Biohub, the nonprofit spearheaded by Mark Zuckerberg and Priscilla Chan, is committing $500 million to help create better AI simulations of the human body, officials shared first with Axios.

Why it matters: The bet is that more data and compute will produce more useful models.

The big picture: Zuckerberg said last year that Biohub's long-term goal is to cure all human disease through the intersection of AI and biology.

• The effort goes beyond frontier AI companies' focus on drug discovery.

What they're saying: "I think there is a path to building accurate predictive models of the cell," Biohub chief Alex Rives told Axios.

• The key is having more data, Rives said. The usefulness of an AI model's prediction increases exponentially as the scale of the data grows.
• Most current datasets cover about a billion cells, Rives said. "We hope to get an order of magnitude or more beyond that."

Zoom in: Of the $500 million over five years, Biohub will spend $400 million on its own work and $100 million to spur others.

• Partners include chipmaker Nvidia, leading research organizations such as the Allen Institute as well as the Human Cell Atlas and the Human Protein Atlas.

Yes, but: Reaching the ambitious goal of curing all disease is likely to take more than five years or the $500 million that Biohub is providing. Rives said that he hopes other funders will expand on the $100 million that Biohub is making available to other entities.

• "The advances that we've had in protein biology came through decades of investment," Rives said.

What we're watching: The big unknown is how much data it will take to make cellular models accurate enough to produce useful predictions.

• "We don't yet know what the slope of the scaling law is with cellular biology," Rives said.

After reports of OpenAI missing its own revenue targets, investors soured on tech stocks yesterday, putting the entire market under pressure as a result. But investors in the private AI labs are unfazed.

Why it matters: The AI bulls aren't going anywhere.

What they're saying: "We're not slowing down just because of a negative article in the Wall Street Journal," Ben Reitzes, managing director at Melius Research, said on Bloomberg Television.

• "Investors who are selling on the thesis that this whole AI thing is about to implode based on these new reports are going to be disappointed," Mark Malek, chief investment officer at Siebert Financial, tells Axios.

Between the lines: An investor in both OpenAI and Anthropic tells Axios that we're still in the early innings of the AI race, and there won't be one winner.

• That same investor sees AI's math problem (the high cost of compute pressuring revenue) resolving when AI labs raise prices.
• While that could lead to a near-term dip in demand, the net revenue result will be positive for AI companies.

Yes, but: Venture investors seem increasingly skeptical about OpenAI, amid multiple reports of outsize demand for Anthropic shares.

• OpenAI "should be scrambling," Malek notes, because competitors like Anthropic and Google's Gemini are taking market share.

Flashback: OpenAI had its "Code Red" moment earlier this year, when the company decided to focus on growing revenue in its enterprise business.

• It has taken steps to do that, partnering with consultants to grow Codex subscriptions, for example.
• Since then, Codex has hit 4 million users, and the company tells Axios it's "firing on all cylinders."

What we're watching: Earnings for some of the biggest public tech companies start today, and that should give investors more clarity on demand, capital expenditure and revenue guidance.

Elon Musk took the stand as the first witness in his billion-dollar lawsuit against OpenAI yesterday in federal court in Oakland, California.

Why it matters: A verdict against OpenAI could reshape who controls one of the most valuable private companies in the world.

The big picture: Musk is asking a court to put legal limits on OpenAI's transformation from charity-backed research lab to AI superpower.

• Musk said it's very simple. "It's not OK to steal a charity."

Zoom in: Musk is seeking sweeping remedies, including changes to OpenAI's corporate structure, removal of OpenAI CEO Sam Altman and president Greg Brockman, and damages that he has said should go back to OpenAI's nonprofit arm.

• His side has put the potential damages in the hundreds of billions.

The other side: OpenAI says Musk is motivated by jealousy and regret for walking away from OpenAI.

What we're watching: Musk is scheduled for several more hours of testimony and cross-examination today.

Keep reading.

Amazon and OpenAI yesterday announced an expanded deal that will make the AI startup's models available from Amazon's cloud.

Why it matters: The move comes a day after OpenAI and Microsoft announced a revised partnership that had required developers to use Microsoft Azure to access most OpenAI services.

Driving the news: OpenAI and Amazon said customers will soon be able to access OpenAI models and its Codex tool via AWS.

Between the lines: OpenAI is betting that it can win more business customers if its models are available through whatever cloud provider they are already using, which for many companies is AWS.

Keep reading.

• Exclusive: OpenAI and Anthropic briefed House Homeland Security Committee staff on their new cyber-capable AI models and what they could mean for cybersecurity. (Axios)
• What really matters in the Musk vs. Altman trial is how much damaging information each side can reveal about the other. (Axios)
• Google is said to have a deal with the Pentagon to allow its AI models to be used on classified servers for "any lawful government purpose." (The Information)
• Microsoft says it's working to improve reliability for GitHub after two recent outages.

If you are into AI but still long to travel to the past, check out talkie, a vintage language model trained exclusively on pre-1931 data.

• If you'd rather laugh at AI's failures, savor the fact that South Africa had to withdraw its draft AI policy after it was found to contain what appear to be AI hallucinations.