the age of agents is such a beautiful thing for solo founders you can scale your impact like never before literal bundles of state of the art intelligence just WAITING for you to command them it's just crazy to think about, and probs why i'm all in

with the latest Claude Code beta feature I've changed my mind you can get subagents to execute specific workflows with the NEW SlashCommand tool call so my view of subagents is now: - specific context injectors - decision makers and task routers AKA "which slashcommand workflow

It's arrived...and better than Opus 4.1.

if sonnet 4.5 is BETTER THAN opus and cheaper... you already KNOW Opus 4.5 is going to reignite that magic we felt not just 3 months ago

Since Claude just launched its Sonnet 4.5 update, now's probs a good time to share this for y'all to get prepared: 2 years ago, I said AI would become our operating systems. A bunch of followers said “no” and “never”... Let me tell you this now: In the next 12 months,

I'm driving blind here...

so they lured us in with a brand new model... to distract us from the fact usage limits have become 5x as strict apparently we're ALL in the 1% of powerusers

my favourite use case of Claude in Chrome: UX testing and bug hunting, then producing actionable plans in Notion and Github it's surprisingly smooth and a unique experience that's hard to replicate in claude code - let alone do in any other tool very valuable

deactivating for now have a son incoming

In the early days... Mind-numbing boredom. Physical exhaustion. Social discomfort. All expected prices of admission. Onwards.

A thing often in common among great startup investors, founders, and researchers: Trading making a lot of small mistakes in exchange for getting a few giant wins. (Surprisingly many people seem to prefer a few big mistakes in exchange for a lot of small wins.)

Tom Mueller on SpaceX's early days: “If people were negative, they were not in the next meeting.” “Elon was technically very strong, even though he was just learning about rockets. The one thing I noticed, if people were negative, they were not in the next meeting. He said, a

I have a pro google account Used Gemini 3 high for 1 prompt It read 3 files and hit the quota limit ...really?

Opus 4.5 is surprising. - Less usage than Opus 4.1. - Costs one third as much - Outperforms Gemini 3.0 Pro. Anthropic also boosted Max users. Opus 4.5 now gets the same allocation Sonnet 4.5 had before, plus extra Sonnet. They even said Opus 4.5 is meant for daily work.

After playing Arc Raiders over this week I can 100% say that Embark studies will sell Arc AI "people hunting" training data to the military.

GitHub is your project manager. Claude is your developer. You're the founder making decisions.

use the Codex mcp to help extend your Claude usage + improve your debugging workflow with this: paste this in your terminal: claude mcp add codex -s user -- codex -m gpt-5.1-codex-max -c model_reasoning_effort="high" mcp-server

remember, whenever you're stressed and overwhelmed it's ALWAYS a skill gap