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

Photo via Getty

The Enterprise Data War Hits OpenAI

By Stephanie Palazzolo · Jul 23, 2025 7:00am PDT

Slack is the glue that holds OpenAI together: Everyone at the 3,000-person company constantly uses the messaging app to discuss their work, former employees have said.So perhaps it’s no surprise that as OpenAI tried to turn ChatGPT into a core enterprise productivity app that connects to all the other apps people use for work, it wanted... Slack is the glue that holds OpenAI together: Everyone at the 3,000-person company constantly...

Art by Mike Sullivan

Where Reinforcement Learning is Going

By Stephanie Palazzolo · Jul 22, 2025 7:00am PDT · 1 comment

Ever since researchers began noticing a slowdown in improvements to large language models using traditional training methods, they’ve shifted to reinforcement learning, a technique that rewards a model for accomplishing certain goals and penalizes it for other behaviors, as we covered last week.That typically means human experts giving feedback... Ever since researchers began noticing a slowdown in improvements to large language models using...

Art via Mike Sullivan

Where LLMs Are Falling Short

By Stephanie Palazzolo · Jul 21, 2025 7:00am PDT · 5 comments

I’m back from the International Conference on Machine Learning in Vancouver, one of the biggest annual meetups for artificial intelligence researchers. And this year, the conference underscored all the ways in which large language models are still falling short of everyone’s expectations, despite the immense progress that got us to this point.... I’m back from the International Conference on Machine Learning in Vancouver, one of the biggest...

Art via Mike Sullivan

OpenAI’s Clever Coding Hack with ChatGPT ‘Startup-Killing’ Agents

By Stephanie Palazzolo · Jul 17, 2025 7:00am PDT · 1 comment

Art via Mike Sullivan

OpenAI’s Clever Coding Hack with ChatGPT ‘Startup-Killing’ Agents

By Stephanie Palazzolo · Jul 17, 2025 7:00am PDT · 1 comment

OpenAI is continuing its push into the application layer with ChatGPT “agents.” As we scooped this week, it plans new features that will allow users to create and edit spreadsheets and presentations, generate reports and automate tasks on their browsers—all through ChatGPT.There was one interesting detail in that news that you might have missed.... OpenAI is continuing its push into the application layer with ChatGPT “agents.” As we scooped...

Image via Getty

Why xAI Spent So Much on Reinforcement Learning

By Stephanie Palazzolo · Jul 16, 2025 7:00am PDT · 2 comments

Watching last week’s Grok 4 release, one statistic stood out to us. XAI said it spent 10 times as much computational power on reinforcement learning for Grok 4 as it laid out for Grok 3.RL is a common AI development technique that rewards a model for accomplishing certain goals and penalizes it for other behaviors. The xAI disclosure provided... Watching last week’s Grok 4 release, one statistic stood out to us. XAI said it spent 10 times as...

Mark Zuckerberg wearing Orion AR glasses. Photo by Getty

Zuckerberg Says His AI Will Solve ‘Simpler Things’ Than Its Rivals

By Stephanie Palazzolo · Jul 15, 2025 7:00am PDT · 1 comment

Before we get to today’s column about Cognition buying Windsurf, Meta Platforms CEO Mark Zuckerberg on Monday spoke to Jessica Lessin, our editor in chief, for the inaugural episode of TITV, a new daily tech show, about the question on everyone’s mind these days: Why exactly would an AI researcher come to Meta? (Other than the $100 million pay... Before we get to today’s column about Cognition buying Windsurf, Meta Platforms CEO Mark...

Photos via Getty

Grok 4 is the Real Deal

By Rocket Drew and Amir Efrati · Jul 14, 2025 7:00am PDT

Before we get to today’s main column about Grok 4, we wanted to share four things we learned from the craziness on Friday, when Google snatched the CEO and technology behind Windsurf, a competitor to Cursor that OpenAI had agreed to buy for $3 billion just two months ago. Before we get to today’s main column about Grok 4, we wanted to share four things we learned from...

Art by Mike Sullivan

What the OpenAI Web Browser Means; Musk Says Grok 4 is No. 1

By Amir Efrati · Jul 10, 2025 7:00am PDT

Art by Mike Sullivan

What the OpenAI Web Browser Means; Musk Says Grok 4 is No. 1

By Amir Efrati · Jul 10, 2025 7:00am PDT

OpenAI’s embrace of Google Cloud servers isn’t stopping it from launching a direct assault on several of Google’s core products. Its next target: Chrome.Reuters on Wednesday reported that OpenAI’s web browser is coming soon. It’s been eight months since we broke the first news about OpenAI’s development of the browser, including its hiring of... OpenAI’s embrace of Google Cloud servers isn’t stopping it from launching a direct assault on...

Art by Mike Sullivan

Why Cloudflare Can’t Block Google From Scraping Websites For Its AI Products

By Erin Woo · Jul 9, 2025 7:00am PDT · 2 comments

Cloudflare, which powers many of the world’s most prominent websites, made waves last week by introducing a default setting for new customers that would block bots that artificial intelligence firms such as OpenAI and Anthropic use to scrape sites to train artificial intelligence. What Cloudflare didn’t highlight was that Google’s AI... Cloudflare, which powers many of the world’s most prominent websites, made waves last week by...

Yichao "Peak" Ji, co-founder and Chief Scientist at Manus AI. Screenshot via YouTube

OpenAI, Investors Eye Agent Startups Led by Chinese Founders

By Juro Osawa and Jing Yang · Jul 8, 2025 7:00am PDT

Chinese tech founders with strong pedigrees are starting to attract attention in Silicon Valley, including from OpenAI, for launching artificial intelligence agents targeting the U.S. market. They’re following in the footsteps of Manus, a browser-using agent developed by Chinese founders that briefly went viral earlier this year and got... Chinese tech founders with strong pedigrees are starting to attract attention in Silicon Valley,...

An underground robot fight in San Francisco. Photo by Sam D'Amico

Two State AI Laws to Watch

By Rocket Drew · Jul 7, 2025 7:00am PDT

Artificial intelligence firms’ failure last week to convince U.S. lawmakers to ban state-based laws targeting AI means all eyes are now on state legislatures that aim to pass them.One law that’s close to the finish line is New York’s Responsible AI Safety and Education Act. The state legislature passed the bill in June, and eventually it will... Artificial intelligence firms’ failure last week to convince U.S. lawmakers to ban state-based...

OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei. Photos via Getty

It’s Time to Take Anthropic and OpenAI’s Wild Revenue Projections Seriously

By Stephanie Palazzolo and Amir Efrati · Jul 3, 2025 7:00am PDT · 2 comments

OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei. Photos via Getty

It’s Time to Take Anthropic and OpenAI’s Wild Revenue Projections Seriously

By Stephanie Palazzolo and Amir Efrati · Jul 3, 2025 7:00am PDT · 2 comments

Six months ago or so, when OpenAI and Anthropic projected how much revenue they would generate this year, the figures might have sounded crazy. Not anymore! This week, Natasha and I reported that Anthropic has passed $4 billion in annualized revenue, up from a $3 billion rate it was at just a month ago and about $1 billion at the start of... Six months ago or so, when OpenAI and Anthropic projected how much revenue they would generate...

Photo via Getty

Why Meta Should Give Up on Superintelligence

By Stephanie Palazzolo · Jul 2, 2025 7:00am PDT · 7 comments

Meta Platforms CEO Mark Zuckerberg is on a tear in recruiting for a team focused on “superintelligence,” roughly defined as a more powerful version of artificial general intelligence, AI that can handle most of the economic work people do. Here’s a question, though: Why is Meta going after superintelligence?We can understand why Meta, as a... Meta Platforms CEO Mark Zuckerberg is on a tear in recruiting for a team focused on “...

Photo via Getty

What Zuckerberg’s New AI Team is Good At

By Stephanie Palazzolo and Kalley Huang · Jul 1, 2025 7:15am PDT

In recent weeks, Meta Platforms has probably generated more buzz from hiring two executives and 14 artificial intelligence researchers—mostly from OpenAI—than from its prior AI efforts altogether. Meta CEO Mark Zuckerberg on Monday officially announced 11 of those researcher hires as part of a new organization, Meta Superintelligence Labs.The... In recent weeks, Meta Platforms has probably generated more buzz from hiring two executives and...

Photo via Getty

AI Is Getting Cheaper, Right?

By Stephanie Palazzolo · Jun 30, 2025 7:00am PDT · 3 comments

The overarching narrative of the past two years has been that AI models are getting cheaper for customers. Just take a look at this chart, showing how the cost of purchasing access to a GPT-4 level AI model from OpenAI has dropped more than 85% since it was released in early 2023. And that’s without getting into the rise of cheap,... The overarching narrative of the past two years has been that AI models are getting cheaper for...

Art by Mike Sullivan

Will Chains of Thought Stay Readable for Long?

By Rocket Drew · Jun 26, 2025 7:22am PDT

Art by Mike Sullivan

Will Chains of Thought Stay Readable for Long?

By Rocket Drew · Jun 26, 2025 7:22am PDT

Chains of thought, the “reasoning” steps that many AI models take before delivering their final answer to a query, offer AI developers a precious window into their models’ decision-making. That’s why Google faced a backlash from its users earlier this month when it followed the lead of OpenAI and Anthropic by deciding to hide its models’ chains... Chains of thought, the “reasoning” steps that many AI models take before delivering their final...

Chart by Shane Burke

Which Coding Assistants Retain Their Customers and Which Ones Don’t

By Stephanie Palazzolo and Shane Burke · Jun 25, 2025 7:00am PDT · 4 comments

The explosion in AI coding assistants, used by developers to make apps and websites, has raised questions about whether developers will be loyal to any of them or if they’ll switch around based on which one has access to the best coding model or capabilities at the moment.But new data from research firm Indagari, based on credit card transaction... The explosion in AI coding assistants, used by developers to make apps and websites, has raised...

Art by Mike Sullivan

A Wave of AI M&A Is Coming; The Most Hated AI Startup

By Stephanie Palazzolo and Kevin McLaughlin · Jun 24, 2025 7:00am PDT

Before we get to the column, I’d recommend you check out this story Erin, Rocket and I published on Monday about what sorts of products Thinking Machines Lab, the AI startup founded by ex-OpenAI Chief Technology Officer Mira Murati, is looking to build. It gives a sneak peek into how Murati plans to catch up with OpenAI and other AI rivals,... Before we get to the column, I’d recommend you check out this story Erin, Rocket and I published...

Art by Mike Sullivan

The Great Data-Labeling Debate: Synthetic or Human?

By Stephanie Palazzolo · Jun 23, 2025 7:00am PDT

The news around data-labeling startups is stirring a debate in the industry about whether artificial intelligence developers should improve their models with cheap synthetic data (data that are generated by other AI models) or with a much smaller amount of expensive, human-labeled data via firms such as Scale or Surge.Until recently, AI... The news around data-labeling startups is stirring a debate in the industry about whether...

Photo by Mike Sullivan via Midjourney

Zuckerberg Fills Out His AI Dream Team; Developers Are Cheating on LMArena

By Stephanie Palazzolo · Jun 19, 2025 7:00am PDT

Photo by Mike Sullivan via Midjourney

Zuckerberg Fills Out His AI Dream Team; Developers Are Cheating on LMArena

By Stephanie Palazzolo · Jun 19, 2025 7:00am PDT

Before we get to today’s column, my colleagues Kalley and Cory broke some important news on Wednesday: Meta Platforms is close to hiring former GitHub CEO Nat Friedman, who now moonlights as a prominent AI investor, and Daniel Gross, Friedman’s investment partner and a cofounder of Safe Superintelligence with Ilya Sutskever. In the process,... Before we get to today’s column, my colleagues Kalley and Cory broke some important news on...