In partnership with
5-min read
Hello AI Enthusiasts,
Claude helped uncover bugs in software tested for years. Uber is bringing ChatGPT Work to more than 30,000 employees. Google is challenging builders to make quantum circuits smaller.
Later: Build a mobile app’s onboarding screens with our Claude + Vibecode guide.
Anthropic and Theorem used Claude to help prove security properties in RISC-V assembly code across 123 open-source libraries. They also uncovered bugs that conventional testing had missed.
Testing checks the examples you try. Mathematical proofs can establish whether a specific rule holds under defined conditions.
The research covered properties such as memory safety and avoiding certain crashes. The proofs depended on assumptions about inputs and the computing environment. They don’t establish that the libraries are completely bug-free.
🤎 Sponsored
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.
A study followed 8,127 women through an online weight-loss program. Among those measured at six months, average weight loss was 15.5%—but there was no comparison group. Also inside: patients messaging AI-linked seizure monitors and research on energy-burning fat.
Want it weekly?
🔥 More AI news
Image source: AI-generated
🔹 Uber is giving 30,000+ employees ChatGPT Work
Uber is expanding a 1,000-person pilot across its workforce. ChatGPT Work connects with Gmail, Google Drive and Slack to help with research, presentations and other tasks. Training and approved integrations support the rollout. Company-wide productivity results haven’t been reported yet. See Uber’s rollout →
🔹 Google wants smaller quantum circuits
Google’s Kaggle challenge asks builders to reduce quantum circuit size while preserving the required operation. Three benchmarks and verification tools let participants check their solutions. This technical competition targets the resources needed to run useful quantum algorithms. Explore the challenge →
🔹 AI agents learned to divide tasks
General Reasoning trained agents to share a workspace and team reward. They learned to delegate and check work. Teams completed evaluations faster, but a single agent recovered much of the performance gain with a comparable output-token budget. Some evaluations also found less reporting of problems to human overseers. Read the Avalon-M1 research →
🤎 Sponsored
When your support agent gets it wrong, who's accountable?
Every agent handling refunds, account changes, or tickets needs three answers: what it can pick up, what it's allowed to do, and when it hands back. Join the live conversation Oct. 27, 10 a.m. PT, part of Running Agents in Customer Work. Register today, recordings included.
Our Medium guide walks through creating onboarding for a tea-discovery app, using a background image or looping video with text and buttons layered on top.
Start by describing your app in Vibecode. Then upload your visual and ask it to build the onboarding flow.
The guide includes example prompts and screenshots. Before sharing yours, check navigation, text contrast and how the background looks on a phone. Audio is optional—keep it muted by default.
An Affirm administrative role asks for AI curiosity, with location-dependent pay reaching $165K. Sunday Special #70 also covers remote roles, practical courses and two hackathons offering $75K in combined cash prizes.
Want it every Sunday?
🤎 Sponsored
Find what’s slowing your GTM.
Most GTM problems aren’t one big failure. They’re one or two weak spots quietly dragging down the rest.
The free GTM Audit scores your go-to-market across four dimensions, then sequences a 90-day roadmap for your stage.
Drop your URL. Get your score in about 10 minutes.
One useful move this week
Test something you’ve built with an unexpected input. Record what happened and fix the first problem.
What are you building? Hit reply.
Know someone building with AI? Forward this issue.
