Robby Starbuck rose to fame as an influencer campaigning against Diversity, Equity and Inclusion (DEI) in the USA. He often sued companies to force them to end such policies, and even sued Meta after their AI wrongly implicated him in the events of January 6.
Now the suit is settled and he has a new job offer; as a Meta advisor to address «ideological and political bias» in their AIs. This is what Trump meant when he went against «woke» AI, and Meta says they have made «tremendous strides to improve the accuracy of Meta AI and mitigate ideological and political bias» since working with Starbuck.
More at The WSJ, The Verge, Mashable and MSNBC.
Gemini now defaults to remembering previous chats
Google Gemini’s new feature is always on by design, and will remember your older chats without specifically asking. The feature delivers «more personalized responses the more you use it,» Google says. It will remember «key details and preferences you’ve shared, leading to more natural and relevant conversations, as if you’re collaborating with a partner who’s already up to speed.»
It can be turned off by going to Settings, then «Personal context.» There is also an option called «Temporary Chats» that won’t be remembered.
More at Google’s launch post, The Verge and 9to5Google.
Co-founder of x.ai leaves company to pursue venture capital
Igor Babuschkin, who joined forces with Elon Musk to create x.ai and turned it into a Silicon Valley darling in less than two years, has decided to leave the company. He gives no reason for the departure other than his desire to work with venture capital in the AI space, with an emphasis on safety and to the betterment of future generations. W
hile waxing poetic about x.ai in his departure post, he also says that «Elon and I realized that we had a shared vision of AI used to benefit humanity,» and we can all guess that was before the sexy robots.
More at: Babuschkin’s exit post, writeup on TechCrunch.
Today we're launching new ways to learn in Claude Code and the Claude app.
First up: Claude Code now lets you customize communication styles with /output-style pic.twitter.com/vU6p8cIZbF
— Claude (@claudeai) August 14, 2025
Anthropic reveals learning mode
Catching up to Google and OpenAI, Anthropic continues its busy week by rolling out a couple of new features for Claude Code. Instead of having it run through code on automatic, you can have it explain its reasoning step by step, or use the code as a training project, teaching you the solutions it is developing. T
hey are also rolling the features out all app users with a new «Learning style» that guides you through «hard concepts rather than providing immediate answers.»
More at Anthropic’s launch tweet.
Romanticizing kids? That’s allright, according to leaked Meta AI guidelines
Reuters got access to a leaked moderation document from inside Meta, laying out some ground rules for dealing with sexualization, sensuality and even racism, yesterday. It attempts to draw a line between acceptable and noncompliant use of sensualizing kids, and finds it okay as long as it doesn’t explicitly sexualizes them. So sensual and romantic interactions were deemed «acceptable.»
It also says it’s okay to help explain how black people are dumber than whites, as long as it doesn’t dehumanize anyone. Meta says these guidelines were a mistake, and that «The examples and notes in question were and are erroneous and inconsistent with our policies, and have been removed.»
Read the scoop by Reuters and comments on The Verge.