Developer @ AV Labs 🇨🇦 - Genevieve and Anindya, I love you.❤️
Toronto
Joined May 2025
- Tweets1.2K
- Following960
- Followers219
- Likes1.4K
Everyone should have Bitchat on their phone. You never know when something’s going to crack off.
Messages hop from phone to phone over Bluetooth, with nearby Bitchat users relaying them. Private messages are end-to-end encrypted. No internet or mobile service needed for the local mesh.
There’s no central server to switch off, but offline reach depends on nearby users. It can also be removed from your region’s App Store.
Wherever you are in the world, download it before you need it.
apps.apple.com/app/bitchat-m…
Quote-post:
Fact-check sources: [Official Bitchat documentation](github.com/permissionlesstec…) · [India App Store removal report](techcrunch.com/2026/10/03/ja…) xitter.cf/greybtc/status/2107564…
If you like Claude, you should be rooting for OpenAi to make a comeback. Their competition is the only thing keeping Claude fair and affordable. If Claude had their way, you’d all be paying thousands a month through api usage and your speech and project ideas would be entirely at their approval.
OpenAI's 28-day "ship something every day or you get a reset" campaign looks generous. I think it's actually a way to hand out fewer resets without anyone getting angry, and likely because they have a compute problem. Here's my read.
Honestly, I also struggle to not grind my teeth when I see users cheering on large AI corporations, or any corporation for that matter. We have to remember that, at the end of the day (or fiscal year), they don’t really care about us, as long as we are paying. They will always give as little as possible to retain our subscription. They don’t give anything for free. Ever.
Every task, or project, no matter how small, should be started with your agents first. Plan it with the agent, then move it to a project. Otherwise, you're creating more work for yourself and making it more difficult for your agent to run operations autonomously later.
I haven't yet seen Sol 6.1 on Ultra spin up a significant number of agents in parallel. Astra on Ultra is a beast and does this well. Yes, I'm aware of the usage. The point is that Sol 6.1 should be running more agents in parallel or its not "Ultra." I'll update if I see a change
Claude or Codex, the models are only as smart as you are. In October 2026 there’s nothing you can’t build.
