@NeatoRobotics I'm sad that my xv-11 had its drive motor spring plastic tabs break after only a couple of weeks on the job :(

RT Lucy Kafanov 🏳️‍🌈: RT Occupy Wall Street with a General Strike: For live coverage when the Feds roll in: Ustre.am/DVEZ #Ows #usdor

Verifying myself: I am egalanos on Keybase.io. Eq5fMqPorkEFOFW4PyTTDciDCL0BsDPIuIay / keybase.io/egalanos/sigs/…

🚀 Our community never ceases to amaze us! Check out this awesome app built in Tana by Emmanuel Galanos - Combining GPT with Tana Commands, they were able to replicate an approach which can reduce AI error rates by double digits! #BuiltInTana #AI #KnowledgeManagement #productivity

An animated continuation of some of the best TV sci-fi ever made

A useful ChatGPT Plus custom instruction setting (unless you live in the US): "Always provide units in metric units and never provide imperial units."

Very impressive demo

Why did OpenAI use thumbnails that look like they belong to a B grade horror flick?

Asking Nano Banana Pro its unfiltered thoughts on AGI ☕️ 🐶 🎩 🔥

Claude Code's new 'context: fork' directive for Skills/slash commands is such a powerful feature for primary agent context preservation. It just needs some less bloat with the <local-command-caveat>. Not sure about model being told not to consider the output though. hrmm

🦞 BIG NEWS: We've molted! Clawdbot → Moltbot Clawd → Molty Same lobster soul, new shell. Anthropic asked us to change our name (trademark stuff), and honestly? "Molt" fits perfectly - it's what lobsters do to grow. New handle: Mr. Lobster🦞 Same mission: AI that actually does

Claude Code's new persistent auto memory directory feature is worse than useless for worktree users: ${HOME}/.claude/projects/${worktree-dir}/memory/MEMORY.md Claude will now write memories that will never get used. Sigh.

Claude Code TUI has overly dumbed down the tool calls display - Read() does not even show the filename. Verbose mode can show it, but is too much detail. Not a good change. CC Thariq

Bad experimental design leads to incorrect conclusions. > Anthropic learning: "Users don't want high speed + large context" > Actual learning: "Users aren't willing to pay $225/M token for high speed + large context"

Claude Code v2.1.71 adds /clear to the conversation messages which then confuses Opus CC: Thariq a quick one to fix

Codex default sandbox is broken on Linux: Ran git diff 2a2d1810efcd15fed8417dcfb011a5887bfb6aaa --stat && echo '---DIFF---' && git diff │ 2a2d1810efcd15fed8417dcfb011a5887bfb6aaa └ bwrap: Can't mkdir /home/devuser/.codex/memories: No such file or directory Vaibhav (VB) Srivastav

Claude Code update made bypass permissions not mean it Version 2.1.78: * Fixed .git, .claude, and other protected directories being writable without a prompt in bypassPermissions mode FYI: Thariq this stops work where one expects it to be working but it gets stuck on a prompt

Claude Code sanity tip: pin to v2.1.77 to avoid incessant prompts editing "protected" config dirs. No telling when it will be fixed. curl -fsSL claude.ai/install.sh | bash -s 2.1.77 add to settings.json: "env": { "DISABLE_AUTOUPDATER": "1" }