I went through 140-odd @bot templates so you don't have to.
Most are toys. About forty would run part of a small business tomorrow.
Here they are, sorted by the job they do. Start with the two that check the others.
I still think the fire metaphor for AI is underused:
"Is fire good or evil?" That's a weird question
"Should we give everyone access to fire?" Yes, so they can cook and heat their homes
"Should we give everyone flamethrowers?" Probably not
"Should we give everyone firebombs?" Definitely not
"Should we require companies that produce enormous combustion reactions to do so safely?" Obviously yes
“Should we wait until the city burns down to establish a fire department?” Obviously no, and we should also build in safeguards like smoke detectors
Jacob is correct here—we really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade. I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to.
@free_ai_guides Interesting - I’m using something that combines all three to get my Instinct speaking to someone else’s Instinct, and it seems to work quite well so far…
Some weeks (not all) on Friday morning I have spare Claude tokens (the weekly reset hits at 10am). So I have a new ritual: one hour. Take a real product from its current version to the next. Ship nothing. Review everything.
At minute 60 you get a review file: every item with baseline, after-value, commit hash, screenshots side by side - plus the list of what didn't fit, which seeds the thinking for next week, to be validated against first principles.
Every firm in town is trusted. Almost none is distinct.
Trusted, responsive, tailored is not a position. It's camouflage.
We fix what you stand for, then build the engine that says it. One firm per patch.
Start with a Local Market Read.
https://t.co/Z0zMC9dEDP
I went through 140-odd @bot templates so you don't have to.
Most are toys. About forty would run part of a small business tomorrow.
Here they are, sorted by the job they do. Start with the two that check the others.
The useful bots don't do the job. They keep the record, check the assumption, or find the person. Judgement stays with you.
Capacity without headcount, not headcount without judgement.
They pay for themselves in a month.
Chained Oblivion — unused SaaS seats and recurring charges
https://t.co/ZkduEMY1hs
Watchdog — inbox receipts and trials, before they bill
https://t.co/1BE1fjewEC
Lockdown — checks your controls against SOC 2
https://t.co/QByrC0hSil