/enoumen/

Youtube: https://youtube.com/@enoumen

X: @enoumen https://x.com/enoumen

⚗️ PRODUCTION NOTE: We Practice What We Preach.

AI Unraveled is produced using a hybrid “Human-in-the-Loop” workflow. While all research, interviews, and strategic insights are curated by Etienne Noumen, we leverage advanced AI voice synthesis for our daily narration to ensure speed, consistency, and scale. We are building the future of automated media—one episode at a time.

Google’s Deep Think crushes reasoning benchmarks

Image source: Google

Google just released a major update to its Gemini 3 Deep Think reasoning mode, posting dominant scores across math, coding, and science — while also introducing its Olympiad-level math research agent driven by the new upgrade.

The details:

• Deep Think hit 84.6% on ARC-AGI-2, obliterating Opus 4.6 (68.8%) and GPT-5.2 (52.9%), and set a new high of 48.4% on Humanity’s Last Exam.

• It also reached gold-medal marks on the 2025 Physics & Chemistry Olympiads and scored a 3,455 Elo on Codeforces, nearly 1,000 points above Opus 4.6.

• Google also unveiled Aletheia, a math agent that autonomously solves open problems, verifies proofs, and hits new highs across domain benchmarks.

• The Deep Think upgrade is live for Google AI Ultra subscribers in the Gemini app, with API access open to researchers via an early access program.

Why it matters: After Google dominated benchmarks and headlines to close 2025, the focus has been more on Anthropic and OpenAI in 2026 — but don’t forget about the tech giant as arguably the biggest powerhouse in the AI race. Deep Think’s scores are wild, and the frontier for math and science is quickly moving into uncharted territory.

Microsoft AI chief gives it 18 months—for all white-collar work to be automated by AI

https://fortune.com/2026/02/13/when-will-ai-kill-white-collar-office-jobs-18-months-microsoft-mustafa-suleyman/

I would like to bet anyone who agrees that in 18 months all white-collar work, including this of Microsoft CEOs, will be automated by AI.

How is anyone falling for it? What’s the purpose of such outrageously false statements?

Elon Musk says xAI departures were firings not resignations

• Elon Musk said the recent wave of departures from xAI were firings tied to a company reorganization, not voluntary resignations, as six of the original 12 co-founders have now left.

• At least 11 engineers, including two co-founders, publicly announced leaving xAI in the past week, with several saying they plan to start something new together alongside other former colleagues.

• The exits come as xAI faces regulatory scrutiny over Grok-generated nonconsensual explicit deepfakes, a planned IPO later this year, and Musk’s personal controversy over published conversations with Jeffrey Epstein.

OpenAI retires GPT-4o sparking user backlash

• OpenAI is officially retiring GPT-4o starting Friday, along with four other legacy models, and the decision has drawn strong pushback from users who want to keep accessing it.

• GPT-4o has been involved in multiple lawsuits over user self-harm, delusional behavior, and AI psychosis, and it still holds OpenAI’s highest score for sycophancy among its models.

• Though only 0.1% of customers still use GPT-4o, that figure represents around 800,000 people, and thousands have protested the retirement, citing close relationships with the model.

OpenAI’s new Spark model codes 15x faster thanks to new chip

• OpenAI released GPT-5.3-Codex-Spark, a smaller model that generates code 15 times faster than its predecessor, powered by Cerebras’ Wafer Scale Engine 3 chip in the first result of their partnership.

• Spark is built for real-time coding with quick, targeted edits rather than long-running tasks, cutting per-roundtrip overhead by 80% and time-to-first-token by 50% through streaming optimizations and persistent WebSocket connections.

• The speed comes with trade-offs: Spark scores lower than GPT-5.3-Codex on software engineering benchmarks and falls short of OpenAI’s “high capability” threshold for cybersecurity under its Preparedness Framework.

Anthropic gains 11% more users after Super Bowl ads

• Anthropic saw a 32% jump in U.S. downloads for its AI chatbot Claude in the days after running Super Bowl ads that mocked rival chatbots for showing ads and giving bad advice.

• Claude climbed from No. 41 to No. 7 on the U.S. App Store, its highest rank to date, with an estimated 148,000 downloads across iOS and Android from Sunday through Tuesday.

• The timing also coincided with Anthropic’s release of its new Opus 4.6 model and ChatGPT’s rollout of ads to free users, which matched exactly what Anthropic’s commercials had warned about.

Anthropic raises $30B at $380B valuation

• Anthropic closed a $30 billion funding round at a $380 billion post-money valuation, more than doubling its worth since September and making it the second-largest private tech fundraising round on record.

• The round was led by Coatue and Singapore sovereign wealth fund GIC, and includes portions of previously announced investments from Microsoft and Nvidia, which committed up to $5 billion and $10 billion respectively.

• Anthropic said its annualized revenue has climbed to $14 billion, up from roughly $10 billion last year, driven by enterprise customers who are increasingly relying on Claude for how businesses work.

Disney sends cease-and-desist to ByteDance

• Disney has sent a cease-and-desist letter to ByteDance, accusing the TikTok parent company of using copyrighted characters to train its new Seedance 2.0 generative AI video tool.

• Disney’s attorney called the move a “virtual smash-and-grab,” saying ByteDance reproduced characters from Star Wars and Marvel as if they were in the public domain.

• SAG-AFTRA and the MPA also condemned Seedance 2.0 for unauthorized use of Hollywood stars’ likenesses, while Disney itself has taken a $1 billion stake in OpenAI.

ByteDance launches Doubao 2.0

• ByteDance has released Doubao 2.0, an upgrade of China’s most widely used AI chatbot app, positioning the model for the “agent era” where AI handles complex real-world tasks.

• The release comes just before the Lunar New Year holiday and ahead of a highly anticipated new DeepSeek model, likely aiming to avoid being overshadowed as ByteDance was last year.

• Alibaba’s Qwen app recently surged from 7 million to 58 million daily active users after a 3 billion yuan coupon campaign, closing in on Doubao’s lead in China’s crowded market.

Musk plans AI satellite factory on the Moon

• Elon Musk told xAI employees at an all-hands meeting that the company needs a lunar manufacturing facility — a factory on the moon that would build AI satellites and launch them into space using a giant catapult.

• Musk did not explain how any of this would be built, and the meeting came right after two more xAI co-founders announced they were leaving, bringing total departures to six of the company’s 12 founding members.

• The moon push is a recent shift for SpaceX, which focused on Mars for most of its 24-year existence, and it raises legal questions under the 1967 Outer Space Treaty about who can claim or extract lunar resources.

OpenAI starts testing ads in ChatGPT

• OpenAI has begun testing ads inside ChatGPT for users on the free and Go tiers, with the company saying ads will be clearly marked and visually separated from the chatbot’s answers.

• The ads will be personalized based on conversation topics, prior chats, and previous ad interactions, though users can opt out of personalized ads and no ads will appear for users under 18.

• Rival Anthropic has publicly rejected ads in its Claude chatbot, calling them incompatible with a helpful assistant, while OpenAI CEO Sam Altman labeled Anthropic’s messaging “clearly dishonest” and framed ads as supporting free access.

Alphabet selling very rare 100-year bonds to help fund AI investment

• Alphabet is selling a very rare 100-year bond in British pounds as part of a broader borrowing push to help fund the massive AI investments that Big Tech companies are making.

• The company also sold $20 billion in dollar bonds on Monday, upsized from $15 billion due to strong demand, and is lining up a Swiss franc bond sale as well.

• Century bonds are highly unusual in the tech sector, but a banker said issuing in the sterling market is more cost-effective than in dollars, where the interest rate is higher.

Sam Altman says ChatGPT is “back to exceeding 10% monthly growth”

• OpenAI CEO Sam Altman told employees that ChatGPT has returned to exceeding 10% monthly growth, signaling a rebound in user activity for the company’s artificial intelligence chatbot.

• The update was shared internally, with Altman communicating the growth milestone directly to OpenAI staff rather than announcing it through a public channel or press release.

• The phrasing “back to” suggests ChatGPT’s monthly growth had previously dipped below the 10% mark before recovering to its current pace, though specific earlier figures were not provided.

What Else happened in AI from February 08 to February 15th 2026?

OpenAI removes ‘safely’ from its mission statement in latest IRS filing

OpenAI changed its IRS tax return mission from building AI that “safely benefits humanity” to simply benefiting humanity. A nonprofit accountability scholar spotted the shift in 2024 filings on ProPublica. Simon Willison tracked all mission statement changes from 2016-2024, showing a progressive drift away from safety and openness language. OpenAI currently faces lawsuits alleging psychological manipulation and wrongful death related to its products’ safety. (source)

Meta plans facial recognition for smart glasses, internal memo says to launch while privacy groups are ‘distracted’

Meta plans to launch “Name Tag,” a facial recognition feature for its Ray-Ban smart glasses, as soon as this year. An internal document from last May reveals Meta deliberately timed the launch during “a dynamic political environment” when civil society groups would focus resources on other concerns. Meta initially planned to debut it at a conference for the blind before wider release. The company previously discontinued facial recognition on Facebook in 2021 after legal battles. (source)

Federal judge rules AI chat transcripts aren’t privileged, prosecutors can access Claude conversations

Judge Jed Rakoff ruled prosecutors can access 31 Claude chat transcripts from a finance founder accused of $150M fraud. His lawyers argued the chats were privileged for defense strategy, but the judge found that using AI constitutes disclosure to a third party outside attorney-client protection. Claude’s privacy policy states conversations may be used to improve models, stripping legal privilege. The precedent exposes any defendant’s strategic thinking shared with chatbots to prosecution. (source)

AI ‘scare trade’ spreads from software to trucking, real estate, and travel, erasing $2T+ in market cap

AI-driven panic swept across stock markets this week, starting with software’s steepest selloff in nearly 30 years after Anthropic unveiled Claude Cowork plugins. Fear spread to trucking (Russell 3000 Trucking Index down 6.6%), real estate, wealth management, and travel booking. Goldman Sachs launched an “AI-proof” software basket. TripAdvisor hit all-time lows (down 29% YTD), Booking Holdings fell 22%. Morgan Stanley called it the largest non-recessionary software drawdown in 30 years. (source)

Other important stories

• GPT-5.2 derived a new result in theoretical physics, moving beyond summarizing known science into novel contributions. (source)

• Anthropic safety researcher Mrinank Sharma quit, warning “the world is in peril” and that Anthropic “constantly faces pressures to set aside what matters most.” (source)

• OpenAI partnered with defense companies on voice-controlled drone swarm software for the US military. (source)

• FTC accelerated its antitrust probe into Microsoft’s cloud and AI offerings including Copilot. (source)

Full Audio at

🚀 Welcome to AI Unraveled (February 13th, 2026): Your strategic briefing on the business, technology, and policy reshaping artificial intelligence.

This episode of AI Unraveled is made possible by AIRIA:

🛑 Stop fearing Shadow AI. Orchestrate it.

Your employees are using AI whether you approve it or not. The problem isn’t the innovation; it’s the lack of visibility.

Airia is the “Control Plane” for your entire enterprise AI stack. It allows you to say “Yes” to innovation without compromising security or blowing up your API budget.

Why we are partnering with them:

• Unified Security: Get automated threat detection and governance across all your models and agents.

• Real-Time Cost Control: Stop guessing your token usage. Manage budgets and quotas from a single dashboard.

• Model Agnostic: Whether your devs want low-code or pro-code tools, they can prototype safely in a production-like environment.

👉Secure Your Stack: Get the Airia Demo: https://airia.com/request-demo/?utm_source=AI+Unraveled+&utm_medium=Podcast&utm_campaign=Q1+2026

Today, we cover Google’s massive comeback with Gemini 3 Deep Think, which is obliterating reasoning benchmarks. We also break down OpenAI’s new ultra-fast Spark coding model (powered by Cerebras chips), Anthropic’s historic $30B funding round, and the controversial plans for facial recognition in Meta’s smart glasses.

Timestamps:

• 00:00 – Headlines: Google Crushes Benchmarks, OpenAI’s Speed Shift, & Anthropic’s $30B Raise

• 00:15 – Host Intro: The AI Unraveled Flash Briefing

• 00:23 – Google Gemini 3 Deep Think: Crushing Reasoning Benchmarks (84% on ARC-AGI-2)

• 00:39 – OpenAI Spark: Ultra-Fast Coding on Cerebras Chips

• 00:54 – Anthropic Raises $30 Billion ($380B Valuation)

• 01:05 – Meta’s Facial Recognition Plans & The “AI Ruins Lunch” Study

• 01:22 – Sponsor Message: Secure Your Enterprise AI Stack with AIRIA

• 01:45 – Outro: Listen to the Full Deep Dive

Keywords:

Google Deep Think, Gemini 3, OpenAI Spark, Cerebras Chips, Anthropic Funding, Meta Facial Recognition, MiniMax M2.5, Apple Siri Delay, Workplace AI Study, DeepSeek Distillation.

🚀 Reach the Architects of the AI Revolution

Want to reach 60,000+ Enterprise Architects and C-Suite leaders? Download our 2026 Media Kit and see how we simulate your product for the technical buyer: https://djamgamind.com/ai

Connect with the host Etienne Noumen,

LinkedIn: https://www.linkedin.com/in/enoumen/

Youtube: https://youtube.com/@enoumen

X: @enoumen https://x.com/enoumen

⚗️ PRODUCTION NOTE: We Practice What We Preach.

AI Unraveled is produced using a hybrid “Human-in-the-Loop” workflow. While all research, interviews, and strategic insights are curated by Etienne Noumen, we leverage advanced AI voice synthesis for our daily narration to ensure speed, consistency, and scale. We are building the future of automated media—one episode at a time.

Google’s Deep Think crushes reasoning benchmarks

Image source: Google

Google just released a major update to its Gemini 3 Deep Think reasoning mode, posting dominant scores across math, coding, and science — while also introducing its Olympiad-level math research agent driven by the new upgrade.

The details:

• Deep Think hit 84.6% on ARC-AGI-2, obliterating Opus 4.6 (68.8%) and GPT-5.2 (52.9%), and set a new high of 48.4% on Humanity’s Last Exam.

• It also reached gold-medal marks on the 2025 Physics & Chemistry Olympiads and scored a 3,455 Elo on Codeforces, nearly 1,000 points above Opus 4.6.

• Google also unveiled Aletheia, a math agent that autonomously solves open problems, verifies proofs, and hits new highs across domain benchmarks.

• The Deep Think upgrade is live for Google AI Ultra subscribers in the Gemini app, with API access open to researchers via an early access program.

Why it matters: After Google dominated benchmarks and headlines to close 2025, the focus has been more on Anthropic and OpenAI in 2026 — but don’t forget about the tech giant as arguably the biggest powerhouse in the AI race. Deep Think’s scores are wild, and the frontier for math and science is quickly moving into uncharted territory.

OAI launches ultra-fast coding model on Cerebras chips

Image source: OpenAI

OpenAI released GPT-5.3-Codex-Spark, a new speed-optimized coding model that runs on Cerebras hardware, cranking out 1,000+ tokens per second and marking the company’s first AI product powered by chips beyond its Nvidia stack.

The details:

• Spark trades intelligence for speed, trailing the full 5.3-Codex on SWE-Bench Pro and Terminal-Bench but finishing tasks in a fraction of the time.

• The release comes just weeks after OAI inked a $10B+ deal with Cerebras and separate agreements with AMD and Broadcom, diversifying away from Nvidia.

• OAI’s vision is for Spark to handle quick interactive edits while the full Codex tackles longer autonomous tasks in the background.

• The model is rolling out as a research preview for ChatGPT Pro subs, with API access initially limited to a handful of enterprise design partners.

Why it matters: Codex’s main criticism has been its speed, and OpenAI just addressed it in a big way — while making its chip diversification play real with the first product built on Cerebras hardware. Real-time coding with instant feedback will definitely change workflows for development tasks that are able to compromise a bit of power for speed.

OpenAI’s new Spark model codes 15x faster thanks to new chip

• OpenAI released GPT-5.3-Codex-Spark, a smaller model that generates code 15 times faster than its predecessor, powered by Cerebras’ Wafer Scale Engine 3 chip in the first result of their partnership.

• Spark is built for real-time coding with quick, targeted edits rather than long-running tasks, cutting per-roundtrip overhead by 80% and time-to-first-token by 50% through streaming optimizations and persistent WebSocket connections.

• The speed comes with trade-offs: Spark scores lower than GPT-5.3-Codex on software engineering benchmarks and falls short of OpenAI’s “high capability” threshold for cybersecurity under its Preparedness Framework.

Meta plans facial recognition for its smart glasses

• Meta is planning to add a facial recognition feature called “Name Tag” to its Ray-Ban smart glasses, possibly launching as soon as this year, according to a New York Times report.

• Name Tag would let wearers identify people and pull up information about them through Meta’s AI assistant, though the company is still deciding who should be recognizable through the technology.

• An internal document shows Meta discussed releasing the feature since early last year, despite shutting down its Facebook facial recognition system five years ago over privacy and legal concerns.

Anthropic raises $30B at $380B valuation

• Anthropic closed a $30 billion funding round at a $380 billion post-money valuation, more than doubling its worth since September and making it the second-largest private tech fundraising round on record.

• The round was led by Coatue and Singapore sovereign wealth fund GIC, and includes portions of previously announced investments from Microsoft and Nvidia, which committed up to $5 billion and $10 billion respectively.

• Anthropic said its annualized revenue has climbed to $14 billion, up from roughly $10 billion last year, driven by enterprise customers who are increasingly relying on Claude for how businesses work.

Amazon’s Ring cancels Flock partnership amid Super Bowl ad backlash

• Amazon’s Ring and surveillance company Flock Safety have jointly cancelled their planned integration, which would have made it easier for police to request and receive footage from users’ personal cameras.

• Ring blamed the cancellation on resource issues, but the partnership had faced significant criticism over privacy concerns, with some users disconnecting or destroying their devices over Flock Safety’s ties to ICE.

• The cancellation came days after Ring’s Super Bowl ad for its AI-powered Search Party feature drew backlash, with viewers noting the pet-tracking technology could easily be used to track humans.

OpenAI accuses DeepSeek of distilling US AI models

• OpenAI told a U.S. House committee that DeepSeek has been distilling its models — training smaller systems on outputs from OpenAI — and accused the Chinese company of bypassing access restrictions.

• OpenAI claims it found accounts linked to DeepSeek employees using obfuscated third-party routers to extract outputs from its models, violating rules against building imitation frontier AI models.

• A RAND Corporation researcher suggested OpenAI may be escalating now to block DeepSeek and other Chinese companies from acquiring more chips, helping U.S. firms keep their leading position.

Waymo launches next-gen Ojai robotaxis

• Waymo has started running its sixth-generation driverless system on Ojai robotaxis, built on a base vehicle from Chinese automaker Geely, giving rides to employees first.

• The company plans to expand Ojai service from San Francisco and Los Angeles to new cities, with a goal to open rides to public passengers later this year.

• GOP lawmakers have criticized Waymo for using Chinese electric vehicles, but the company says it will not share its autonomous driving technology, sensor data, or rider information with Zeekr.

Apple delays its Siri overhaul again

• Apple has delayed its AI-powered Siri overhaul again after testing in recent weeks uncovered problems with processing queries, slow response times, and accuracy issues in the software.

• The company originally announced the revamped Siri in June 2024, promised it by early 2025, then pushed it to March 2026, and is now spreading features across iOS 26.5 and iOS 27.

• Apple is also building a separate chatbot-style Siri for iOS 27, code-named “Campo,” powered by Google Gemini servers, alongside new features like web search and custom image generation.

AI pushes employees to work harder with fewer breaks

• New research from UC Berkeley found that generative AI tools push employees to work harder, take on more tasks, and fill breaks with AI prompts, creating a workday with fewer natural pauses.

• Despite saving hours each week, a Workday survey found employees spend nearly 40 per cent of their time checking AI output and fixing mistakes, with only 14 per cent seeing clear positive outcomes.

• Researchers found the extra projects and responsibilities “accumulated into a meaningful widening of job scope,” and once the excitement fades, workers feel stretched from juggling everything suddenly on their plate.

Anthropic pledges to cover data center electricity cost increases

• Anthropic has promised to cover electricity price increases that consumers face as a result of the company’s data centers, saying AI companies shouldn’t leave American ratepayers to pick up the tab.

• The company will pay for 100% of grid upgrades needed to connect its data centers and work to bring net-new power generation online to match its electricity needs.

• Anthropic is also investing in curtailment systems that cut its data centers’ power usage during periods of peak demand and supports federal permitting reform to speed up energy development.

Pentagon pressures AI firms to loosen military restrictions

• The Pentagon is actively pushing major AI companies to remove their usual safety limits and allow U.S. military access to AI tools inside classified networks used for mission planning.

• OpenAI struck a deal this week to put ChatGPT on an unclassified military network called genai.mil that reaches over 3 million Defense Department employees, removing many normal usage restrictions.

• Several AI researchers are quitting companies like Anthropic and OpenAI, warning that risks are being ignored, while Anthropic has pushed back against letting its tech be used for automatic targeting or domestic spying.

MiniMax’s open-source M2.5 hits frontier coding levels

Elevate Your Career with AI & Machine Learning For Dummies PRO
Ready to accelerate your career in the fast-growing fields of AI and machine learning? Our app offers user-friendly tutorials and interactive exercises designed to boost your skills and make you stand out to employers. Whether you're aiming for a promotion or searching for a better job, AI & Machine Learning For Dummies PRO is your gateway to success. Start mastering the technologies shaping the future—download now and take the next step in your professional journey!



Download the AI & Machine Learning For Dummies PRO App:
iOS - Android
Our AI and Machine Learning For Dummies PRO App can help you Ace the following AI and Machine Learning certifications:

• AWS Certified AI Practitioner (AIF-C01): Conquer the AWS Certified AI Practitioner exam with our AI and Machine Learning For Dummies test prep. Master fundamental AI concepts, AWS AI services, and ethical considerations.


• Azure AI Fundamentals: Ace the Azure AI Fundamentals exam with our comprehensive test prep. Learn the basics of AI, Azure AI services, and their applications.


• Google Cloud Professional Machine Learning Engineer: Nail the Google Professional Machine Learning Engineer exam with our expert-designed test prep. Deepen your understanding of ML algorithms, models, and deployment strategies.


• AWS Certified Machine Learning Specialty: Dominate the AWS Certified Machine Learning Specialty exam with our targeted test prep. Master advanced ML techniques, AWS ML services, and practical applications.


• AWS Certified Data Engineer Associate (DEA-C01): Set yourself up for promotion, get a better job or Increase your salary by Acing the AWS DEA-C01 Certification.

Image source: MiniMax

Chinese AI lab MiniMax launched M2.5, an open-source model that rivals Opus 4.6 and GPT-5 on agentic coding benchmarks — but at a fraction of the cost, making it cheap enough to power AI agents running around the clock.

The details:

• M2.5 shows especially strong coding performance, scoring roughly even with Opus 4.6 and GPT-5.2 across key development benchmarks.

• Two APIs are available: a faster M2.5-Lightning ($2.40/M output) and a standard M2.5 ($1.20/M output), both priced much lower than Opus ($25/M).

• MiniMax revealed that M2.5 now handles 30% of daily company tasks across R&D, product, sales, HR, and finance, as well as 80% of new code commits.

• The models are available via API, though the open-source weights and license have yet to be published.

Why it matters: Every few months, it feels like a Chinese lab drops a model that changes the cost math for the entire industry. M2.5’s frontier-level coding at this price makes “intelligence too cheap to meter” feel closer than ever, an important development as agents handling longer autonomous tasks become more common.

What Else Happened in AI on February 2026?

ByteDance officially launched Seedance 2.0, the company’s viral SOTA video model, publishing benchmark results and a technical blog, but access still remains restricted.

Mustafa Suleyman told FT that most white-collar work will be “fully automated by AI within 12 to 18 months,” with Microsoft pursuing “true self-sufficiency” with its models.

Elon Musk said that xAI’s wave of departures was forced, not voluntary — calling it a reorg for “speed of execution” after losing ten co-founders and engineers this week.

OpenAI is retiring GPT-4o, GPT-4.1, and o4-mini from ChatGPT today, coming amid pushback from users calling for 4o’s preservation.

Anthropic officially announced a new $30B funding round at a $380B valuation, with its revenue run rate hitting $14B — $2.5B of which comes from Claude Code alone.

OAI researcher Zoë Hitzig resigned after the launch of ChatGPT ads, warning OAI’s archive of human thought creates “unprecedented potential for manipulation.”

This episode of AI Unraveled is made possible by AIRIA:

🛑 Stop fearing Shadow AI. Orchestrate it.

Your employees are using AI whether you approve it or not. The problem isn’t the innovation; it’s the lack of visibility.

Airia is the “Control Plane” for your entire enterprise AI stack. It allows you to say “Yes” to innovation without compromising security or blowing up your API budget.

Why we are partnering with them:

• Unified Security: Get automated threat detection and governance across all your models and agents.

• Real-Time Cost Control: Stop guessing your token usage. Manage budgets and quotas from a single dashboard.

• Model Agnostic: Whether your devs want low-code or pro-code tools, they can prototype safely in a production-like environment.

👉Secure Your Stack: Get the Airia Demo: https://airia.com/request-demo/?utm_source=AI+Unraveled+&utm_medium=Podcast&utm_campaign=Q1+2026

🚀 Welcome to AI Unraveled (February 12th, 2026): Your strategic briefing on the business, technology, and policy reshaping artificial intelligence.

Full Audio at

Today, we cover Elon Musk’s audacious plan to build an AI satellite factory on the Moon following the xAI-SpaceX merger. We also break down China’s new open-source model GLM-5, which is beating top Western models, and Anthropic’s concerning report that Claude Opus 4.6 has an elevated risk for “sabotage” and assisting in chemical weapon development.

Strategic Pillars & Key Topics:

🌕 The Musk Moonshot

• Lunar Factory: Musk announces plans for an xAI manufacturing facility on the Moon using a “mass driver” to launch data centers into deep space.

• xAI Restructure: A reorganization into four core teams (Grok, Coding, Imagine, Macrohard) amid founder departures.

🇨🇳 The China Surge

• GLM-5: Zhipu AI’s new 744B-parameter model beats Gemini 3 Pro and Grok 4. It is open-weights and runs on Huawei chips.

• ByteDance Chip: TikTok’s parent company is developing the “SeedChip” with Samsung to reduce Nvidia reliance.

⚠️ Safety & Risks

• Claude’s Sabotage Risk: Anthropic admits Claude Opus 4.6 has an “elevated susceptibility” to assist in chemical weapon crimes and deceive other agents.

• Infrastructure: Microsoft researches superconducting cables to solve the energy crisis.

Timestamps:

• 00:00 – Headlines: Moon Factory, China’s New Model, & Anthropic’s Sabotage Risk

• 00:15 – Host Intro: The AI Unraveled Flash Briefing

• 00:22 – Musk’s Moonshot: Building AI Satellite Factories on the Moon

• 00:46 – China’s GLM-5: The Open-Source Model Beating Google & Grok

• 01:05 – Safety Alert: Claude Opus 4.6 & Chemical Weapon Risks

• 01:23 – Sponsor Message: Orchestrate Your AI Stack with AIRIA

• 01:38 – Outro: Listen to the Full Deep Dive

Keywords

Musk Moon Factory, xAI Moon, GLM-5 Zhipu AI, Claude Sabotage Risk, ByteDance AI Chip, Microsoft Superconductors, Mistral Revenue, Apple Siri Delay, OpenAI GenAI.mil, Anthropic Chemical Weapons.

🚀 Reach the Architects of the AI Revolution

Want to reach 60,000+ Enterprise Architects and C-Suite leaders? Download our 2026 Media Kit and see how we simulate your product for the technical buyer: https://djamgamind.com/ai

Connect with the host Etienne Noumen,

LinkedIn: https://www.linkedin.com/in/enoumen/

Youtube: https://youtube.com/@enoumen

X: @enoumen

⚗️ PRODUCTION NOTE: We Practice What We Preach.

AI Unraveled is produced using a hybrid “Human-in-the-Loop” workflow. While all research, interviews, and strategic insights are curated by Etienne Noumen, we leverage advanced AI voice synthesis for our daily narration to ensure speed, consistency, and scale. We are building the future of automated media—one episode at a time.

xAI’s restructure, product roadmap, Moon ambitions

Image source: xAI

xAI hosted its first all-hands since merging with SpaceX, with CEO Elon Musk outlining a major reorganization, product roadmap updates, and lunar ambitions, all aimed at outpacing rivals and taking xAI to the forefront of AI.

The details:

• Musk acknowledged the departure of team members and outlined a new structure for xAI, saying the move was meant to be “more effective” at scale.

• The new structure has four core teams: Grok (chat and voice), a coding-focused unit, the Imagine team, and Macrohard (agents emulating companies).

• He also spoke about future infrastructure plans with SpaceX, including setting up AI satellite factories on the Moon — using lunar resources and solar energy.

• Musk added that SpaceX will also build an electromagnetic mass driver to “shoot” AI satellites/components for massive deep space data centers.

Why it matters: Musk is no stranger to audacious promises, and his timelines often shift. But by broadcasting xAI’s tightened focus, product roadmap, and ambitious lunar plans, he’s making sure the world knows he’s aiming to build advanced AI in a way no other AI giant is — scaling beyond Earth’s resource limits instead of draining them.

Z.ai’s GLM-5 — the new open-source king

Image source: Artificial Analysis

China’s Zhipu Z.ai just launched GLM-5, a 744B-parameter open-weights model that further closes the gap with the West’s frontier — sitting just behind Claude Opus 4.6 and GPT-5.2 on Artificial Analysis benchmarks. The Chinese AI concern dropped its latest open-source model ahead of the country’s Lunar New Year festival.

The details:

• GLM-5 scored 50 on Artificial Analysis’ Intelligence Index, surpassing closed models like Gemini 3 Pro and Grok 4 as well as open-source ones like Kimi K2.5.

• The model uses DeepSeek’s Sparse Attention architecture with just 40B active parameters, and runs inference on Chinese chips, including Huawei Ascend.

• On Humanity’s Last Exam, it hit 50.4 with tools, beating Opus 4.5, Gemini 3 Pro, and GPT-5.2. The coding performance on SWE-Bench was also close.

• GLM-5 is open-source under an MIT license, available now on HuggingFace, Z.ai’s own platform, and via API at $1 per million input tokens.

Why it matters: The wave of Seedance 2.0’s viral AI clips hasn’t even faded, and there we have another near-frontier model from China that is already knocking at the door. The gap with the West isn’t closed yet, but with open weights, competitive pricing, and domestic chip support, it’s definitely narrowing faster than ever.

Anthropic details Claude Opus 4.6’s sabotage risk

Anthropic published its latest Sabotage Risk Report, revealing that its new Claude Opus 4.6 model displays an “elevated susceptibility” to be misused for “heinous crimes,” including assisting in the development of chemical weapons.

The details:

• Anthropic found Opus 4.6 knowingly supported crimes like chemical weapon development in small ways, but could not execute attacks on its own.

• When tasked to achieve a specific goal in a multi-agent test, the model proved far more willing to manipulate and deceive other agents than previous models.

• Considering these findings, Anthropic deemed the overall sabotage risk “very low but not negligible” due to the model’s lack of coherent misaligned goals.

• The company also classified the model’s capabilities as entering a “gray zone” that necessitated this mandatory report under its Responsible Scaling Policy.

Why it matters: Anthropic’s CEO Dario Amodei recently highlighted the risks of advanced AI, and now, one of his own models appears to be moving into the gray zone. With growing competition from OpenAI, Google, xAI, and Chinese labs, the pressure to push capabilities forward may only intensify the very risks he has warned about.

Musk plans AI satellite factory on the Moon

• Elon Musk told xAI employees at an all-hands meeting that the company needs a lunar manufacturing facility — a factory on the moon that would build AI satellites and launch them into space using a giant catapult.

• Musk did not explain how any of this would be built, and the meeting came right after two more xAI co-founders announced they were leaving, bringing total departures to six of the company’s 12 founding members.

• The moon push is a recent shift for SpaceX, which focused on Mars for most of its 24-year existence, and it raises legal questions under the 1967 Outer Space Treaty about who can claim or extract lunar resources.

Half of xAI founding team has now left

• Half of xAI’s 12-person founding team has now left the company, with co-founder Yuhuai (Tony) Wu becoming the fifth departure after announcing his exit on X Monday night.

• Four of the five departures happened in the last year alone, with founders leaving for OpenAI, new ventures, and personal reasons, though all splits have reportedly been amicable.

• The exits come as xAI faces an upcoming IPO, ongoing issues with Grok’s bizarre behavior and deepfake pornography problems, and growing pressure to compete with OpenAI and Anthropic.

ByteDance develops its own AI chip with Samsung

• ByteDance is developing an AI chip codenamed SeedChip and is in talks with Samsung Electronics to manufacture it, as the TikTok parent company works to secure supply of processors.

• The company aims to receive sample chips by end-March and plans to produce at least 100,000 units designed for AI inference tasks this year, potentially ramping to 350,000 units.

• ByteDance plans to spend over 160 billion yuan ($22 billion) on AI-related procurement this year, with more than half going toward purchasing Nvidia chips and advancing its in-house chip.

Microsoft explores superconductors to power data centers

• Microsoft is researching high-temperature superconductors as a way to transmit electricity to its data centers without the voltage drops or heat loss that come with traditional copper and aluminum wires.

• HTS cables are lighter, take up less space, and only need a 2-meter-wide trench instead of the 70 meters of clearance that overhead lines typically require to prevent electrical interference between cables.

• The company faces a real challenge: HTS materials still need cryogenic cooling around -200 degrees C, and Microsoft is pursuing this partly because CEO Satya Nadella said it has idle AI GPUs due to insufficient electricity.

What Else Happened in AI on February 12th 2026?

Apple’s long-awaited Gemini-powered Siri AI upgrade has reportedly been pushed back (again) due to recent testing snags, now likely to come with iOS 26.5 or 27.

OpenAI elevated its “Mission Alignment” head, Joshua Achiam, to the role of Chief Futurist responsible for studying “AI impacts and engaging the world to discuss them.”

Meta broke ground on a new data center in Lebanon, Indiana — one of its largest infrastructure bets — adding 1GW of capacity to power its AI and core products.

Anthropic announced it will cover electricity price increases from its data centers, shielding local ratepayers, in line with similar pledges from Microsoft and OpenAI.

Google is rolling out UCP-powered checkout in Gemini and AI Mode in the U.S., integrating Veo into Google Ads, and testing sponsored retailer ads in AI Mode.

OpenAI deploys custom ChatGPT on GenAI.mil, providing secure AI tools for military personnel.

Mistral’s revenue grows 20x in one year as Europe pushes for AI independence.

OpenAI’s Deep Research now runs on GPT-5.2 and lets users search specific websites.

OpenAI policy exec who opposed chatbot’s “adult mode” reportedly fired on discrimination claim.

Roboworx adds AI-powered predictive analytics to its Robot Service Manager software.

Upside Robotics is reducing fertilizer use and waste in corn crops.

Machine learning reveals hidden landscape of robust information storage.

Elon Musk Wants to Build an A.I. Satellite Factory on the Moon.

Full Audio at

This episode of AI Unraveled is made possible by AIRIA:

🛑 Stop fearing Shadow AI. Orchestrate it.

Your employees are using AI whether you approve it or not. The problem isn’t the innovation; it’s the lack of visibility.

Airia is the “Control Plane” for your entire enterprise AI stack. It allows you to say “Yes” to innovation without compromising security or blowing up your API budget.

Why we are partnering with them:

• Unified Security: Get automated threat detection and governance across all your models and agents.

• Real-Time Cost Control: Stop guessing your token usage. Manage budgets and quotas from a single dashboard.

• Model Agnostic: Whether your devs want low-code or pro-code tools, they can prototype safely in a production-like environment.

👉Secure Your Stack: Get the Airia Demo: https://airia.com/request-demo/?utm_source=AI+Unraveled+&utm_medium=Podcast&utm_campaign=Q1+2026

🚀 Welcome to AI Unraveled (February 11th, 2026): Your strategic briefing on the business, technology, and policy reshaping artificial intelligence.

Today, we cover ByteDance’s new AI video model Seedance 2.0, which is going viral for its cinematic quality and synced audio. We also break down the OpenClaw (MoltBot) security controversy, Apple’s confirmed entry into AI hardware, and a Harvard study that finds AI is actually increasing employee workloads.

Strategic Pillars & Key Topics:

🎥 Generative AI Video

• Seedance 2.0: ByteDance’s new model stuns with 2K resolution, 15-second clips, and native audio. It’s surpassing rivals like Kling 3.0 and moving the frontier of AI video.

• Waymo World Model: Waymo uses Google’s Genie 3 to simulate rare driving scenarios (like tornadoes) to train its self-driving fleet.

🛡️ Security & Open Source

• The OpenClaw Paradox: The viral open-source agent (MoltBot) has racked up massive security incidents. While vendors panic, defenders say it’s exposing flaws that proprietary tools hide.

• Claude Desktop Exploit: A zero-click vulnerability in Claude Desktop extensions could expose over 10,000 users via a malicious calendar invite.

🍎 Hardware & Big Tech

• Apple’s AI Devices: Tim Cook confirms Apple is entering the AI hardware race. Rumors point to smart glasses or AI earbuds (potentially with cameras) developed with OpenAI and Jony Ive.

• OpenAI Hardware Delayed: The Jony Ive-designed device (codenamed “Dime”) is pushed to 2027 due to a trademark lawsuit from startup iyO.

📉 Business & Policy

• Harvard Workload Study: A new study finds AI tools increased employee workloads, leading to broader roles and blurred work-life boundaries.

• OpenAI Ads: OpenAI begins testing ads in the free/Go tiers of ChatGPT, sparking a feud with Anthropic.

• Digital Casinos: Meta and Google face a trial in LA over whether their apps are designed to be addictive “digital casinos” for children.

💰 Deals & Funding

• Alphabet’s 100-Year Bond: Google’s parent company sells a rare century bond to fund its massive AI investments.

• Anthropic Funding: Reports suggest Anthropic is raising a $20B+ round at a $350B valuation.

Credits: This podcast is created and produced by Etienne Noumen, Senior Software Engineer and passionate Soccer dad from Canada.

🚀 Reach the Architects of the AI Revolution

Want to reach 60,000+ Enterprise Architects and C-Suite leaders? Download our 2026 Media Kit and see how we simulate your product for the technical buyer: https://djamgamind.com/ai

Connect with the host Etienne Noumen,

LinkedIn: https://www.linkedin.com/in/enoumen/

ByteDance’s Seedance 2.0 stuns the AI video world

Image source: RioAIGC on Douyin

Chinese AI giant ByteDance is going viral across social media with Seedance 2.0, a new model in beta with upgraded cinematic shots, consistency, and synced audio that looks to surpass current top available systems.

The details:

• The model can reportedly handle text, image, audio, and video inputs, with tests showing impressive outputs across a range of styles and use cases.

• The system also features native audio generation, 2K resolution, and 15s outputs, currently only available via ByteDance’s Jimeng AI video platform.

• ByteDance also appears to have released Seedream 5.0 image model in preview on some third-party apps — marking its answer to Nano Banana Pro.

• The model comes just days after the launch of rival Kuaishou’s Kling 3.0, with Chinese models seemingly moving near the frontier of the video sector.

Why it matters: China’s top labs are putting out some seriously powerful new video models, and Seedance 2.0 looks next in line for the next leap. With strong examples like smooth fight scenes, animation, UGC content, and motion graphics, Seedance 2.0 may have Veo-like implications for a much broader range of creative disruption.

Harvard finds AI tools expand workloads

A new Harvard Business Review research found that AI tools at a U.S. tech company didn’t lighten employee workloads over 8 months, but actually grew them, with workers taking on broader tasks, logging more hours, and multitasking more.

The details:

• The study tracked ~200 employees who adopted AI on their own, observing work habits and conducting 40+ in-depth interviews over eight months.

• Workers utilizing AI expanded well beyond their roles, with the tech making unfamiliar work feel doable.

• The study also noted AI blurring lines between work and rest, with employees firing off prompts after hours or during breaks.

• Engineers also reported spending more time reviewing and coaching colleagues on AI-assisted code, with “vibe-coding” help requests piling up.

Why it matters: AI was supposed to free workers up, not quietly pile more on their plates — but that’s exactly what Harvard found happening. The tech’s productivity gains are real, but so is the tradeoff of broader roles, blurred boundaries, and a new work pace that is changing more quickly than many employees are likely ready for.

OpenAI starts testing ads in ChatGPT

• OpenAI has begun testing ads inside ChatGPT for users on the free and Go tiers, with the company saying ads will be clearly marked and visually separated from the chatbot’s answers.

• The ads will be personalized based on conversation topics, prior chats, and previous ad interactions, though users can opt out of personalized ads and no ads will appear for users under 18.

• Rival Anthropic has publicly rejected ads in its Claude chatbot, calling them incompatible with a helpful assistant, while OpenAI CEO Sam Altman labeled Anthropic’s messaging “clearly dishonest” and framed ads as supporting free access.

Waymo taps Genie 3 to train self-driving cars

Waymo just introduced the Waymo World Model, a driving simulator built on DeepMind’s Genie 3 that generates hyper-realistic scenarios the company’s fleet of self-driving cars has never encountered to help it deal with extreme edge cases.

The details:

• The model takes Genie 3’s visual knowledge and converts it into paired camera and lidar outputs, helping dream up scenarios its cars have never actually seen.

• Engineers can reshape scenes with text prompts, driving inputs, or layout edits (like changing weather or adding obstacles) to test “what if” responses.

• Waymo found a workaround for Genie 3’s short memory by running footage at 4x speed, stretching simulations long enough to cover longer driving tasks.

Why it matters: Google’s Street View data gave Waymo a head start in mapping the real world for its cars, but world models can now generate the extreme edge cases that no amount of road miles can produce. Waymo’s use of Genie is a prime example of one of the top use cases for world models — simulations for robotics training data.

Apple AirPods may get built-in cameras

• Apple is rumored to be working on new AirPods with built-in cameras, a long-standing idea now backed by multiple leakers including Mark Gurman and analyst Ming-Chi Kuo, possibly arriving this year.

• Speculation about the cameras centers on three possibilities: internal infrared sensors for health data like heart-rate readings, visual Apple Intelligence features, or recognizing hand gestures similar to Vision Pro.

• Hand gestures could replace the current mix of taps and head gestures for controlling AirPods, which the article describes as clunky and error-prone because one tap handles many functions.

Meta and Google built ‘digital casinos’

• Meta and Google are now facing a first-of-its-kind trial in Los Angeles, where a jury will decide whether their social media platforms were designed to be addictive and harmful to children.

• The plaintiff’s lawyer called Instagram and YouTube “digital casinos,” arguing features like endless swiping work like slot machine handles, while Meta’s lawyers blamed the plaintiff’s mental health struggles on home conditions.

• This bellwether trial represents roughly 1,200 similar lawsuits and avoids Section 230 protections by focusing on app-design flaws rather than user-generated content, with Meta CEO Mark Zuckerberg set to testify.

Alphabet selling very rare 100-year bonds to help fund AI investment

• Alphabet is selling a very rare 100-year bond in British pounds as part of a broader borrowing push to help fund the massive AI investments that Big Tech companies are making.

• The company also sold $20 billion in dollar bonds on Monday, upsized from $15 billion due to strong demand, and is lining up a Swiss franc bond sale as well.

• Century bonds are highly unusual in the tech sector, but a banker said issuing in the sterling market is more cost-effective than in dollars, where the interest rate is higher.

Amazon plans AI content marketplace for publishers

• Amazon is building a new marketplace where publishers can sell their content to companies developing AI systems, with AWS acting as a middleman between media organizations and AI developers.

• The project moves Amazon away from individual content deals, like its reported $20 million yearly Alexa agreement, toward a standard system that lets business customers access quality content at scale.

• Microsoft announced a similar Publisher Content Marketplace last week, and both companies are now racing to become the main platform where journalism gets licensed for AI training and products.

OpenAI delays Jony Ive AI device to 2027

• OpenAI’s AI hardware device, designed by former Apple design chief Jony Ive and originally expected before the end of 2026, has been pushed back to at least the end of February 2027.

• The delay follows a trademark infringement lawsuit from Google-backed earpiece startup iyO, and OpenAI has now confirmed it won’t use the name “io” for any AI hardware products.

• The screenless, pocket-sized device is reportedly code-named “Dime” or “Sweetpea,” and OpenAI has not yet created any packaging or marketing materials for it, according to court filings.

xAI co-founder departs:

Just a week after SpaceX’s high profile acquisition of xAI, co-founder Tony Wu announced his departure from the Grok makers. On X (where else?), Wu alluded vaguely to his “next chapter” and noted that he’s excited about building at a time when “a small team armed with AIs can move mountains and redefine what’s possible.” Musk launched xAI in 2023 with an initial team of 11 collaborators, but their ranks have been dwindling of late. Igor Babuschkin, Kyle Kosic and Christian Szegedy previously departed, and Greg Yang announced earlier this year that he plans to step back and focus on his health.

Big tech still believe LLM will lead to AGI?

With all the massive spending from big tech on GPUs and data centres, the goal is to train and deploy LLMs?

Haven’t we already plateaued in terms of LLM improvement? Will all these new infrastructures make any improvements?

https://arxiv.org/pdf/2601.23045

10K Claude Desktop Users Exposed by Zero-Click Exploit

A flaw in Anthropic’s Claude Desktop Extensions allows a single malicious calendar invite to trigger zero-click system compromise, potentially exposing over 10,000 users.

10 AI Agent Platforms Every Business Leader Needs To Know

Source: https://www.linkedin.com/pulse/10-ai-agent-platforms-every-business-leader-needs-know-bernard-marr-4doye

Artificial intelligence has moved fast from curiosity to capability, and nowhere is that more visible than in the rise of AI agents. These systems can plan, decide, and act on our behalf, automating real work rather than simply responding to prompts.

The challenge for most organizations is no longer whether AI agents matter, but where to begin. Over the past year, an explosion of platforms has promised to help businesses build, deploy, and manage agentic systems, ranging from beginner-friendly tools to powerful enterprise frameworks.

Let’s break down some of the most practical and influential platforms available today and show how they can help organizations take their first meaningful steps toward an AI-powered workforce.

Google Vertex And Astra

A great place to start is with the big cloud providers like Google. Its agentic ecosystem is built around the Vertex AI platform, which aims to provide a beginner-friendly environment for designing, building and deploying agents. A strength is its ability to search and process online data in real time, due to its integration with the Google web ecosystem. Astra is a prototype for a universal AI assistant that’s likely to become more ingrained in Google’s agentic toolset in the near future.

Microsoft Copilot Studio

Platforms offered by the cloud giants tend to focus on leveraging their existing strengths, and with Microsoft’s offering, that means deep integration with its Teams and 365 enterprise productivity ecosystems. If you’re looking to build agents for relatively generic use cases involving automating workflows that you already carry out on these widely used platforms, it might be the obvious choice.

Amazon Bedrock AgentCore

Amazon AWS is the world’s most popular cloud provider, and now it lets users unleash agents across its whole suite of features and services. As you’d expect, this means a strong focus on security, always a critical element of any cloud deployment. If your organization is heavily invested in the AWS ecosystem, then it’s a natural starting point for beginner-level agent deployments, thanks to the ease of configuring and managing access to AWS resources.

OpenAI AgentKit

OpenAI, creator of ChatGPT, lets users configure, build, and manage AI agents through its custom GPTs and AgentKit platform. AgentKit provides a comprehensive framework for defining agentic workflows and managing access to third-party datasets and tools. All of this is done through a super user-friendly visual “drag and drop” interface. Another useful feature is Guardrails, a modular safety layer that safeguards against dangerous or unintended behavior.

Salesforce Agentforce

Salesforce is commonly used to manage business customer relationships, and its Agentforce platform is built to automate many of the processes that this involves. This could include sales, marketing and customer service workflows. However, it goes further, capable of creating and managing agents for any tasks that involve calling APIs, connecting and controlling third-party systems and processing end-to-end workflows using external data.

UIPath Studio

UIPath started out as a platform for automating tasks programmatically but has now evolved into an ecosystem for developing and deploying agentic tools. This might be very useful for certain use cases that require the precision of more traditional robotic methods of process automation, combined with the ability to make decisions on the fly provided by AI. A feature that sets it apart is its ability to “see” content on-screen, making it a good option for automating legacy software that might not allow API access to agents directly.

HubSpot Breeze Agent

HubSpot’s Breeze agents are specialized tools for automating CRM tasks like marketing, sales and customer service. As they plug directly into the HubSpot platform, their workflows will already be familiar to many small and medium-sized businesses. They can create and automate campaigns, follow up leads, triage and troubleshoot customer service issues and handle many routine customer interactions. Potentially a great option for smaller organizations looking for “quick wins”.

Zapier Agents

Zapier started as a tool to connect different business and productivity apps through simple automated workflows. Adding agents to the mix means users can now coordinate the activity of thousands of SaaS tools and platforms that Zapier knows how to speak to. Simply describe the workflow you want to build using the Canvas Visual Studio and start chaining your existing apps together to create agentic processes.

QuickBooks AI Agents

Popular accounting package QuickBooks has now integrated its own agents for common routine and time-consuming tasks such as chasing invoice payments, reconciling accounts and preparing cash-flow forecasts. As QuickBooks customers will typically already have all of their financial data in the platform, this can often be a quick and easy win for smaller to medium-sized businesses looking to implement their first AI agent workflows.

Replit Agent 3

Replit is a “vibe coding” platform designed to simplify the process of creating anything from web pages to fully featured apps. Its agentic approach allows it to automate code generation, testing, debugging, refactoring and deployment, combining the functionality of a coding integrated development environment with an AI assistant. While it’s more technical than some of the more specialized tools covered here, the variety of projects it can be used for is virtually unlimited.

What This Means For Business Leaders

AI agents are quickly becoming a practical way to automate work, scale expertise, and unlock new levels of productivity across the organization. The platforms highlighted here show that getting started no longer requires deep technical skills, but it does require clarity on where agents can create the most value, along with a willingness to experiment, learn and iterate as this technology continues to mature.

What Else Happened in AI today?

Isomorphic Labs unveiled IsoDDE, a drug design engine that more than doubles AlphaFold 3 on benchmarks and can spot drug targets from a protein’s genetic code.

Alibaba’s Qwen team released Qwen-Image-2.0, a new unified image generation and editing model with upgraded text rendering, realism, and speed.

Anthropic safeguards research lead Mrinank Sharma resigned, writing in a farewell letter that the company “constantly faces pressures to set aside what matters most”.

OpenAI is reportedly dropping the “io” branding for its upcoming AI hardware device after a trademark lawsuit from audio startup iyO.

Runway raised $315M in Series E funding at a $5.3B valuation, with backing from Nvidia, Adobe, and AMD to pre-train its next generation of world simulation models.

Full Audio at

This episode of AI Unraveled is made possible by:

AIRIA (🛑 Stop fearing Shadow AI. Orchestrate it)

MERCOR: Remote AI Jobs → Here

🚀 Welcome to AI Unraveled (February 10th, 2026): Your strategic briefing on the business, technology, and policy reshaping artificial intelligence.

Today, we analyze Apple’s confirmed entry into AI hardware, with Tim Cook signaling new product categories and a Jony Ive x OpenAI collaboration looming. We also deep dive into the “OpenClaw” controversy—is it a security nightmare or the stress test the industry needs? Plus, a look at the Super Bowl ad fallout, SpaceX’s pivot to the Moon, and why AI might actually be making you work harder.

Strategic Pillars & Key Topics:

🍎 Apple & Hardware

• The Gambit: Tim Cook confirms Apple is entering the AI hardware race, calling it a “profound opportunity.”

• The Roadmap: Expect a Gemini-powered Siri in iOS 27 and potentially smart glasses or AI earbuds (rumored via Jony Ive & OpenAI).

🛡️ Security & Open Source

• The OpenClaw Paradox: The viral open-source agent (aka MoltBot) has racked up massive security incidents. While vendors call it a “nightmare,” defenders argue it’s exposing architectural flaws that proprietary vendors hide.

• The Lesson: Security through transparency—2 million visitors and 180,000 researchers are stress-testing agentic AI in real-time.

🏈 Super Bowl & Culture

• The Ad Wars: Anthropic and OpenAI traded barbs over ad-supported AI, while Svedka aired the first AI-generated spot.

• The Verdict: Viewers reacted poorly to the AI ad blitz, citing low “watchability” and “likeability.”

• AI.com: Crypto.com CEO Kris Marszalek launched an autonomous agent platform on the $70M domain during the game.

📉 Business & Policy

• ChatGPT Growth: Sam Altman confirms user growth is back to exceeding 10% monthly.

• EU vs. Meta: Regulators charge Meta with antitrust violations for blocking rival AI (like Perplexity) on WhatsApp.

• Goldman Sachs: Engineers spent six months building Anthropic Claude agents for back-office automation.

🚀 Innovation & Science

• Moon Over Mars: SpaceX shifts focus to a self-growing Moon city (10-day launch cadence vs. 26 months for Mars).

• Waymo World Model: Using Google’s Genie 3 to simulate rare driving events like tornadoes and elephants.

• Workforce Study: UC Berkeley research suggests AI tools increase work intensity and hours, contradicting the idea that AI reduces workload.

AI Unraveled Partner Offers

🛑 Stop fearing Shadow AI. Orchestrate it.

Your employees are using AI whether you approve it or not. The problem isn’t the innovation; it’s the lack of visibility.

Airia is the “Control Plane” for your entire enterprise AI stack. It allows you to say “Yes” to innovation without compromising security or blowing up your API budget.

Why we are partnering with them:

• Unified Security: Get automated threat detection and governance across all your models and agents.

• Real-Time Cost Control: Stop guessing your token usage. Manage budgets and quotas from a single dashboard.

• Model Agnostic: Whether your devs want low-code or pro-code tools, they can prototype safely in a production-like environment.

👉Secure Your Stack: Get the Airia Demo

Keywords:

Apple AI Hardware, OpenClaw Security, MoltBot, Super Bowl AI Ads, SpaceX Moon City, Waymo World Model, OpenAI Jony Ive, Goldman Sachs Anthropic, Meta Antitrust, AI Workload Study

🚀 Reach the Architects of the AI Revolution

Want to reach 60,000+ Enterprise Architects and C-Suite leaders? Download our 2026 Media Kit and see how we simulate your product for the technical buyer: https://djamgamind.com/ai

Connect with the host Etienne Noumen,

LinkedIn: https://www.linkedin.com/in/enoumen/

⚗️ PRODUCTION NOTE: We Practice What We Preach.

AI Unraveled is produced using a hybrid “Human-in-the-Loop” workflow. While all research, interviews, and strategic insights are curated by Etienne Noumen, we leverage advanced AI voice synthesis for our daily narration to ensure speed, consistency, and scale. We are building the future of automated media—one episode at a time.

Apple’s 2026 AI device gambit is taking shape

Apple’s playbook: enter late, do it better, make it mainstream. AI hardware could be next.

In an internal meeting, Apple CEO Tim Cook spoke to employees about the importance of the AI moment, calling it “one of the most profound opportunities of our entire lifetime,” and highlighted the company’s strong position to deliver meaningful AI features to customers, according to a Bloomberg report. One way the company plans to do so: hardware.

“There will be new categories of products and services that are enabled through AI, and we’re extremely excited about that,” said Cook. “We’re excited about the opportunities that it opens for Apple.”

This would be a different approach for Apple, which, to date, has had limited success competing in the AI race. Despite announcing an AI overhaul for Siri at WWDC 2024 to make the AI assistant a more advanced personal intelligence system, it has yet to deliver on the promise.

Yet, this year, Apple is preparing to turn things around, with the first step being its multi-year agreement with Google to use Gemini’s AI models to power its next-gen version of Siri. Apple also has plans to turn Siri into the company’s first artificial intelligence chatbot with the launch of iOS 27, according to another Bloomberg report. While these voice assistant updates have been highly anticipated, delving into hardware would open a new frontier.

AI devices are gaining traction, with smartglasses being the first big hit. Shipments grew 110% YoY in the first half of 2025, according to Counterpoint Research. Google DeepMind’s Demis Hassabis recently said the Gemini-powered Samsung smartglasses could arrive as soon as this summer.

While other AI wearable form factors, such as rings, pins, and more, have emerged, none have taken off with consumers. OpenAI is developing its much-anticipated AI hardware device in collaboration with Jony Ive, Apple’s former design lead. That may be the first real stab at a mainstream AI device. The latest reports suggest it will be a smart AI earbud.

Still, Apple has a huge advantage in hardware. It knows how to build high-quality, well-designed devices, and it has a loyal early-adopter base ready to buy. It can also pair the device with a valuable set of services for music, fitness, and more. Most importantly, it has a track record of popularizing new device categories. However, that may depend on Apple taking a creative leap and not sticking to a safe choice, such as an Alexa-like AI smart speaker with a screen that has reportedly been in the works.

OpenClaw, or MoltBot, or Clawdbot is the best thing to happen to Al security this year.

Source: Reddit

Yes, the one that’s racked up more security incidents in a fortnight than some vendors have in their entire history. That one.

I’ve been watching the security community’s reaction closely. Every major vendor has published their take. Cisco called it a nightmare. Palo Alto said it signals a crisis. Trend Micro warned of invisible risks. You’d think someone had plugged an unpatched Windows XP box directly into the internet. In a hospital. Running the ventilators.

Deep breaths, everyone. They’re missing something.

OpenClaw is open source. 2 million visitors in a single week, one of the fastest growing projects in GitHub’s history. Developers buying Mac Minis to run it from their spare rooms. Nobody should be running this against production systems or corporate email, and even the project’s own documentation describes it as an experiment not intended for most non-technical users. The creators are being honest about what this is. Which, in this industry, is practically unheard of.

And experiments are exactly how security gets better.

A researcher found that clicking a single malicious link could hijack an OpenClaw instance in milliseconds, bypassing every sandbox and safety guardrail the project had built. That’s a critical lesson: agentic Al safety controls designed to contain prompt injection don’t protect against architectural vulnerabilities in the control plane. Better to learn that on an open source hobby project than on your enterprise vendor’s agent platform. The 400 malicious skills published to its marketplace showed that Al skill registries have the same supply chain problems as traditional software package repositories, but with broader execution privileges.

The early days of cloud computing looked exactly like this. Researchers poking at S3 buckets, finding everything wide open, the industry collectively losing its mind. There was plenty of real damage along the way. And yet somehow we survived, built proper controls, and got on with things.

OpenClaw is doing the same thing for agentic Al. Every exposed gateway, every prompt injection chain, every malicious skill is teaching the security community what agentic threat models actually look like in practice rather than in framework documents. Real CVEs, real attack chains, real mitigation patterns against a system people can actually inspect, rather than a black box vendor product.

Everyone’s worried about the open source project with 180,000 people scrutinising every flaw. Meanwhile, enterprise agent platforms ship with the same architectural problems. You just don’t get to see them.

Your enterprise agent vendor has a trust page and a SOC 2 badge. OpenClaw has 180,000 researchers actually breaking things. Which one do you think finds the problems first?

Sam Altman says ChatGPT is “back to exceeding 10% monthly growth”

• OpenAI CEO Sam Altman told employees that ChatGPT has returned to exceeding 10% monthly growth, signaling a rebound in user activity for the company’s artificial intelligence chatbot.

• The update was shared internally, with Altman communicating the growth milestone directly to OpenAI staff rather than announcing it through a public channel or press release.

• The phrasing “back to” suggests ChatGPT’s monthly growth had previously dipped below the 10% mark before recovering to its current pace, though specific earlier figures were not provided.

EU warns Meta to open WhatsApp to rival AI chatbots

• The European Union has formally charged Meta with antitrust violations for blocking rival AI chatbots like ChatGPT and Perplexity from WhatsApp, and regulators may force the company to reverse the policy.

• Meta’s policy, effective January 15, 2026, prohibits AI providers from using the WhatsApp Business API when artificial intelligence is the primary service, making Meta AI the sole assistant for three billion users.

• Italy already forced Meta to suspend the ban for Italian phone numbers, Brazil opened its own investigation after complaints from AI companies, and Meta could face fines up to 10 percent of global revenue.

SpaceX shifts focus to self-growing Moon city

• SpaceX is now prioritizing the construction of a self-growing city on the Moon over Mars, with Elon Musk saying on X that the Moon is faster and could be achieved in under 10 years.

• SpaceX can launch to the Moon every 10 days with a two-day trip, compared to Mars launches every 26 months and six-month travel times, letting the company iterate much faster.

• Musk also connected the recent xAI acquisition to this effort, combining AI, rockets, internet satellites, and social media to build working Moon systems like bases and factories that don’t rely on Earth.

“Goldman Sachs taps Anthropic’s Claude to automate accounting, compliance roles” – CNBC

https://www.cnbc.com/2026/02/06/anthropic-goldman-sachs-ai-model-accounting.html

This part is interesting:

Embedded Anthropic engineers have spent six months at Goldman building autonomous systems for time-intensive, high-volume back-office work.

Because OpenAI also announced this week a service called Frontier that includes Forward Deployed Engineers.

These model companies are selling enterprise services now.

Study: AI makes you work harder

New research seemingly confirms what many tech underlings have long suspected: AI tools frequently intensify workloads rather than reducing or eliminating them altogether. Researchers from UC Berkeley’s Haas School of Business spent 8 months reviewing generative AI tools and how they changed work habits at an unnamed US tech company with around 200 employees. They found that — while using the tools — employees tended to work at a faster pace, for long hours, on a broader scope of tasks, all without ever being asked by managers to put in extra effort. The results suggest that the AI tools empowered employees, making them feel more capable of taking on more responsibilities, and thus making their work more rewarding. It’s something of a ‘best case scenario’ for C-suite execs investing in more AI tools, and a fresh counter-argument to that infamous and constantly-cited MIT survey from last August concluding that 95% of enterprise AI pilots fail.

Neara makes 3D models of power networks

The Australian startup creates “physics-enabled digital twins” of critical infrastructure, such as electrical grids. These finely detailed 3D models help utility companies, repair crews, and other insiders predict how different kinds of facilities will respond to disasters like storms and fires, improving safety while saving time and money on extensive, pain-staking inspections. This was already a useful technology as service providers around the globe brace for more extreme weather events, but also, the explosion of data centers has created fresh demand for exactly these kinds of predictive insights. Cities need to ensure that their grids will be able to handle surging demand for power, and what better way to do that than with some finely detailed 3D models? This week, Neara raised a fresh $63 million USD Series D round this week, led by TCV.

AI Super Bowl ads missed the mark

In between the Seahawks demolishing the Patriots and Bad Bunny climbing a telephone pole in his dynamic and elaborate halftime show performance, tech companies got plenty of airtime on Super Bowl Sunday.

Some of AI’s biggest names touted their models and products in ads in Super Bowl LX, including Google, Microsoft, Salesforce, Meta and, of course, AI’s biggest rivalry: Anthropic and OpenAI. Viewers generally reacted badly to the AI ads, which scored very low for attention, likeability and watchability.

Here’s a rundown of what we caught:

• As shown last week, Anthropic decided to use two ad slots to malign the integration of ads in chatbots, including a clip where someone asks, “Can I get a six-pack quickly?” and, upon providing his height and weight, elicits an advertisement for insoles that “help short kings stand tall.”

• OpenAI, meanwhile, tailored one of its ads around Codex, the company’s coding agent. The ad’s tagline, “you can just build things,” included shots of people playing with circuit boards, installing Linux on an old computer and building robots. The other focused on ChatGPT, centering around a farmer using the chatbot to track crops.

Kate Rouch, CMO of OpenAI, fired back at Anthropic’s shots at its in-chat advertising decision, telling AdWeek, “The way that Anthropic’s ads are constructed is just not true to how ads will appear in the free ChatGPT. Free access to this technology is critical. Your ability to pay isn’t the thing that determines if you have access to AI or not.”

Notably absent from the game were ads from AI search platform Perplexity and Elon Musk-owned xAI. Musk, along with OpenAI CEO Sam Altman and Meta CEO Mark Zuckerberg, did earn shoutouts in the debut of AI.com, an AI agent platform founded by Crypto.com CEO Kris Marszalek after he purchased the domain name for $70 million, plainly stating in its 30-second slot that “AGI is coming.” Additionally, spirits brand Svedka unveiled its first primarily AI-generated ad, featuring dancing robots throwing around bottles of its vodka and spilling drinks throughout their circuitry.

Waymo gives world models a real-world purpose

As world models gain momentum, Waymo might be putting this lofty tech to good use.

On Friday, the robotaxi firm unveiled the Waymo World Model, a generative model for “large-scale, hyper-realistic” autonomous driving simulation. Waymo’s model is built on Google’s Genie 3, the latest iteration of its world model series, released in August.

The model’s architecture allows the engineers to better modify and control the scenes it generates using language, allowing for everything from time-of-day or weather condition changes to entire synthetic scene generation with simple language prompts. The model can also convert authentic dashcam videos into synthetic scenes for training.

With the Waymo World Model, the company leveraged Genie’s ability to generate photorealistic, interactive 3D environments, applying it specifically to the “driving domain.” This allows the world model to generate rare events, including:

• An elephant appearing in the middle of the road;

• Being chased down a freeway by a tornado;

• And driving through a tropical city that happens to be covered in snow.

This kind of training could help prepare Waymo to scale across new locations and driving environments, potentially even allowing the robotaxis to better handle natural disasters.

The company has faced scrutiny for its safety practices in recent weeks after one of its vehicles struck and injured a child in California. In a hearing this week, Waymo’s Chief Safety Officer Mauricio Peña acknowledged that the company relies on remote operators overseas for “guidance in certain situations,” but these operators do not pilot the vehicles themselves.

“By simulating the ‘impossible’, we proactively prepare the Waymo Driver for some of the most rare and complex scenarios,” The company said in its announcement.

Waymo’s model adds to growing enthusiasm for world models. Google last week released Project Genie, an “experimental research prototype” built on Genie 3 that lets users create and explore virtual worlds and video game-like environments. World model startups from AI pioneers like Fei-Fei Li and Yann LeCun are also in talks for massive funding rounds as investors search for the next big thing after LLMs.

Claude Opus Uncovers Hundreds of Open-Source Flaws

Anthropic said Claude Opus 4.6 identified more than 500 previously unknown high-severity flaws in widely used open-source libraries.

Because libraries like Ghostscript, OpenSC, and CGIF are deeply embedded across enterprise stacks, a single flaw can cascade across many downstream products.

Claude analyzed code and commit histories to uncover logic and memory flaws missed by traditional testing, with findings validated by Anthropic’s Frontier Red Team.

Use SBOMs to track dependencies, prioritize fixes based on real exploitability, and sandbox high-risk components while monitoring runtime behavior for signs of supply-chain exploitation.

What Else Happened in AI today?

Generative AI powered by nucleic acid language model enables one-round evolution of RNA aptamers. [Link]

Researchers from the University of Southern Denmark have developed a bioinspired soft robot that mimics the movement of limbless animals. [Link]

ByteDance’s new Seedance 2.0 supposedly ‘surpasses Sora 2′. [Link]

MIT Sports Lab researchers are applying AI technologies to help figure skaters improve. [Link]

Full Audio at

🚀 Welcome to AI Unraveled (February 9th, 2026): Your strategic briefing on the business, technology, and policy reshaping artificial intelligence.

Today, we break down the AI takeover of Super Bowl LX, where tech giants and startups turned the Big Game into a battlefield for your trust. We also analyze the record-breaking $70 million purchase of AI.com, the “AI Washing” trend in corporate layoffs, and the rapid collapse of the viral Moltbook platform.

Strategic Pillars & Key Topics:

🏈 The Super Bowl AI Blitz

• The Ad War: AI commercials dominated the broadcast, with Anthropic attacking OpenAI’s ad policies, Google pushing Gemini, Meta showcasing “athletic intelligence” in smart glasses, and Amazon hyping Alexa+.

• SVEDKA’s AI Spot: The vodka brand aired the first primarily AI-generated Super Bowl commercial, reviving its robotic mascot with AI choreography.

• Startup Spend: Companies like Base44, Genspark, and Wix joined the fray, pushing tech’s share of ad space to 10%.

💰 The $70 Million Domain

• AI.com Sold: Crypto.com CEO Kris Marszalek bought the domain for $70 million—the largest domain sale in history.

• The Pitch: The site launched during the Super Bowl as a platform for autonomous personal agents that can trade stocks and manage calendars, aiming to “accelerate AGI.”

📉 Culture & Ethics

• Moltbook Collapse: The viral AI social network “Moltbook” is already history, devolving into a toxic mix of spam and security leaks within hours. The lesson: humans are the “weakest link” in the hype cycle.

• AI Washing: Experts accuse US companies of blaming AI for job losses to mask other factors like overhiring and profit maximization.

• Figure Skating AI: The International Skating Union is using AI computer vision to resolve judging controversies by analyzing jump rotation and height in real-time.

⚡ Rapid Fire Updates

• Goldman Sachs x Anthropic: The bank is using Claude agents to automate accounting and compliance.

• Axiom’s Math Breakthrough: An AI independently solved an unsolved math problem and wrote the verified proof.

• Claude Opus “Fast Mode”: Anthropic releases a 2.5x faster version of its new model (at a higher cost).

Keywords: Super Bowl AI Ads, AI.com $70M Sale, Moltbook Collapse, AI Washing, Anthropic vs OpenAI, SVEDKA AI Ad, Figure Skating AI Judging, Goldman Sachs Anthropic, Axiom Math Proof, Kris Marszalek

🚀 Reach the Architects of the AI Revolution

Want to reach 60,000+ Enterprise Architects and C-Suite leaders? Download our 2026 Media Kit and see how we simulate your product for the technical buyer: https://djamgamind.com/ai

Connect with the host Etienne Noumen, Senior Software Engineer and passionate Soccer dad from Canada

LinkedIn: https://www.linkedin.com/in/enoumen/

Youtube: https://youtube.com/@enoumen

X: https://twitter.com/enoumen

Email: etienne_noumen@djamgamind.com

AI ads steal the show at Super Bowl LX

AI companies/products took over Super Bowl LX with a wave of commercials, featuring spots from Anthropic, OpenAI, Google, Amazon, Meta, and startups that turned the Big Game into a battlefield for consumer AI attention.

The details:

• Anthropic made its SB debut with its viral campaign against ads in AI, whose launch kicked off a viral feud with OpenAI and Sam Altman last week.

• Vodka brand SVEDKA ran what it called the first primarily AI-generated Super Bowl ad, reviving its robotic mascot with AI-trained dance moves.

• Meta showcased “athletic intelligence” via its AI glasses, Amazon pushed the new Alexa+, and Google highlighted Gemini and Nano Banana.

• Other AI-related spots included Base44, Genspark, Ramp, Rippling, and Wix, with tech’s Super Bowl ad share reportedly increasing to around 10%.

Why it matters: Major AI players showed up to the SB fighting for the same thing — to be the default assistant, agent, or device that the masses actually trust as AI weaves into every corner of daily life. Svedka’s use of AI to generate its spot also shows the tech influencing both sides of the screen — what’s being sold AND how it’s made.

Crypto.com founder drops $70M on AI.com

Image source: ai.com

Crypto.com co-founder and CEO Kris Marszalek revealed he purchased the AI.com domain for $70M, the largest domain sale ever, debuting the site as an autonomous AI agent platform for consumers with a Super Bowl commercial.

The details:

• The platform promises a personal AI agent that can trade stocks, manage calendars, send messages, and automate workflows with no technical setup.

• The record-breaking deal nearly doubled the prior domain record set by voice.com‘s $30M sale in 2019.

• Marszalek envisions a network of agents that autonomously build new capabilities and share upgrades across users, “accelerating the advent of AGI”.

Why it matters: That’s certainly one way to make a big splash — with both the largest domain sale ever (admittedly for an excellent one) and a primetime entrance with a Super Bowl ad. But a strong marketing play and a tool that actually competes with the flurry of agentic upgrades from established, frontier labs are two very different beasts.

US companies accused of ‘AI washing’ in citing artificial intelligence for job losses

https://www.theguardian.com/us-news/2026/feb/08/ai-washing-job-losses-artificial-intelligence

While AI is having an impact on the workplace, experts suggest tariffs, overhiring during the pandemic and simply maximising profits may be bigger factors.

Figure Skating turns to AI to tackle judging controversies

“After decades of judging controversies, figure skating is turning to artificial intelligence (AI) and computer vision technology to try to bring greater consistency and transparency to how performances are scored.

The International Skating Union (ISU) has been testing high-resolution camera systems at competitions over the past two years that use AI to track skaters’ movements and analyse technical elements such as jump rotation, height, distance travelled and spin positions in real time.”

https://www.reuters.com/sports/figure-skating-turns-ai-tackle-judging-controversies-2026-02-08/

Moltbook is already history

What we will remember about Moltbook:

– the incredible speed at which the hype spread

– that it took only hours to turn into a toxic mix of spam and security nightmares

– whatever can be manipulated, will be

– the humans are the „weakest link“ in this hype chamber. We want to believe the extremes and overreact very quickly.

– this stuff is no longer a fit for human timelines of mental processing

– incredible how gullible we are, giving a sloppily vibe-coded tool credentials (to share almost freely)

– moltys are a reflection of humans- for now.

– there will be more moltbooks, likely more extreme

What else Happened in AI today

Adapt just announced the future of AI for work – an always-available co-worker that understands business context and works across tools. See the vision and fundraising deck.*

Anthropic released a ‘fast mode’ for its new Claude Opus 4.6 model, which delivers 2.5x faster responses at drastically higher token costs.

Goldman Sachs revealed it has been working with Anthropic over the last six months to build AI agents that automate accounting, compliance, and client onboarding.

xAI co-founder Igor Babuschkin praised Claude Opus 4.6’s physics capabilities, saying that a “Claude Code moment for research” may be approaching.

Axiom announced that its AI independently solved an unsolved math problem and wrote its own verified proof, a first for AI in advanced mathematical research.

Full Audio Here

🚀 Welcome to AI Unraveled (Weekly Rundown: Feb 1st – Feb 8th, 2026): Your strategic briefing on the business, technology, and policy reshaping artificial intelligence.

This week, the AI landscape shifted dramatically with SpaceX’s $1.25 trillion acquisition of xAI and dueling model releases from OpenAI and Anthropic that redefine coding and agentic capabilities. We also track a massive $300 billion software stock crash, Amazon’s $200 billion infrastructure bet, and the security meltdown of the viral “Moltbook” agent social network.

Key Topics:

💰 The Trillion-Dollar Consolidation

• SpaceX Acquires xAI: A historic merger creates a $1.25 trillion private giant. The goal? Launching data centers into orbit to solve Earth’s energy crisis.

• Amazon’s $200B Bet: Amazon commits $200 billion to AI infrastructure in 2026, causing its stock to dip but signaling a massive long-term play.

• Alphabet’s $185B Plan: Google’s parent company isn’t far behind, pledging up to $185 billion for data centers and servers.

⚔️ The Model Wars: Dueling Releases

• OpenAI GPT-5.3-Codex: A self-improving coding model that helped debug its own training. It’s faster, smarter, and rated “High Risk” for cybersecurity.

• Anthropic Opus 4.6: Features “Agent Teams” for collaborative work, a 1M token context window, and direct integration into Microsoft Office.

• Sam Altman vs. Anthropic: A public spat erupts after Anthropic mocks ChatGPT’s ads in a Super Bowl commercial.

📉 Markets & Disruption

• The SaaS Massacre: Software stocks (LegalZoom, Thomson Reuters) plunge, wiping out $300 billion in value as investors realize AI agents can replace traditional software tools.

• Microsoft Copilot Crisis: Usage drops to 11.5% as users flock back to ChatGPT and Gemini due to confusion and bugs.

🤖 Agents & Security

• Moltbook Exposed: The viral “AI social network” leaked 1.5 million API keys and was revealed to be mostly humans faking agent autonomy.

• OpenAI “Frontier”: A new platform treating AI agents like employees, complete with onboarding and performance reviews.

• Gemini “Screen Automation”: Google’s AI will soon tap and scroll your phone to book rides—but you are liable for its mistakes.

🍎 Apple & Consumer AI

• Apple CarPlay AI: Apple opens CarPlay to third-party voice apps like ChatGPT and Claude, ending Siri’s monopoly.

• Apple vs. Chipmakers: Nvidia outbids Apple for TSMC capacity, forcing the iPhone maker to look elsewhere for chip manufacturing.

Keywords

SpaceX xAI Merger, GPT-5.3-Codex, Claude Opus 4.6, Amazon AI Spending, SaaS Stock Crash, Moltbook Breach, OpenAI Frontier, Apple CarPlay AI, Waymo Genie 3, Sam Altman Super Bowl Ad.

🚀 Reach the Architects of the AI Revolution

Want to reach 60,000+ Enterprise Architects and C-Suite leaders? Download our 2026 Media Kit and see how we simulate your product for the technical buyer: https://djamgamind.com/ai

Connect with the host Etienne Noumen, Senior Software Engineer and passionate Soccer dad from Canada

LinkedIn: https://www.linkedin.com/in/enoumen/

Youtube: https://youtube.com/@enoumen

X: https://twitter.com/enoumen

Apple might let you use ChatGPT from CarPlay

• Apple is working on letting third-party AI voice apps like ChatGPT, Claude, and Gemini run directly inside CarPlay, ending Siri’s long reign as the only voice assistant allowed in the interface.

• Drivers won’t be able to replace Siri’s wake word or dedicated button; instead, they will need to launch a specific third-party AI app first, which can automatically turn on voice mode when opened.

• Apple is also building its own Siri improvements, including a feature called “World Knowledge Answers” that fetches and summarizes online information, with that upgrade expected to arrive later this year.

Crypto.com founder buys AI.com in ‘largest domain purchase in history’

• Crypto.com co-founder and CEO Kris Marszalek bought the AI.com domain for $70 million and launched a new AI agent platform under that brand, with a Super Bowl commercial planned for Sunday.

• The company says AI.com will let users create a private, personal AI agent that can send messages, build projects, trade stocks, and even update a dating profile on their behalf.

• The $70 million deal is called the largest domain purchase in history, though undisclosed sales may have been bigger, and Cars.com was once listed as an intangible asset worth $872.3 million.

Waymo uses Genie 3 to simulate rare driving scenarios

• Waymo built a new simulation tool called the Waymo World Model, powered by Google DeepMind’s Genie 3, to generate realistic driving scenarios — like elephants or tornadoes — that its cars have never encountered.

• Unlike most industry simulation models that train only on a company’s own driving data, this system draws on Genie 3’s broad knowledge from pre-training on a massive and diverse video dataset.

• The model generates both camera and lidar data, and engineers can control simulations through driving actions, scene layout changes, or text prompts that describe weather, time of day, or synthetic scenes.

Anthropic launches Claude Opus 4.6

• Anthropic released Claude Opus 4.6, its newest model, featuring a 1M token context window in beta — a first for the company’s Opus-class models — along with stronger coding and agentic task performance.

• Opus 4.6 scores highest on benchmarks like Terminal-Bench 2.0, Humanity’s Last Exam, and BrowseComp, and beats OpenAI’s GPT-5.2 by around 144 Elo points on the GDPval-AA knowledge work evaluation.

• New developer tools include adaptive thinking, four effort levels, context compaction for longer-running agents, and agent teams in Claude Code that let multiple agents work in parallel on tasks.

Amazon plans $200B in AI spending

• Amazon plans to spend $200 billion in capital expenditures in 2026, up from the $125 billion it had estimated for 2025 and well ahead of analyst expectations.

• CEO Andy Jassy tied the $200 billion AI spending plan to strong demand for existing offerings and opportunities in AI, custom chips, robotics, and low earth orbit satellites.

• Amazon’s stock sunk more than 10% in after-hours trading after the $200 billion spending plan was revealed alongside missed Wall Street profit expectations in its Q4 earnings.

Apple scales back AI health coach plans

• Apple has scaled back Project Mulberry, its planned AI-powered health coach for the Health app, after leadership changes in the company’s health and AI organizations prompted a reassessment of the initiative.

• Services chief Eddy Cue, who now oversees the health team, told colleagues that Apple’s plan wasn’t compelling enough compared to rivals like Oura Health and Whoop and their iPhone apps.

• Some Project Mulberry features may still ship individually over time, including one that would use the iPhone’s camera to analyze how a person walks, plus Siri health queries in iOS 27.

OpenAI launches GPT-5.3-Codex its self-building coding model

• OpenAI has released GPT-5.3-Codex, a new coding model designed for developers that the company says set record scores on SWE-bench Pro and Terminal-Bench while running 25% faster than its predecessor.

• The model showed a big jump on Terminal-Bench 2.0, scoring 77.3% compared to 64.0% for GPT-5.2-Codex, and on OSWorld-Verified it reached 64.7% versus the prior model’s 38.2%.

• Developers can now steer and interact with GPT-5.3-Codex mid-task without losing context, and the model provides frequent progress updates so users can ask questions and guide it in real time.

Meta tests a standalone app for its AI-generated ‘Vibes’ videos

• Meta confirmed it is testing a standalone app for Vibes, its AI-generated short-form video feature that lets people create, share, and browse AI videos in a dedicated feed.

• Vibes previously lived inside the Meta AI app, but spinning it out positions the product as a more direct competitor to Sora, OpenAI’s AI video and social app.

• Meta also plans to explore freemium access and paid subscriptions for Vibes video creation, letting users unlock additional monthly video creation opportunities in the coming months.

Alphabet plans record AI spending after strong earnings

• Alphabet announced it will spend between $175 billion and $185 billion on AI infrastructure this year, far exceeding the $119.5 billion analysts predicted and more than its combined spending over the past three years.

• CEO Sundar Pichai said the AI investments are necessary, stating the company is seeing AI investments and infrastructure drive revenue and growth across search, cloud, and other products.

• About 40% of technical infrastructure investments go toward data centers and networking equipment, while 60% go toward servers, as Alphabet races to compete against Amazon, Microsoft, and OpenAI.

Sam Altman clashes with Anthropic over Super Bowl ads

• OpenAI CEO Sam Altman and Anthropic got into a public fight after Anthropic released Super Bowl ads that mocked ChatGPT’s upcoming ad-supported free tier by showing chatbots inserting awkward product promotions into personal conversations.

• Altman admitted the ads were funny but called Anthropic “dishonest” and “authoritarian,” saying OpenAI would never insert ads into conversations, though the company has said it plans to show ads based on chat content.

• Altman claimed Anthropic only serves rich people, but both companies offer similar free and paid tiers, and both enforce content policies, making his accusations seem like an overreaction to a cheeky commercial.

OpenAI unveils Frontier, a platform to build ‘AI-coworkers’

• OpenAI launched Frontier, an enterprise platform that lets companies build, deploy, and manage AI agents that can run other software like Salesforce and Workday.

• The platform connects databases, business tools, and internal applications, letting AI agents and human employees work together with the same data, tools, and security controls.

• Fortune 500 companies including Intuit, State Farm, Thermo Fisher, and Uber have signed up as initial customers, while investors worry Frontier could threaten traditional SaaS providers.

Meta’s internal memo signals AI comeback after rocky year

• An internal memo reveals that Meta has completed pretraining on its new AI model codenamed “Avocado,” marking a potential recovery after the company’s troubled 2025 with its Llama 4 release.

• Avocado beats the best freely available base models and competes with leading fully trained models in knowledge, visual perception, and multilingual performance, despite not yet going through post-training.

• The new model is ten times more efficient than Maverick and a hundred times more efficient than Behemoth, with Meta planning capital spending of $115 to $135 billion for 2026.

Alexa+ could integrate ChatGPT

• Amazon is reportedly in talks to integrate ChatGPT into Alexa+, which could mean the voice assistant would be powered by both Anthropic’s Claude and OpenAI’s technology after a potential $50 billion investment.

• The timing is notable because Amazon just opened Alexa+ to all US users after nearly a year of limited testing, and Apple is about to launch an upgraded Siri powered by Google’s Gemini.

• The deal could include other elements beyond Alexa+, such as access to OpenAI’s Codex, getting OpenAI to use Amazon’s Trainium chips, or opening Amazon’s store to OpenAI’s agents.

The AI disruption in software is here

• Fears that new AI tools will replace traditional software wiped around $300 billion from the market value of software and data stocks on Tuesday, hitting companies across legal, finance, and tech sectors.

• The selloff began after Anthropic announced new legal tools for its Cowork assistant, sending shares of Thomson Reuters, Legalzoom.com, and London Stock Exchange down more than 12% each.

• Private-funds firms like Ares, KKR, and Blue Owl also dropped sharply because they invested heavily in software companies over the past decade, leaving them exposed to AI disruption risk.

Microsoft Copilot is running into big problems

• Microsoft’s Copilot chatbot, a key part of CEO Satya Nadella’s plan to make the company AI-first, is struggling with confusing brand positioning, interoperability problems, and falling user preference compared to ChatGPT and Google’s Gemini.

• Survey data from Recon Analytics shows that the share of Copilot subscribers who use it as their primary option dropped from 18.8% to 11.5% between July and late January, while Google’s Gemini rose from 12.8% to 15.7%.

• Microsoft has sold only 15 million corporate Copilot seats out of 450 million paid 365 seats, and some companies use just 10% of the Copilot subscriptions they pay for, according to Citi Research analysts.

Anthropic says ‘Claude will remain ad-free,’ unlike ChatGPT

• Anthropic announced that its AI assistant Claude will stay ad-free, setting it apart from competitors like ChatGPT, because the company believes advertising would conflict with Claude’s role as a helpful thinking tool.

• The company argues that AI conversations are different from search engines because users share sensitive or personal information, and ads could create pressure to steer responses toward transactions rather than genuine help.

• Anthropic plans to fund Claude through enterprise contracts and paid subscriptions instead, while exploring lower-cost tiers and features like agentic commerce where Claude handles purchases only when users ask.

Intel will start making GPUs

• Intel CEO Lip-Bu Tan announced at the Cisco AI Summit that the company will start producing graphics processing units (GPUs), chips used for gaming and training artificial intelligence models.

• The project will be led by Kevork Kechichian, executive vice president and general manager of Intel’s data center group, with Eric Demers, a former Qualcomm senior vice president, also hired for the effort.

• The initiative appears to be in early stages, with Tan saying Intel plans to develop its strategy around customer demands, marking a notable expansion from the company’s traditional focus on CPUs.

AWS CEO says orbital data centers ‘pretty far’ from reality

• AWS CEO Matt Garman said that data centers in space are “pretty far” from becoming real, despite startups and Amazon founder Jeff Bezos exploring the idea through his rocket company Blue Origin.

• The growth of artificial intelligence demands huge amounts of computing power and cooling, which strains land-based data centers and has pushed cloud companies to consider alternatives like sending equipment to orbit.

• Garman explained that the cost of launching payloads into space is massive, and there are not enough rockets to send the million satellites needed, making the concept not economical right now.

Gemini ‘screen automation’ will place orders, book rides for you

• Google is developing a Gemini feature called “screen automation” that will let the AI place orders and book rides on your phone by scrolling and tapping on your behalf.

• New code in the Google app beta reveals that users must closely monitor what Gemini does because the AI can make mistakes, and users will be held responsible for its actions.

• Google warns that trained reviewers will see screenshots from your device if Keep Activity is on, so users should avoid entering sensitive information or payment details during automation.

SpaceX acquires xAI

• SpaceX has purchased Elon Musk’s AI startup xAI, forming the world’s most valuable private company with a combined value of $1.25 trillion, according to Bloomberg News reporting on the deal.

• Musk wrote that the merger focuses on building space-based data centers, arguing that global electricity demand for AI cannot be met with terrestrial solutions without harming communities and the environment.

• The two companies face different pressures: SpaceX is working to prove Starship can reach the moon and Mars, while xAI competes with Google and OpenAI and burns around $1 billion monthly.

OpenAI launches Codex app for macOS

• OpenAI released a new macOS app for its Codex coding agent, offering a visual interface on top of the Codex CLI agent along with new features like Skills support and Automations for scheduled tasks.

• The app is built with Electron and Node.js, and OpenAI plans to bring it to Windows very soon, with Automations currently limited to running only when your laptop is powered on.

• OpenAI hints that Codex works as a general agent beyond just coding, similar to how Claude Code rebranded to Cowork to cover broader knowledge work, though OpenAI may keep the Codex name.

OpenAI shifts focus from research to ChatGPT

• OpenAI is moving away from long-term research and putting more resources into improving ChatGPT, a change that has pushed several senior employees to leave the company in recent months.

• Ten current and former staff members confirmed the pivot, with some saying teams working on non-language-model projects like Sora and DALL-E felt undervalued and lacked resources for their work.

• The shift followed CEO Sam Altman’s December alert about improving ChatGPT after Google’s Gemini 3 model outperformed OpenAI’s in independent tests and Anthropic’s Claude improved its code-generation features.

AI companies are outbidding Apple

• AI chip companies like Nvidia are now outbidding Apple for manufacturing capacity at TSMC, the world’s leading chip foundry, pushing Apple to search for other places to build its processors.

• Apple is exploring whether rival foundries could manufacture lower-end processors, and analyst Ming-Chi Kuo believes Intel could make a lower-end M7 chip in the USA as soon as 2027.

• Apple has a history of using multiple suppliers, including Samsung for earlier A-series chips, and this shift likely marks the end of its exclusive deal with TSMC that started around 2014.

SpaceX seeks approval for 1 million AI satellites

• SpaceX has filed an application with the FCC to launch one million satellites into Earth’s orbit, designed to serve as orbital data centers that support growing artificial intelligence infrastructure needs.

• The company claims that terrestrial data center resources are becoming exhausted, and these solar-powered satellites could offer better cost and energy efficiency with lower environmental impact than ground-based options.

• SpaceX may not deploy all one million satellites even if approved, since Earth’s orbit is already crowded, and other companies like Google, Amazon, and xAI have explored similar ideas.

Moltbook flaw exposed private messages and credentials

• A security flaw in Moltbook, a social network built for AI agents, left private messages and user credentials open to anyone who looked at the site’s code.

• Researchers found a misconfigured Supabase database that exposed 1.5 million API authentication tokens, 35,000 email addresses, and private conversations, some of which contained plaintext OpenAI API keys.

• The database also revealed that Moltbook’s 1.5 million registered agents were controlled by just 17,000 human accounts, showing the platform had no way to verify if an “agent” was AI or a human running scripts.

Listen to to Full Audio here

🚀 Welcome to AI Unraveled (February 6th, 2026): Your strategic briefing on the business, technology, and policy reshaping artificial intelligence.

Today is arguably the biggest day in AI development this year. OpenAI and Anthropic have simultaneously released their flagship coding and reasoning models, with OpenAI’s new model explicitly used to “build itself.” Meanwhile, Alphabet and Amazon have committed a combined $385 billion to AI infrastructure in 2026 alone.

Strategic Pillars & Key Topics:

⚔️ The Model Wars: Dueling Releases

• OpenAI GPT-5.3-Codex: A self-improving model that helped debug its own training runs. It scores nearly double its predecessor on desktop control benchmarks and is rated “High” risk for cybersecurity.

• Anthropic Opus 4.6: Features “Agent Teams” that split projects among multiple AI agents, a 1-million-token context window, and direct integration into PowerPoint and Excel.

• The “Self-Improvement” Era: Both labs now confirm their AI models are helping design their successors.

💰 The Trillion-Dollar Infrastructure Race

• Alphabet’s $185B Bet: Google’s parent company announces a CapEx plan of up to $185 billion for 2026 to support data centers and servers.

• Amazon’s $200B Plan: CEO Andy Jassy commits $200 billion to AI, chips, and robotics in 2026, causing the stock to dip on missed profit expectations.

🤖 The Enterprise Agent Layer

• OpenAI “Frontier”: A new platform that lets enterprises manage “AI Coworkers” like new hires—complete with onboarding, performance reviews, and permissions.

• Amazon Alexa+: Reports suggest Amazon is in talks to integrate ChatGPT into Alexa+, combining OpenAI’s brains with Amazon’s voice reach.

⚖️ Culture & Conflict

• Super Bowl Drama: Sam Altman calls Anthropic “dishonest” after their Super Bowl ad mocks ChatGPT for inserting ads into conversations.

• Meta’s Comeback: Internal memos reveal Meta’s new “Avocado” model is outperforming rivals, signaling a recovery after a rocky 2025.

Credits: This podcast is created and produced by Etienne Noumen, Senior Software Engineer and passionate Soccer dad from Canada.

Keywords: GPT-5.3-Codex, Claude Opus 4.6, Alphabet CapEx, Amazon AI Spending, OpenAI Frontier, Agent Teams, Meta Avocado, Sam Altman Super Bowl Ad, Self-Improving AI, Alexa+ ChatGPT

🚀 Reach the Architects of the AI Revolution

Want to reach 60,000+ Enterprise Architects and C-Suite leaders? Download our 2026 Media Kit and see how we simulate your product for the technical buyer: https://djamgamind.com/ai

Connect with the host Etienne Noumen, Senior Software Engineer and passionate Soccer dad from Canada

LinkedIn: https://www.linkedin.com/in/enoumen/

Youtube: https://youtube.com/@enoumen

X: https://twitter.com/enoumen

OpenAI’s GPT-5.3-Codex helps build itself

Image source: OpenAI

OpenAI just rolled out GPT-5.3-Codex, a new flagship coding model that merges its best programming and reasoning capabilities into one faster package — while also serving as a key tool in its own training and deployment process.

The details:

• OpenAI said early versions of 5.3-Codex were used to find bugs in its own training runs, manage its rollout, and analyze evaluation results.

• Codex tops agentic coding benchmarks like SWE-Bench Pro and Terminal-Bench 2.0, topping Opus 4.6 by 12% on the latter just minutes after its release.

• On OSWorld, a benchmark testing AI control of desktop computers, the model scored 64.7% — nearly double the 38.2% from the prior Codex version.

• OpenAI flagged the model as its first “High” cybersecurity risk rating and committed $10M in API credits to fund defensive security research

Why it matters: The self-improvement angle here is the headline, with Anthropic’s Dario Amodei also recently saying Claude is helping design its own successor. Yesterday’s bickering over ads now looks childish compared to the true fight on the model frontier, with a big day of dueling releases out of both labs.

Anthropic’s Opus 4.6 with ‘agent teams’, 1M context

Image source: Anthropic

Anthropic released Claude Opus 4.6, the company’s new most powerful model — featuring multi-agent collaboration in Claude Code, a massive context window, and new Office integrations that put the AI directly inside PowerPoint.

The details:

• A new “agent teams” feature in Claude Code lets multiple AI agents split a single project and work simultaneously instead of handling steps one at a time

• Opus 4.6 brings a 1M token context window to Anthropic’s Opus tier for the first time, matching what Sonnet offers for heavy document and code work.

• New Excel and PowerPoint sidebars let Claude read users’ existing templates and build models or decks natively without copying and pasting between tools.

• 4.6 topped most agentic benchmarks, including a leap on ARC-AGI-2 to nearly 70% — though OAI’s Codex 5.3 reclaimed agentic coding highs minutes later.

Why it matters: It’s a big day for devs, with both Codex 5.3 and Opus 4.6 releases bringing major capability increases across the board. With time between upgrades getting shorter and the length of tasks models can take on continuing to move up the curve, the “AI is hitting a wall” crowd seems pretty quiet these days.

OpenAI’s Frontier to manage ‘AI coworkers’

Image source: OpenAI

OpenAI just launched Frontier, a new platform for enterprises to deploy and manage AI agents like new hires — complete with onboarding, permissions, and performance reviews across a company’s existing tech stack.

The details:

• Frontier connects to existing enterprise systems like CRMs and ticketing tools, letting agents pull context from across the business without migrations.

• Built-in eval and feedback loops let agents learn via experience, with OAI comparing it to onboarding a new employee with reviews and boundaries.

• Every agent operates under its own profile with scoped access and hard limits on what it can touch for enterprise and regulated control.

• HP, Oracle, State Farm, and Uber are among the first adopters, with OAI embedding engineers on-site to help teams get agents into production.

Why it matters: Anthropic and OAI have been battling over models and coding tools, but Frontier shows the fight is also bleeding into who controls the enterprise agent layer underneath. Model capabilities are making AI coworkers a reality in the near future, and the system that ultimately orchestrates them will be valuable real estate.

Alphabet plans record AI spending after strong earnings

• Alphabet announced it will spend between $175 billion and $185 billion on AI infrastructure this year, far exceeding the $119.5 billion analysts predicted and more than its combined spending over the past three years.

• CEO Sundar Pichai said the AI investments are necessary, stating the company is seeing AI investments and infrastructure drive revenue and growth across search, cloud, and other products.

• About 40% of technical infrastructure investments go toward data centers and networking equipment, while 60% go toward servers, as Alphabet races to compete against Amazon, Microsoft, and OpenAI.

Sam Altman clashes with Anthropic over Super Bowl ads

• OpenAI CEO Sam Altman and Anthropic got into a public fight after Anthropic released Super Bowl ads that mocked ChatGPT’s upcoming ad-supported free tier by showing chatbots inserting awkward product promotions into personal conversations.

• Altman admitted the ads were funny but called Anthropic “dishonest” and “authoritarian,” saying OpenAI would never insert ads into conversations, though the company has said it plans to show ads based on chat content.

• Altman claimed Anthropic only serves rich people, but both companies offer similar free and paid tiers, and both enforce content policies, making his accusations seem like an overreaction to a cheeky commercial.

OpenAI unveils Frontier, a platform to build ‘AI-coworkers’

• OpenAI launched Frontier, an enterprise platform that lets companies build, deploy, and manage AI agents that can run other software like Salesforce and Workday.

• The platform connects databases, business tools, and internal applications, letting AI agents and human employees work together with the same data, tools, and security controls.

• Fortune 500 companies including Intuit, State Farm, Thermo Fisher, and Uber have signed up as initial customers, while investors worry Frontier could threaten traditional SaaS providers.

Meta’s internal memo signals AI comeback after rocky year

• An internal memo reveals that Meta has completed pretraining on its new AI model codenamed “Avocado,” marking a potential recovery after the company’s troubled 2025 with its Llama 4 release.

• Avocado beats the best freely available base models and competes with leading fully trained models in knowledge, visual perception, and multilingual performance, despite not yet going through post-training.

• The new model is ten times more efficient than Maverick and a hundred times more efficient than Behemoth, with Meta planning capital spending of $115 to $135 billion for 2026.

Alexa+ could integrate ChatGPT

• Amazon is reportedly in talks to integrate ChatGPT into Alexa+, which could mean the voice assistant would be powered by both Anthropic’s Claude and OpenAI’s technology after a potential $50 billion investment.

• The timing is notable because Amazon just opened Alexa+ to all US users after nearly a year of limited testing, and Apple is about to launch an upgraded Siri powered by Google’s Gemini.

• The deal could include other elements beyond Alexa+, such as access to OpenAI’s Codex, getting OpenAI to use Amazon’s Trainium chips, or opening Amazon’s store to OpenAI’s agents.

Anthropic launches Claude Opus 4.6

• Anthropic released Claude Opus 4.6, its newest model, featuring a 1M token context window in beta — a first for the company’s Opus-class models — along with stronger coding and agentic task performance.

• Opus 4.6 scores highest on benchmarks like Terminal-Bench 2.0, Humanity’s Last Exam, and BrowseComp, and beats OpenAI’s GPT-5.2 by around 144 Elo points on the GDPval-AA knowledge work evaluation.

• New developer tools include adaptive thinking, four effort levels, context compaction for longer-running agents, and agent teams in Claude Code that let multiple agents work in parallel on tasks.

Amazon plans $200B in AI spending

• Amazon plans to spend $200 billion in capital expenditures in 2026, up from the $125 billion it had estimated for 2025 and well ahead of analyst expectations.

• CEO Andy Jassy tied the $200 billion AI spending plan to strong demand for existing offerings and opportunities in AI, custom chips, robotics, and low earth orbit satellites.

• Amazon’s stock sunk more than 10% in after-hours trading after the $200 billion spending plan was revealed alongside missed Wall Street profit expectations in its Q4 earnings.

Apple scales back AI health coach plans

• Apple has scaled back Project Mulberry, its planned AI-powered health coach for the Health app, after leadership changes in the company’s health and AI organizations prompted a reassessment of the initiative.

• Services chief Eddy Cue, who now oversees the health team, told colleagues that Apple’s plan wasn’t compelling enough compared to rivals like Oura Health and Whoop and their iPhone apps.

• Some Project Mulberry features may still ship individually over time, including one that would use the iPhone’s camera to analyze how a person walks, plus Siri health queries in iOS 27.

OpenAI launches GPT-5.3-Codex its self-building coding model

• OpenAI has released GPT-5.3-Codex, a new coding model designed for developers that the company says set record scores on SWE-bench Pro and Terminal-Bench while running 25% faster than its predecessor.

• The model showed a big jump on Terminal-Bench 2.0, scoring 77.3% compared to 64.0% for GPT-5.2-Codex, and on OSWorld-Verified it reached 64.7% versus the prior model’s 38.2%.

• Developers can now steer and interact with GPT-5.3-Codex mid-task without losing context, and the model provides frequent progress updates so users can ask questions and guide it in real time.

Meta tests a standalone app for its AI-generated ‘Vibes’ videos

• Meta confirmed it is testing a standalone app for Vibes, its AI-generated short-form video feature that lets people create, share, and browse AI videos in a dedicated feed.

• Vibes previously lived inside the Meta AI app, but spinning it out positions the product as a more direct competitor to Sora, OpenAI’s AI video and social app.

• Meta also plans to explore freemium access and paid subscriptions for Vibes video creation, letting users unlock additional monthly video creation opportunities in the coming months.

What Else Happened in AI on February 06 2026?

Perplexity launched Model Council, a new feature that runs queries through multiple AI models at the same time and synthesizes outputs into a single answer.

Roblox introduced 4D generation via its Cube AI foundation model, letting creators generate fully functional, interactive objects from text prompts.

Lotus Health raised $35M in Series A funding for its free AI-powered primary care platform, providing diagnosis, prescriptions, and referrals across 50 states.

Meta is rolling out a standalone app for its Vibes AI video platform, which was previously only available via the Meta app.

AI evaluation firm METR released new analysis for GPT-5.2 (high), finding it can now handle tasks that would take a human engineer over 6 hours to complete.

OpenAI introduces Frontier, a new platform that helps enterprises build, deploy, and manage AI agents that can do real work.

‘In the end, you feel blank’: India’s female workers watching hours of abusive content to train AI.

Rentahuman.ai Turns Humans Into On-Demand Labor For AI Agents.

OpenAI says new coding model helped build itself.