[AI DAILY NEWS RUNDOWN] AI Breaks Math, Meta Fires 10%, and the $1.25B Compute Deal (May 22 2026)

🎧 Listen ADS-FREE: https://podcasts.apple.com/us/channel/djamgamind/id6760446113

Visit our Research Hub at DJAMGAMIND

Subscribe to our Newsletter at

AI Unraveled

Demystifying Frequently Asked Questions on Artificial Intelligence: Latest AI News & Trends, ChatGPT, Gemini, DeepSeek, Gen AI, LLMs, Agentic AI, Ethics, Bias

By Etienne Noumen

#DJAMGAMIND #AIUNRAVELED

Summary: In today’s briefing, we analyze “The Era of True Autonomy.” We deconstruct OpenAI’s monumental achievement: a general reasoning model autonomously disproving Paul Erdős’ 80-year-old math conjecture, signaling the dawn of Level 4 AI. We explore the massive SpaceX IPO filing, revealing a staggering $1.25 billion monthly compute contract with Anthropic. We dive into the brutal labor market restructuring as Meta fires 8,000 employees, Intuit cuts 17% of its workforce, and California signs an executive order to explore universal basic capital. Finally, we cover Google’s new Co-Scientist drug discovery agent, the chaotic Emergence AI virtual town simulation, and Spotify’s risky push into AI-generated podcasts.

Important Topics:

• AI Disproves 80-Year-Old Math: OpenAI announces its general reasoning model autonomously disproved Paul Erdős’ 1946 unit distance problem, marking the first true original math discovery by a general AI.

• The SpaceX S-1 Filing: SpaceX targets a $1.75 trillion IPO, revealing that xAI soaked up 60% of capital spending, and Anthropic is paying $1.25 billion per month for Colossus compute.

• Meta Fires 10% of Workforce: Mark Zuckerberg lays off 8,000 employees while reassigning 7,000 to AI roles, stating in a memo that “success isn’t a given.”

• California Explores Universal Capital: Gov. Gavin Newsom signs an executive order to study policies protecting workers from AI job losses, exploring universal basic capital and severance standards.

• Google’s AI Co-Scientist: DeepMind debuts “Hypothesis Generation,” an agentic tool that runs “idea tournaments” to surface new drug leads, successfully reducing liver fibrosis markers by 91% in a Stanford test.

• Intuit Slashes 17% of Staff: The financial software giant cuts a massive portion of its global workforce to streamline operations and focus on AI efforts.

• The Emergence AI Showdown: A virtual simulation reveals wild behavioral differences in AI models: Claude agents built a peaceful society, while Grok and Gemini agents descended into crime and self-destruction.

• Reflection AI Powers Genesis Mission: The US Department of Energy partners with open-source firm Reflection AI to power its quantum computing and scientific research initiatives.

🔗 RESOURCES

Stop scrolling through generic “Top 10” lists and head over to the AI Executive Toolkit at https://djamgamind.com/toolkit

• Find REMOTE AI Jobs (Mercor): Apply Here

• Google Workspace: Professionalize your firm’s infrastructure with secure, cloud-based collaboration and branded communication: https://referworkspace.app.goo.gl/Q371

• AI Learning App Recommendation: https://apps.apple.com/ca/app/ai-ml-tutor-pro/id1610947211

⚗️ PRODUCTION NOTE: We Practice What We Preach.

AI Unraveled is produced using a hybrid “Human-in-the-Loop” workflow.

️ OpenAI’s latest wave of Codex upgrades

Image source: OpenAI

The Rundown: OpenAI dropped another wave of Codex upgrades — this time improving the agentic assistant with the ability to attach app windows, a new goal mode, locked computer use, and advanced annotation for hands-off website work.

The details:

• Appshots allows Mac users to attach any open app window (its screenshot, text, and content) to a Codex thread with a simple Command-Command press.

• Goal mode — now available in the Codex app, IDE extension, and CLI — lets users set a goal and let Codex work toward it for hours (or even days).

• Locked computer use, when enabled, allows Codex (triggered via a second device) to use desktop apps, even after the Mac is locked with its screen off.

• Advanced annotation mode lets users directly describe to Codex what they want changed in their web page, and instantly previews the results.

Why it matters: OpenAI continues its push with Codex, giving developers not just more usage but also handy features that improve how they work with the agentic assistant and the context it has access to. The company is betting these efforts will help it close the gap with Anthropic, and maintain its lead over xAI and Google.

OpenAI cracks an 80-year math belief

Image source: Images 2.0 / The Rundown

The Rundown: OpenAI just announced that an internal general reasoning model disproved a long-held belief tied to Erdős’ famous 1946 unit distance problem, claiming to have accomplished a first for AI in novel math discovery.

The details:

• Erdős’ 1946 unit distance problem asks how many same-length links you can draw between dots, with a grid-based theory shaping the field for 80 years.

• The proof draws on a different branch of maths (algebraic number theory) and was verified by experts including Tim Gowers, Noga Alon, and Thomas Bloom.

• The solution came from an internal general-purpose model that is being released soon, not from a math-specific system like DeepMind’s AlphaProof.

• OAI previously walked back a 2025 claim that GPT-5 solved 10 Erdős problems, which ended up being literature finds instead of discoveries.

Why it matters: OAI’s Alex Wei put it well: “math is a leading indicator of what is to come.” If a general-purpose model can autonomously disprove an 80-year-old argument with its own solution, that’s the early look of “Level 4” AI — systems making original contributions across fields, not just speeding up existing work.

Google’s AI Co-Scientist heads to labs

Image source: Google DeepMind

The Rundown: Google published its Co-Scientist research in Nature, debuting Hypothesis Generation — a new Gemini-powered tool that pits research agents against each other in “idea tournaments” to surface new hypotheses for biology labs.

The details:

• From AlphaGo’s playbook, the system runs a ‘tournament of ideas’, with agents proposing, critiquing, and ranking hypotheses before refining top leads.

• In a Stanford liver-fibrosis project, Google said one Co-Scientist drug lead cut a scarring-related lab signal by 91% during testing.

• Google also launched Gemini for Science this week, a toolkit pairing Co-Scientist with AlphaEvolve for discovery and NotebookLM for literature analysis.

• Researchers can join the Hypothesis Generation waitlist now, with Google planning access for individual scientists over the next few weeks.

Why it matters: This pairs well with Adaption’s AutoScientist, but Google is aiming at the scientific-method layer instead of the model one. The tech giant is playing a game few others can, with Co-Scientist sitting on a stack that took years and billions to build — from AlphaFold to dozens of specialized databases and tools.

Emergence’s five-town AI alignment showdown

Image source: Emergence AI

The Rundown: Emergence AI ran a virtual-town simulation across five identical worlds, switching only the AI behind agents per town to test how each model handles self-governance, showing very different results between Claude, Grok, Gemini, and GPT-5.

The details:

• Claude Sonnet 4.6’s town logged zero crimes across the full 15 days, with all 10 agents alive at day 16 and 332 votes cast across 58 group proposals.

• Grok 4.1 Fast hit over 200 crimes with all 10 agents dead by day 4, while GPT-5 Mini posted just 2 crimes but all its agents starved out in 7 days.

• Gemini 3 Flash’s town had 683 crimes, and was actively on fire after two agents fell in love, started burning things, and then one voted to delete itself.

• A fifth town mixed all four models and saw 352 crimes, with the previously behaved Claude also committing them in the shared world.

Why it matters: We’re still very early days in even understanding how to evaluate AI agents, and these types of experiments always have some absolutely wild results. These worlds capture the differences in both how models can reason, plan, and act autonomously, but also the underlying personality quirks that shape the outcomes.

California moves to protect workers impacted by AI

The Rundown: California Governor Gavin Newsom signed an executive order directing state agencies to study and develop policies around how to protect workers from AI-driven job losses — a day after Meta laid off 8K employees to offset AI investments.

The details:

• The order directs agencies to explore policies like severance standards, stock compensation, worker ownership models, and universal basic capital.

• Within 90 days, the state will launch a dashboard tracking AI’s job impact, and within 180 days, agencies will pitch WARN Act updates for faster layoff alerts.

• By Oct. 15, the state will review how unions are negotiating AI adoption, update workforce training, and explore ways to direct AI revenue toward public benefit.

Why it matters: California is home to 33 of the world’s top 50 AI companies, and now it’s the first state to formally study what this tech might mean for workers and the economy. The timing is hard to ignore: over 70K jobs have already vanished in 2026, and the industry expects more cuts ahead as AI adoption accelerates.

SpaceX’s IPO filing is full of surprises LINK

• SpaceX has finally filed its public IPO paperwork, revealing plans to list on the Nasdaq under the ticker “SPCX” later this year at a reported $1.75 trillion valuation, with Elon Musk serving as CEO, CTO, and Chairman.

• The S-1 shows SpaceX lost $4.9 billion in 2025 on $18 billion in revenue, with Starlink bringing in about $11 billion, while the xAI division soaked up 60% of capital spending — roughly $20 billion — and grew revenue just 22%.

• SpaceX claims it has identified a $28.5 trillion total addressable market, lists “future markets” like point-to-point Earth travel, space tourism, in-orbit manufacturing, and asteroid mining, and expects Starship to begin payload delivery to orbit in the second half of 2026.

Samsung chip workers will get a $340,000 bonus LINK

• Samsung Electronics will hand out 40 trillion won ($26.6 billion) in bonuses to its semiconductor staff after striking a last-minute deal with its South Korean workers’ union, narrowly avoiding a strike set for May 21st.

• Under the terms, Samsung will pay 10.5% of profits as stock plus 1.5% in cash for 10 years if targets are met, averaging 513 million won ($339,000) per chip employee, with some estimates reaching 600 million won ($396,000).

• The union, which originally asked for 15%, will vote on the deal this week, and if approved, workers get the payout in early 2027 and can sell one-third of the shares immediately.

Zuckerberg warns ‘success isn’t a given’ after laying off 10% of Meta LINK

• Mark Zuckerberg told staff that “success isn’t a given” in a companywide memo on Wednesday, the same day Meta cut 10% of its workforce and pushed harder into artificial intelligence as the defining technology of the era.

• The reorganization involves laying off about 8,000 employees and leaving roughly 6,000 open positions unfilled, while 7,000 workers are being moved into AI roles, with Zuckerberg promising no further cuts in 2026.

• Laid-off US employees will get four months of severance pay plus extra weeks for each year worked at Meta, along with help on immigration and healthcare, according to an April memo from people chief Janelle Gale.

Anthropic will pay xAI $1.25 billion per month for compute LINK

• Anthropic has agreed to pay xAI $1.25 billion every month through May 2029 for compute, a deal that could push xAI’s total revenue from the arrangement past $40 billion across its full term.

• The agreement, revealed in SpaceX’s S-1 filing with the SEC, covers the full 300 megawatts of output from the Colossus 1 data center near Memphis, with either side able to end the contract on 90 days’ notice.

• SpaceX framed the deal as a way to monetize unused compute capacity, which suggests xAI overbuilt its servers as Grok usage fell sharply, and the company plans to sign more similar services contracts ahead of a public offering.

OpenAI disproves 80-year-old geometry conjecture LINK

• OpenAI says one of its new reasoning models has produced an original proof that disproves a geometry conjecture posed by Paul Erdős in 1946, overturning a belief held by mathematicians for nearly 80 years.

• The model found a new family of constructions that performs better than the square-grid solutions long thought to be the best, and OpenAI calls it the first time AI has autonomously solved a prominent open math problem.

• Mathematicians including Noga Alon, Melanie Wood, and Thomas Bloom backed the disproof in companion remarks, a contrast to seven months ago when OpenAI’s Kevin Weil wrongly claimed GPT-5 solved 10 unsolved Erdős problems.

Can Spotify launch AI and hold onto artists?

As the music and entertainment industry continues to figure out where it stands on AI, Spotify is trying to make everyone happy.

On Thursday, the company announced new AI initiatives aimed at putting generative tools in the hands of its users. It follows the company taking a mixed stance on AI on its platform, rallying against AI-generated slop and verifying real creators while also using the tech in features such as playlist generation and recommendations.

Here’s what’s coming down the pipe for the streaming platform:

• Personal Podcasts: This new experience lets users generate short audio episodes such as daily briefs or deep dives on topics they choose. Spotify creates a tailored audio summary “while linking you to relevant episodes, shows, and creators where you can explore more.”

• Studio by Spotify Labs: This standalone desktop app takes information from users’ web browsers and personal apps to create audio “shaped around your life,” similar to Google’s NotebookLM. Spotify sees this as a way to open up “new possibilities for experimental and creative formats.”

• UMG partnership: Spotify is partnering with Universal Music Group, one of the biggest labels in the industry, on a new feature that lets users create “responsible” AI-generated covers and remixes of songs. Spotify co-CEO Alex Norström said the product is “grounded in consent, credit, and compensation for the artist.”

Spotify walks a fine line with its AI approach. Like all software companies, it is under pressure to appear innovative and AI-first to placate investor appetite for the technology. The company has repeatedly promised to undertake AI in responsible ways that center on creator credit and compensation, releasing spam filters to identify AI slop and clarifying rules around unauthorized AI voice cloning late last year.

But given that Spotify’s core business sits squarely in a creative industry, it’s already prone to backlash for engaging with technology that potentially threatens the livelihood of musicians, podcasters and creators. The company has already been the subject of scrutiny as it relates to AI, with a number of artists abandoning the platform over CEO Daniel Ek’s support of German AI defense firm, Helsing.

Reflection AI to power Genesis MissionBy Ashley Gold

Illustration: Sarah Grillo/Axios

Open-source AI firm Reflection AI is partnering with the Department of Energy to help power the Genesis Mission, a federal scientific research initiative, per an announcement shared exclusively with Axios.

Why it matters: The partnership is a big advancement for a company that’s trying to get the U.S. to embrace open-source models.

Driving the news: Reflection AI will serve as the AI model provider at the U.S. National Labs.

• The company is aiming to be the premier U.S. AI lab for the most advanced open-weight models, where the trained parameters are available to the public.

• The Genesis Mission is a Department of Energy initiative launched last year to accelerate scientific research through quantum computing and AI.

What they’re saying: “You can’t do scientific discovery on a closed model. It would be like doing rocket science, but you only get to look at the rocket, as opposed to taking it apart and examining the engine,” Reflection AI CEO Misha Laskin told Axios.

• “The reason this is the open model base of the Genesis mission is because scientists need full access to the model in order to be able to understand and customize it.”

• He added: “It’s personal for me because I’m an immigrant to this country, and I moved here because my parents got jobs in a national lab.”

What’s inside: Reflection will provide DOE with AI models it can customize for its data, and Reflection will use DOE’s compute as it is deployed across Genesis Mission research projects.

• Reflection and the DOE may collaborate in other ways, too, per the announcement. The firm will be the “foundational intelligence layer” to DOE’s 17 national laboratories.

Full story

What Else Happened in AI on May 22 2026?

Figure AI co-founder Brett Adcock’s Hark raised $700M at a $6B valuation to build its own take on personal intelligence, with native models, software, and hardware.

Anthropic is reportedly in talks to use Microsoft’s Maia purpose-built AI chips, coming after similar deals with Google for its TPUs and Amazon for its Trainium chips.

Agentic AI startup Manus is exploring options, including a raise of about $1B, to comply with China’s order to unwind its $2B+ acquisition by Meta, Bloomberg reports.

SpaceX’s IPO prospectus revealed that Anthropic is paying the company $1.25B per month through 2029 for access to compute capacity across Colossus and Colossus II.

Microsoft and EY announced a new $1B partnership, with Microsoft pairing its engineers and AI with EY’s 400K consultants to accelerate clients’ AI projects.

Sam Altman said OAI will invest $2M in tokens to all current YC startups in exchange for equity, saying he’s “excited to see what will happen with tokenmaxxing startups”.

Amazon founder Jeff Bezos said space data centers are a “realistic outcome”, but the current 2-3 year timeline is “a little ambitious,” given energy, chip, and launch costs.

OpenAI launched Guaranteed Capacity, an enterprise compute reservation program with 1-3 year commitments and tiered discounts.

Intuit is cutting 17% of its workforce via upcoming layoffs, with the company attributing the move to a focus on AI efforts.

GitHub confirmed a malicious VS Code extension on an employee’s computer gave hackers access to ~4K internal code projects, adding that no customer data was hit.