In partnership with
News of the day
1. Anthropic reports AI-driven workforce changes, a 5% public benefit fund, and Replit shifting its default model to OpenAI. → Read more
2. Dozens of WhatsApp help pages accidentally exposed hidden Meta chatbot instructions, including rules against legal advice and regulator referrals. → Read more
3. Thomson Reuters explores how to build trust in legal AI, leveraging OpenAI's ChatGPT Enterprise and their own CoCounsel Legal and HighQ platforms. → Read more
4. NVIDIA, EBU and Adobe Frame.io expand sovereign AI and accelerated video search for media producers and broadcasters at Berlin event. → Read more
Our take
Hi Dotikers!
This week, Anthropic gave us a rare glimpse into what happens when an AI company drinks its own champagne. An internal report reveals the company is hiring far fewer junior engineers, while more employees now spend their days supervising multiple instances of Claude at once. The bottleneck is no longer writing code, it's reviewing the mountains of output the machines produce. And in a twist nobody saw coming, meeting time per employee went up 20 percent. Turns out when robots do the typing, humans do the talking.
Alongside this, Anthropic announced it will commit 5 percent of its shares to a new Public Benefit Fund, with the board able to add 2 percent for every 50 percent jump in company valuation. The fund could finance large nonprofit donations or even direct payments to workers displaced by AI, going beyond the 200 million dollars already pledged to study labor market impacts.
Let's be honest about the timing here. Announcing a generosity fund just before what could be one of the biggest IPOs in history is smart positioning, and the governance details are conveniently promised for "the coming months." That said, Anthropic deserves credit for being the only major lab willing to say out loud that its own products are reshaping its workforce, starting with entry level jobs. That transparency matters, because if it's happening inside the company building the models, it's coming for everyone else.
The week also brought a bruise: Replit's CEO declared the company is done working with Anthropic, making OpenAI its default model provider. A disputed claim, apparently, but a reminder that even AI darlings can get ghosted by their biggest fans.
More on all of this below.
G.
The AI notetaker that gets the hard words right
Your AI is only as good as what you feed it. Feed it a transcript with your product names, acronyms, and numbers spelled wrong, and you spend the afternoon hand-fixing the output.
Wispr Flow Notetaker uses your dictionary and calendar, so product names, acronyms, numbers, and uncommon names come out spelled right. It pops up when your meeting starts and captures Zoom, Google Meet, Teams, and Slack huddles with one click, no bot joining the call.
Then your meetings show up inside Claude or ChatGPT through the built-in connector. No copy paste. Try Notetaker free on Mac and Windows.
Meme of the day
