ChatGPT Health is led by Instagram’s former VP of Product. The person who spent 12 years optimizing for engagement at Instagram and Meta - keeping you scrolling, clicking, coming back - is now designing OpenAI tools for health decisions. What do you think can be the consequences? And do you use AI for your health?

It’s been a while since I’ve posted here. Aside from the holidays, I have spent the last 30 days in full business recovery mode following a disastrous workspace migration in ChatGPT on November 30th. I didn’t want to come back until I had something positive to report. Over the summer, before joining the AI Advantage I launched my startup using a ChatGPT Plus account tied to my personal email. As a (now) professional founder, I eventually realized it was necessary to separate my personal data from my business for IP protection, liability, and future HIPAA compliance with the app I’m building. When I started my ChatGPT account I hadn’t even thought about starting a company yet. Everything was mixed up. All my personal topics and my business building topics were in one account. Which is not a good thing. I knew enough to upgrade to a ChatGPT plus account when I started iterating business ideas, because I did not want them being trained on any of open AI models. I did everything by the book. I followed the documentation and even used ChatGPT to build a migration checklist. What happened next took my business out at the knees. ChatGPT wiped out every last bit of my data. Their support was nonexistent—no apology, no refund, and no path to recovery. $632.00 spent for a two year Business Workspace and nothing to show for it, everything gone. While my "founder discipline" meant I had backed up my core IP and design docs locally, I lost months of context and chat logs that served as the backbone of my cognitive scaffolding system. It was a disaster of epic proportions. I’ll admit it: I cried. Sobbed actually. I’m normally not a crier, but I had worked so damn hard for several months building something that I know is going to help people. I refused to let it be the end. I have spent December reconstituting my operations from the ground up, but I didn't go back to what broke. After learning that this has happened to dozens, if not hundreds of other founders and businesses, I migrated my entire company to Google Business Workspace (Enterprise Standard) and Gemini. The difference in professional stability and support has been night and day. Because of this move, I’ve recovered enough momentum to stay on track for our first product launch this January.

Your own voice can now influence the development of AI. I am thrilled to see this interview, implemented recently. This is exactly the kind of feedback I wanted to become possible, and Anthropic is the first to offer it - initially to a sample of around 1000 professionals, and now to the general users. The outcomes of the first part of the research are available on Anthropic website (google up "Anthropic-interviewer"). Cool :)

Last week's news but still ongoing. 39% of publicly exposed web apps may be affected, with vibe-coded apps at even higher risk because many rely on default framework setups. A maximum-severity security flaw has been disclosed in React Server Components and Next.js that can allow unauthenticated attackers to gain full control of a vulnerable server via a single crafted request. In practical terms, this means potential access to your environment variables, secret keys, databases and backend logic. If you have any public apps: (1) check whether they use Next.js or any framework with server-side components, (2) verify whether your version falls within the affected releases, and (3) upgrade immediately to a patched version and redeploy. Non-vibe-coded apps can be affected as well. Edit: Hacker activity is massive now, to exploit this vulnerability. If your app has this door wide open, they WILL walk in freely. Severity: 10.0 (Critical) Impact: Full system compromise Target: mass exploitation. If you're not tech and have vibe-coded an app, read the comment below by @Alya Naters More info: I can't provide links because they're not allowed in this group, you can google up Security Advisory: CVE-2025-66478.

Well… apparently my little BOT DETECTOR post in the main AI Advantage community caused more turbulence than expected. It survived a whole 2 hours before getting gently escorted off the stage. Reason given: “It might confuse the thread.” Totally fine — I get it. But here’s the part that’s ironic: I posted it precisely to help beginners avoid getting confused. Because half the new members are getting approached by bots and have no idea what’s going on. My intention was simple: ✨ keep humans safe ✨ keep the thread clean ✨ teach new members how to recognize low-effort spam ✨ do it in a fun, memorable way so people actually pay attention But maybe my delivery was a little too… effective 😇 (Moderation bot: “THREAT DETECTED — remove.”) Still love the team — genuinely. No hard feelings. Just amused that the only thing in the community that successfully got flagged was… the anti-bot post 🤖💀 Anyway — since it vanished before some of you saw it, here are the screenshots and the explanation of what it was. If anyone still wants the track or the guide, I can drop them here safely — where context isn’t an issue and beginners actually benefit from it. Humans helping humans. What a concept 😌✨ Also people let us stay inside topic this time in the comments 😉🫰