Founder @actual_ai_ . Previous Head of Product @AWSCloud GameTech. @UnderscoreVC Core Partner.
Seattle, WA
Joined July 2012
- Tweets3.2K
- Following729
- Followers1.2K
- Likes98.6K
Pinned Tweet
Dynamic Workflows for Claude Code make it immeasurably more important to have ADRs. You might be looking at 20 exploratory runs vs getting it right the first time. 5% the token cost with ADRs. We have already seen this play out with open source software factories.
If you are not building a harness, you are missing out. Not necessarily a coding agent, just something, email, creative, organizer. Combining instant software prototyping with inference is magic.
John Kennedy retweeted
Gave a talk last night at the Claude Community Meetup here in Auckland on what I built for my best friend Bruce who died of cancer seven weeks ago.
I built an AI system to process all of his medical data from three different hospitals,
and created a 3D digital twin that realtime produced audio visual explainers of what was happening in his body, we could then ask questions and receive decision support from a total system of recorded + multiple AI models as we interfaced with doctors and nurses to make critical decisions day by day during his final days and hours.
Video of the talk incoming, I’ll share more soon.
Factory dealt with this badly, but a board observer joining a direct competitor with no gap is destructive behavior. Founders will remember.
Accidentally said AI instead of the greatest of all evils of which I only partake as I must to survive and they kicked me out of Seattle.
This is not it. Good devs understand that their job is now to engineer rules, context, and process. It is not horrible, but it does take discipline.
I am done with this shit. It is over. The state of engineering right now is horrible. It has been half a month since I started a new role at a big company. Nobody knows anything here. The specs, code, tests, PRDs, tickets, resolution of those tickets, reports, etc., everything is made by Claude Code. Nobody on my team likes this. They are being forced to ship as much as they can. I have heard multiple times from higher management that pushing code is not a bottleneck, so why are we slow? People are working 12 to 13 hours a day just to press enter. Nobody is reading anything. Humans in corporate are doing nothing on their own. Everyone, literally everyone, from an L1 to an L7 engineer here is doing the same thing. Talk to Claude. There is no sense of victory. Nobody is resolving bugs. In reality, nobody is thinking anymore. Everything is done by LLMs. It is so soul-sucking. I would not mind it, to be honest, if we were at least given the time to check out the code and see what is going where. But no, the goal is to just ship. No matter what happens.
Nice, but AI predictions without timelines are not useful. The capability landscape changes too quickly. Conservative predictions can be useful if they are limited to the next three months.
John Kennedy retweeted
we do things not because they are easy, but because fable oneshotted the first 80% so we thought it would take an hour max.
The SOTA blackhole is roughly equivalent to the @Microsoft Excel blackhole. Yes, you can do *everything* in Excel and F500 companies continue to do some pretty crazy things in Excel to avoid another vendor onboarding, but most others will want OOTB value.
John Kennedy retweeted
We ran Kimi K3 through 3 agent harnesses (Claude Code, Hermes, Kimi Code) on 28 identical tasks.
All 3 harnesses completed the tasks at similar success rates, but the interesting story is token efficiency: the same task cost up to 30x more tokens depending on the harness. 🧵🧵
Just to confirm, @natebjones is not AI-generated. He does in fact live in meat-space.
So @jack is now investing in Nostr Protocol instead of AT Protocol?
block.xyz/inside/introducing…
