I'm one of the first 100 users of Meta's Project Hatch, the company's answer to AI assistants like Town, OpenClaw, and agents from Anthropic and OpenAI (a perk of still being a well-connected former Metamate). Here are my early thoughts: 1. The interface feels far more like texting than any other AI assistant. I use Town religiously but Hatch feels more light-weight and far easier for a non-tech savvy person to use. Makes sense given the consumer GTM 2. It can do surprisingly powerful things with far fewer approval steps compared to the competition. The friction of approvals with other AI assistants can make those experiences feel pointless ("I might as well have done it myself!") 3. Each user gets their own virtual machine to run their Hatch. No one else in the market can claim that. Huge for privacy and security. 4. The initial suggested routines / tasks feel off the mark and not useful. These are supposed to improve with time as Hatch learns more about you. There are improvements coming that with the right marketing will put Meta's Hatch in play as one of the best AI assistants available worldwide. If not the best. The power of this AI will have many people feeling like we're entering a Black Mirror era in AI. It's powerful, very helpful day to day, aaaaaand borderline scary. It's noteworthy the power of this assistant is all free today for those with early access -- who knows what the company will charge in the future. However, for the first time I can now see a difference between a regular AI plan (a Honda Accord) and having no ceiling in what you pay for (a Ferrari). It's truly a different AI experience. Once Hatch is released publicly, I'll share more about the Ferrari experiences this AI makes possible because some of the stories are wild...

We are proud to back this world class team that is building something HUGE! Congrats Jacob, Kelechi, and the entire @Novig team on your Series A🚀