“A lot of the early OpenAI people have achieved financial independence already. At some point, people are seeing a $25 million offer, and they’re just like, ‘Well, after tax it’s half that.’” new from me, inside the AI talent draft:

Exclusive: Microsoft’s Nadella Pressures Deputies to Accelerate Copilot Improvements In recent months, Microsoft CEO Satya Nadella has begun cranking up the pressure on deputies to improve the company’s Copilot AI assistant. Read more from @aaronpholmes, @erinkwoo and @Amir 👇

pre-holiday news dump: Google finally sued SerpApi, a startup that scrapes its search results. We wrote about how OpenAI uses SerpApi in August, noting that Google hadn't sued—maybe due to regulatory scrutiny. That case is now resolved, and here's the suit! https://t.co/y8EqaiMuwQ

new from me this morning: a look inside how Google allocates compute, one of the most important resources of the AI boom https://t.co/4kvljjZ6Jp

Exclusive: Inside the Balancing Act Over Google’s Compute Crunch Google has created a group of top executives to decide how to allocate computing capacity, as it struggles to balance competing demands. Read more from @erinkwoo 👇 https://t.co/tQeY4Cm2gk

Sources: Google has formed a new executive council to allocate computing capacity given a critical shortage; members include Thomas Kurian and Demis Hassabis (@erinkwoo / The Information) https://t.co/fqhkbP5yUl https://t.co/TLHKibMuhT 📥 Send tips!

🚨We’re launching an AI infrastructure newsletter! Building on three years of leading coverage, @theinformation is doubling down on our smart reporting and analysis of the physical infrastructure behind the AI boom. I’m co-leading the newsletter with Ann Davis. Every Monday

What happens when increases in model intelligence don't benefit your most important, revenue-making product? OpenAI is finding this out the hard way. W/ @srimuppidi @amir: https://t.co/LR7e8suOM9

News: Google announced an AI partnerships pilot with news publishers involving experimental features in Google News, including AI-generated article overviews on the publications’ Google News pages and audio briefings https://t.co/Gm0TUmN7hZ

From my communication with MR. X, here’s how he explained why Ai is on the side of humanity based on the volatility in $NXXT (NextNRG Inc.). Hey VR, There’s an analysis circulating that points to a high-probability large move forming, with big upside momentum. The forces

one must imagine sisyphus a little bit sick of his meal prep all the time

anyway Snap later turned off by default the paid "solar system" feature that let you see where you ranked in other people's best friends lists, following reporting from the WSJ that mirrored concerns first raised in this piece

She got it at the comparatively ancient age of 14, which she described as "mind-blowing":

Re the Australia social media ban, I always think about one tidbit from this piece: One teen told me her friends got Snapchat at age EIGHT. As someone who got it at 15, that struck me as a crazy young age to have that complex an online social life. https://t.co/sQhWU1WxJb

You heard it here first folks https://t.co/7jlnYjEA1N

and final judgment here:

full memorandum opinion here:

New: Judge Mehta's final judgment on Search remedies! He sided with the DOJ in ruling that Google can't force partners to distribute Gemini to license other Google apps: "Google cannot be permitted to replay its illegal conduct with its GenAI products."

Also new from me this morning: the DoD is getting ready to release its https://t.co/YVitAa1xm6 chatbot portal. But the most recent version only includes Gemini! The project maven days are over... This is what it's going to look like:

gauche to love a sunset, which insists so loudly upon its own metaphor, and yet and yet and yet

Scoop central! OpenAI has its response to Google’s Gemini 3 and it has a funny allium themed name. “Garlic,” the company’s new pretrained model, is performing well on coding and reasoning benchmarks, according to an internal memo. https://t.co/IMjfDhufnL

In our piece last week about OpenAI grappling with mental wellbeing of users, we reported that the company declared a Code Orange in October because the new, safer version of the chatbot wasn’t connecting with users. Now it’s a Code Red.