In the last month, OpenAI, Anthropic and Google have released more than a dozen new models. If you thought that pace was crazy, don’t expect things to slow down anytime soon.

Anthropic is preparing to release a new model that is better in its reasoning and intuition than the Anthropic models that are currently available commercially, according to a person who has used it. The unreleased model is especially good at discerning a person’s underlying intentions and goals even if they don’t give the model much information, such as explicitly stating those goals, the person said. 

For instance, a person might ask the model to fix a single bug, and the model might go off and instead identify the underlying class of problems that caused the bug and ask whether it should fix the whole class of problems, said the person who has used it.

The Anthropic model is also better at thinking outside of the box rather than regurgitating common sense advice or solutions, the person said. For instance, if a user asks for advice on how to invest in stocks, rather than providing generic advice like looking at price-to-earnings ratios, the model might come up with its own creative algorithms to predict future stock performance, the person said.

Remember when OpenAI landed in hot water earlier this month after mathematicians accused the company of trying to steal their thunder for solving an ultra-difficult math problem? (I know it feels like a million years ago given all the news cycles since then.)

Well, it turns out OpenAI is close to solving another math problem from the Millennium Prize Problems, a set of seven famous math challenges including that earlier “Navier-Stokes existence and smoothness problem,” according to a person with knowledge of the solution.

Employees expect the next problem, known as the Hodge Conjecture, to be solved relatively soon, the person said. The Hodge Conjecture asks whether certain geometric features of shapes defined by polynomial equations can always be described using simpler algebraic pieces. It could take the company longer to announce the solution, though, because it’s trying to figure out how to collaborate with the math community to make the announcement without triggering another PR nightmare, the person said. 

The year-old startup behind Instinct, a personal AI assistant that’s caught fire with Silicon Valley insiders, is seeking more computing power, leveraging its early buzz as it faces new competition from giant Meta Platforms.

Over the last few weeks, Instinct, which launched to a select group of users earlier this year, has sometimes notified users that it is running at full capacity and “responses may be slower” as people use it for tasks such as negotiating bills and answering emails. More cash could help: the company is looking to raise $1 billion in new funding after recently raising $250 million, its founder and CEO Noah Shinn has recently told prospective investors, according to a person familiar with the statement.

Hollywood mogul Jeffrey Katzenberg is teaming up with the former head of OpenAI’s Sora app to launch a new AI startup that would train its own video models and try to win over filmmakers, people familiar with the matter said.

The founders have been talking to potential investors including Andreessen Horowitz about raising significant funding for the new company, the people said.

Anthropic safety researcher Evan Hubinger said in a post on X Tuesday night that his company does “earnestly believe AI could kill all humans,” adding that he believes there’s a greater than a 10% chance that happens within the next decade.

The post immediately sparked outrage and concerns from ...

Gimlet Labs said late last week it had raised $300 million at a valuation of about $3 billion, in a round led by new investor Andreessen Horowitz. Today we have unreported details on how that round, which boosted its valuation by a whopping 16 times from its last PitchBook valuation, came together.

The three-year-old startup, which helps AI developers run models using different AI chips simultaneously, raised the round after the company’s executives told some potential investors that based on signed contracts, OpenAI was likely to spend more than $100 million a year on its services, according to a person familiar with the company. Such outlays, if they transpire, would be significant given it just emerged from stealth in October. 

This windfall would be in the future, however. “We have explored possibilities with Gimlet but we are not a paying customer,” said an OpenAI spokesperson.

Applied Compute, a year-old startup that helps companies run and customize open-source models with their own data, is discussing a new funding round that would value it at around $3 billion, more than double its valuation from a round announced four months ago, according to a person with knowledge of the funding.

The financing follows strong revenue growth in recent months. Applied Compute is currently generating around $50 million in annualized revenue, the person said, up nearly four times from what its CEO Yash Patil has said it was generating last November.

Legal AI startup Harvey is in talks to raise at least $500 million at a $15.5 billion valuation including the investment, a 40% premium to its last valuation five months ago. The fundraising follows a revenue surge at the legal AI startup, according to a person with knowledge of the fundraise.

Four-year-old Harvey is currently generating more than $350 million in annualized revenue, up more than 80% from the $190 million in annualized revenue it was generating in January, according to the person. Lightspeed Venture Partners has been interested in leading the round, the person said.

OpenAI is preparing to release a new model family, tentatively using the name “Astra,” with improved abilities to complete long-running tasks, according to three people briefed on the plans.

OpenAI CEO Sam Altman demonstrated Astra to policymakers and regulators in Washington, D.C., this week, ...