One such slight came in 2018. Brockman asked Dario to double-check a fact on one of his slides for an important meeting. Dario asked who the slides were for. When Brockman said that he and Altman were going to meet former President Barack Obama, Dario got angry that he had been left out of the loop. The following year, Dario asked for a promotion to vice president of research. Altman agreed and sent an email to the board saying Dario would report directly to him and receive equal PR treatment to co-founders Brockman and Sutskever.

A useful reminder that humans have feelings, and petty grievances shape the world.

www.wsj.com/tech/ai/the-...

OpenAI in Talks to Raise At Least $10 Billion From Amazon and Use Its AI Chips Amazon is in talks to invest $10 billion or more in OpenAI, according to three people familiar with the discussions. The valuation would be higher than $500 billion, one of the people said. The Amazon...

Scoop: OpenAI is in talks to raise at least $10 billion from Amazon and use its AI chips.

Read more from Anissa Gardizy, Sri Muppidi, Cory Weinberg and Amir Efrati 👇
https://thein.fo/4oUnAfZ

This is all crucial context for 2026, when many of these companies are planning to double down on their enterprise business.

When asked, the AI labs themselves say their clients are seeing strong ROI.

Screenshot from a story that says: Large language models are rapidly conquering complex tasks in math and coding, but can still fail at comparatively trivial tasks. Researchers call this contradiction in capabilities the “jagged frontier” of AI. “It might be a Ferrari in math but a donkey at putting things in your calendar,” said Anastasios Angelopoulos, the CEO and cofounder of LMArena, a popular benchmarking tool.

We also spoke with @Klarna, @Verizon, @Zendesk and others who say AI is amazing at certain things, but highly limited in others. This is often called the “jagged frontier” as explained by @ml_angelopoulos.

Cando Rail and Terminals found that no model could consistently + accurately summarize a ~100-page doc of Canadian safety rules. tc.canada.ca/sites/default/…

“We all thought it’d be the easy button. And that’s just not what happened.” - Jeremy Nielsen, GM at Cando

Case in point: @cellartracker built an AI sommelier to help users figure out if they’d like a certain wine. But it wasn’t discerning enough!

“We had to bend over backwards to get the models (any model) to be critical and suggest there are wines I might not like,” per CEO Eric LeVine.

AI promised a revolution. Companies are still waiting. Last spring, CellarTracker, a wine-collection app, built an AI-powered sommelier to make unvarnished wine recommendations based on a person’s palate. The problem was the chatbot was too nice.

New from @Reuters: executives *really* want AI to work but getting there is much harder than expected.

Many find the tech v difficult to harness or too unreliable for core functions. Others find their customers actually like humans.

www.reuters.com/business/bus...

Excerpt from a Washington Post story about Olivia Nuzzi and Ryan Lizza with the following sentence highlighted: This new iteration of Nuzzi, though entertaining, is a backstab not just to her former fiancé, but to all journalists who, toiling for insubstantial compensation, wouldn’t consider getting romantically involved with a source or subject.

www.washingtonpost.com/opinions/202...

I’ve spent almost seven years covering Meta and have had a good deal of exposure to internal work product on teen well-being. But a Friday filing in the school district social media litigation breaks some ground on Meta’s causal findings regarding social media harm.

www.reuters.com/sustainabili...

This from WaPo puts Musk’s pay package in perspective. www.washingtonpost.com/technology/i...

OH MY GOD

What I learned about productivity this year What I gave up, what I kept, and what's new. PLUS: How I'm using AI

"[T]he best productivity hack in the world is simply liking your job. If you enjoy what you do, you will find ways to do it, and in 2025 almost any software remotely suited to the purpose will be more than good enough." - @caseynewton.bsky.social www.platformer.news/productivity...

Data Brokers Face New Pressure for Hiding Opt-Out Pages From Google After reporters found dozens of firms hiding privacy tools from search results, US senator Maggie Hassan insists the companies explain their practices—and pledge to improve access to privacy controls.

NEW: Reporters at @calmatters.org and @themarkup.org ‬ found dozens of data brokers hiding opt-out tools from search results. US senator Maggie Hassan is now pressing top offenders to disclose their practices and pledge to improve access to privacy controls.

By me @wired.com

Congratulations to greenhouse gas

Still thinking about this illiterate line in the Trump WSJ lawsuit

Photos: Here Are the Piles of Used Bedding and Children’s Play Sets Left Near DOGE’s Old Offices Elon Musk’s DOGE minions have largely moved out of the General Services Administration. A dystopian scene of bedding and discarded kids’ toys remains.

SCOOP: Elon Musk’s associates at DOGE have begun vacating their headquarters at the General Services Administration in Washington, DC, and they’re leaving behind an apocalyptic scene of toddler play sets, mattresses, and pillows.

We got photos. 👇

When users did not see an AI overview, the authors write, they clicked on links 15 percent of the time. When they did see one, that number fell nearly by half, to 8 percent. The study also found that people are likelier to stop browsing after encountering an AI overview, suggesting that the overview has satisfied their curiosity. Researchers saw that outcome 26 percent of the time when AI overviews were displayed, up from 16 percent when they were not.

A new Pew study offers data to back up what we have suspected for a while now: AI answers are draining all the traffic out of the web www.platformer.news/google-ai-ov...

Even OpenAI suggests bringing back blue books Where does that leave writing teachers?

New in The Important Work: Jenny Lederer and Jennifer Trainor share thoughts about blue books, AI in the classroom, and the value of human interaction after attending an OpenAI webinar sponsored by CSU. theimportantwork.substack.com/p/even-opena... Illustration from @betterimagesofai.bsky.social

This report may have been a good time to note Marc will be testifying live in court this week along with several board members on the largest privacy settlement ever and allegations of insider trading for Zuckerberg and a coverup to protect him. Sandberg, Zuckerberg next week @nitasha.bsky.social

OpenAI Leadership Responds to Meta Offers: 'Someone Has Broken Into Our Home' As Mark Zuckerberg lures away top research talent to Meta, OpenAI executives say they're ‘recalibrating comp,’ according to an internal memo.

NEW: In a forceful memo, OpenAI’s chief research officer has responded to Meta poaching the company’s employees:

“Someone has broken into our home.”

By @zoeschiffer.bsky.social

www.wired.com/story/openai...

OpenAI Starts Selling ChatGPT at a Discount, Hurting Microsoft OpenAI has started discounting enterprise subscriptions to its ChatGPT app when customers agree to spend money on additional AI products, according to an OpenAI spokesperson and two executives at larg...

Exclusive: OpenAI Starts Selling ChatGPT at a Discount, Hurting Microsoft

Microsoft Loses Copilot Deals After OpenAI Discounts ChatGPT

Read more Aaron Holmes and Sri Muppidi 👇
www.theinformation.com/articles/ope...

The 560-pound Twitter sign met a fiery end in a Nevada desert The 560-pound bird logo that used to adorn Twitter's office was blown up in the Nevada desert as part of an elaborate stunt to promote an online marketplace app.

Remember how the Twitter bird sign from its old SF office was auctioned for $34,000 earlier this year? I can exclusively report the fate of Larry: his new owners, a marketplace startup, blew him up in the Nevada desert (with Cybertrucks on the scene) in a massive, fiery spectacle

(Bloomberg) -- Elon Musk’s artificial intelligence startup xAI is trying to raise $9.3 billion in debt and equity, but even before the money is in the bank the company has plans to spend more than half of it in just the next three months, according to deal terms shared with investors. The rate at which the company is raising funds and tearing through cash offers a stark illustration of the unprecedented financial demands of the artificial intelligence industry, and the relatively meager revenues it is bringing in so far, at least when it comes to xAI. Musk’s startup, which is responsible for the AI-powered chatbot Grok, expects to burn through about $13 billion over the course of 2025, or more than $1 billion each month, as reflected in the company’s levered cash flow, according to the people familiar with the deal terms who asked not to be identified because the information is private.  As a result, its prolific fundraising efforts are just barely keeping pace with expenses, the people added.  A spokesperson for the company declined to comment.

LOL yeah I hope they're raising equity.

*MUSK’S XAI BURNING THROUGH $1 BILLION A MONTH AS COSTS PILE UP

They Asked an A.I. Chatbot Questions. The Answers Sent Them Spiraling.

Fantastic and troubling piece by @kashhill.bsky.social about AI chatbots leading people into delusional behavior. One man whose wife was affected said he doesn’t think the AI companies fully understand what they are doing. “You ruin people’s lives,” he said.

www.nytimes.com/2025/06/13/t...

Fake Images and Conspiracy Theories Swirl Around L.A. Protests

“‘Information warfare is always a symptom of conflict, stoked often by those in power to fuel their own illiberal goals … It confuses audiences, scares people who might otherwise have empathy for the cause and divides us when we need solidarity most.’”

Sources: Meta agrees to take a 49% stake in Scale AI for $14.8B, in a deal that gives cash to Scale's shareholders and makes Alexandr Wang a top Meta executive (Cory Weinberg/The Information)

Main Link | Techmeme Permalink

Here's Kate's perspective from the NYT. On the WaPo biz/tech team, we reporters often write multiple potential headlines and then test them among ourselves in a Slack channel. Sometimes overruled by homepage / editors but it's cool being able to focus-group ideas and often they end up way better

The MAHA Report Cites Studies That Don’t Exist The Trump administration’s “Make America Healthy Again” report misinterprets some studies and cites others that don’t exist, according to the listed authors.

“The paper cited is not a real paper that I or my colleagues were involved with.” Just amazing from @notusreports.bsky.social www.notus.org/health-scien...