AI Agenda by Stephanie Palazzolo — The Information

Four times/week (subscribers only)

AI Agenda

By Stephanie Palazzolo

Stay ahead of the technological breakthroughs, startups and leaders in AI with a 14-day free trial of the AI Agenda. After your trial, unlock 25% off on a subscription to The Information for full access to the newsletter, plus all of our original reporting on AI.

By providing your email, you agree to The Information’s Privacy Policy

Recent Editions

Perplexity CEO Aravind Srinivas. Photo via Getty

Is Perplexity Worth $15 Billion?

By Natasha Mascarenhas and Stephanie Palazzolo · Feb 26, 2025 7:14am PST

Before we get into today’s column, leaders at OpenAI have told some employees that GPT-4.5—the model known internally as Orion—will be released this week, according to a person with knowledge of the conversations.Don’t get too excited though. A person who’s tested the model told us that its performance on certain tasks have been mixed; for... Before we get into today’s column, leaders at OpenAI have told some employees that GPT-4.5—the...

Photo by Shutterstock

Why xAI and Anthropic’s New Products Matter

By Stephanie Palazzolo · Feb 25, 2025 7:00am PST · 1 comment

If you’ve been on X at all in the last few days, you’ve probably come across posts in which people described jailbreaking xAI’s new Grok 3 model. For instance, one person asked the chatbot how to make chemical weapons, while another revealed instructions xAI gave the model to stop it from saying that Elon Musk is the top spreader of... If you’ve been on X at all in the last few days, you’ve probably come across posts in which...

Aidan Gomez. Photo by Getty

Cohere Boosts Revenue Pace to $70 Million on Overseas Growth

By Jon Victor, Natasha Mascarenhas and Sri Muppidi · Feb 24, 2025 7:00am PST

Cohere, one of the earliest venture-backed developers of artificial intelligence models, has faced plenty of headwinds in the last two years. The company has been under threat from better-funded competitors like OpenAI and Anthropic and faces competition from cheap, open-source alternatives like DeepSeek.In spite of all that, the company is... Cohere, one of the earliest venture-backed developers of artificial intelligence models, has...

Clara Shih. Photo by Getty

Meta’s Open-Source AI Prepares to Go to Market

By Kalley Huang · Feb 20, 2025 7:00am PST

Clara Shih. Photo by Getty

Meta’s Open-Source AI Prepares to Go to Market

By Kalley Huang · Feb 20, 2025 7:00am PST

As Meta Platforms tries to make its large language model Llama the industry standard for artificial intelligence, it is not only courting big businesses but also taking steps to work with one of the largest customers in the world: the U.S. government. In November, the company poached Clara Shih from Salesforce, where she was CEO of ... As Meta Platforms tries to make its large language model Llama the industry standard for...

Photo by Mike Sullivan via Midjourney

How Can Investors Continue Funding AI Developers?

By Stephanie Palazzolo · Feb 19, 2025 7:00am PST

One of the earliest AI Agenda newsletters in August 2023 made the bear case for large language model developers. Over the last year-and-a-half, our predictions from that day seemed to largely come true, as model makers like OpenAI moved more into applications and cheap AI releases from DeepSeek and others threatened the margins of those selling... One of the earliest AI Agenda newsletters in August 2023 made the bear case for large language...

Photo via Getty

The Three Classes of AI Coding Assistants

By Stephanie Palazzolo · Feb 18, 2025 7:10am PST

The model race continues. On Monday evening, Elon Musk’s xAI released a new family of large language models, Grok 3, which includes a baseline model, a smaller and faster version of the baseline model, and two “reasoning” models.Early reactions seem promising. Andrej Karpathy, one of the cofounders of OpenAI who left the company last year, said... The model race continues. On Monday evening, Elon Musk’s xAI released a new family of large...

Photo by Getty

Anthropic Strikes Back

By Stephanie Palazzolo · Feb 13, 2025 7:00am PST

After OpenAI released its reasoning models last fall, Google, Alibaba, High-Flyer Capital Management and others followed with their own. One prominent competitor, Anthropic, has been noticeably absent from the race. Now we know one reason: Anthropic is taking a slightly different approach to reasoning. It has developed a hybrid AI model... After OpenAI released its reasoning models last fall, Google, Alibaba, High-Flyer Capital...

Thrive Capital founder Josh Kushner.

OpenAI Investor Kushner Tells Startup’s Staff: Leaving Is a Bad Idea

By Natasha Mascarenhas, Erin Woo and Amir Efrati · Feb 12, 2025 11:57am PST

Thrive Capital founder Josh Kushner.

OpenAI Investor Kushner Tells Startup’s Staff: Leaving Is a Bad Idea

By Natasha Mascarenhas, Erin Woo and Amir Efrati · Feb 12, 2025 11:57am PST

How fierce is the fight for talent at OpenAI and its rivals?One sign came Tuesday when Josh Kushner, founder of major OpenAI investor Thrive Capital, gave a presentation to some of OpenAI’s staff. As part of his talk about why Thrive had committed over $1 billion to the ChatGPT maker, Kushner drove home this message: Leaving would be a bad... How fierce is the fight for talent at OpenAI and its rivals?One sign came Tuesday when Josh...

Photo of Elon Musk via Getty

Musk’s OpenAI Bid Leaves Door Open For Boardroom Activism

By Rocket Drew and Stephanie Palazzolo · Feb 12, 2025 7:00am PST

As the OpenAI-Elon Musk drama bogged down in arguments about whether Musk’s group had actually sent the takeover offer to the board, another issue surfaced. While OpenAI is signalling it will dismiss the Musk-led group’s offer, it turns out any of OpenAI’s 10 board directors could take matters into their own hands. That’s because of a... As the OpenAI-Elon Musk drama bogged down in arguments about whether Musk’s group had actually ...

Photos by Getty

The Strategy Behind Musk’s $97 Billion Bid for OpenAI

By Stephanie Palazzolo and Rocket Drew · Feb 11, 2025 7:00am PST · 1 comment

You've got to hand it to Elon Musk: He is great at getting attention for his causes. In the latest example, the OpenAI cofounder-turned-rival made a $97.4 billion bid for the assets of OpenAI’s nonprofit—namely, control over the OpenAI for-profit arm that runs ChatGPT. This is all part of Musk’s monthslong effort to stop the for-profit from... You've got to hand it to Elon Musk: He is great at getting attention for his causes. In the...

Aaron Ginn, cofounder and CEO of Hydra Host

The ‘GPU Whisperer’ Making Software for Data Centers

By Anissa Gardizy · Feb 10, 2025 7:00am PST

Is going “bare metal” in data centers the future for cloud computing? If so, it could signal a shift away from big cloud firms toward a new group of AI computing startups.Bare metal isn’t a subset of rock music but an industry term for renting servers to cloud customers on terms that give them access to the hardware, allowing customers to... Is going “bare metal” in data centers the future for cloud computing? If so, it could signal a...

Photos by Getty

Google Closes In On DeepSeek’s AI Efficiency, While DeepSeek’s Chatbot Blows Past Google’s

By Amir Efrati and Rocket Drew · Feb 6, 2025 7:00am PST

Photos by Getty

Google Closes In On DeepSeek’s AI Efficiency, While DeepSeek’s Chatbot Blows Past Google’s

By Amir Efrati and Rocket Drew · Feb 6, 2025 7:00am PST

Google doesn’t get the accolades OpenAI, Anthropic and DeepSeek are getting in advancing artificial intelligence, but it’s arguably doing some things right.Case in point: On Wednesday, Google made a trio of new Gemini AI models generally available to app developers, including a reasoning model akin to OpenAI’s o1 and the much-discussed DeepSeek... Google doesn’t get the accolades OpenAI, Anthropic and DeepSeek are getting in advancing...

Alexandr Wang, CEO of Scale AI. Photo by Getty

Will DeepSeek Hurt Scale AI’s Business Model?

By Stephanie Palazzolo · Feb 5, 2025 7:00am PST · 1 comment

To understand how OpenAI has been improving the quality of its reasoning models in recent months, check out the following question the company recently asked scores of biology PhDs:“Please write a summary of wet-lab experimental approaches available before August 2024 for assessing in parallel the gene expression effects of many different... To understand how OpenAI has been improving the quality of its reasoning models in recent months,...

Photo by Getty

ChatGPT’s New ‘Research’ Feature: The Good and the Bad

By Stephanie Palazzolo · Feb 4, 2025 7:00am PST

It’s been barely a day since OpenAI released Deep Research, a new ChatGPT feature that helps customers conduct multi-step research. And already it’s gotten at least one rave review.Derya Unutmaz, a professor of immunology at the Jackson Laboratory for Genomic Medicine who’s been testing the feature since he was granted early access on Friday,... It’s been barely a day since OpenAI released Deep Research, a new ChatGPT feature that helps...

Photo via Getty

OpenAI’s DeepSeek Response: Catch Us If You Can

By Stephanie Palazzolo and Rocket Drew · Feb 3, 2025 7:17am PST · 1 comment

OpenAI’s release on Friday of o3-mini, its latest artificial intelligence reasoning model, wasn’t a surprise. CEO Sam Altman had previously said it would drop by then.But it was impossible to ignore the coincidence of the release coming a few days after a blaze of news erupted over DeepSeek R1, a Chinese reasoning model that has matched or even... OpenAI’s release on Friday of o3-mini, its latest artificial intelligence reasoning model, wasn’t...

Photo by AP

Why It’s Cheap and Easy to Mimic OpenAI’s Reasoning Model

By Rocket Drew · Jan 30, 2025 7:06am PST · 1 comment

Photo by AP

Why It’s Cheap and Easy to Mimic OpenAI’s Reasoning Model

By Rocket Drew · Jan 30, 2025 7:06am PST · 1 comment

The word of the moment in artificial intelligence is distillation. Distillation is the technical term for transferring knowledge from one large language model, which acts like a teacher, to another model that acts as a student. And it’s likely one of the key ways a Chinese quant fund replicated OpenAI’s o1 reasoning model and set the... The word of the moment in artificial intelligence is distillation. Distillation is the...

Photo via Getty

OpenAI Tightens Grip on High End of App Market; Murati’s Startup Gets a Name

By Stephanie Palazzolo and Amir Efrati · Jan 29, 2025 7:00am PST

As questions swirl about whether super-cheap-yet-high-quality Chinese artificial intelligence threatens American-made AI, a new data point suggests OpenAI has some major advantages, relative to its compatriots.We reported Tuesday that OpenAI leaders recently made comments implying that ChatGPT Pro subscriptions—which cost $200 per month, or 10... As questions swirl about whether super-cheap-yet-high-quality Chinese artificial intelligence...

Photo via Getty

Is DeepSeek an Extinction Event?; A New Reasoning-Focused AI Lab Lands Funding

By Stephanie Palazzolo · Jan 28, 2025 7:00am PST · 4 comments

If you spent any time yesterday or this weekend on X, you might have concluded that AI developers, venture capitalists, Nvidia and cloud providers are totally screwed, thanks to new uber-cheap and impressive AI models from High-Flyer Capital Management, a Chinese hedge fund.The truth is a lot more nuanced, though, and there are still a lot of... If you spent any time yesterday or this weekend on X, you might have concluded that AI...

Photo by Getty

DeepSeek’s Implications For AI; Why Stargate Should Be Taken Seriously

By Anissa Gardizy · Jan 27, 2025 7:00am PST · 2 comments

Before we get to our main item, I’d recommend you read yesterday’s scoop about Meta Platforms and other American developers reacting to the threat posed by DeepSeek, the hottest new artificial intelligence model (and consumer mobile app!) that has been sweeping the industry.On to the news…Last week I was at one of the largest data center... Before we get to our main item, I’d recommend you read yesterday’s scoop about Meta Platforms and...

Photo by Getty

An OpenAI Math Test Raises Eyebrows

By Rocket Drew · Jan 23, 2025 7:00am PST

Photo by Getty

An OpenAI Math Test Raises Eyebrows

By Rocket Drew · Jan 23, 2025 7:00am PST

When OpenAI previewed o3, its latest reasoning model, in December, one of the ways the company demonstrated its strength was by showing off its 25.2% score on FrontierMath, a test of AI models’ math abilities that Epoch AI developed. (You might remember Epoch, the nonprofit research organization, from their work on AI scaling laws.)That’s not... When OpenAI previewed o3, its latest reasoning model, in December, one of the ways the company...