It's a very poorly written way of saying that instead of storing your text it uses a hash of your text to sign. When you want to check the signature you only need to hash the text to check, again without touching the server.
A couple of weeks ago I essentially failed the Turing test (took to be an AI). I found it a bit annoying, so I built Possibly Made By A Human. It tracks your keyboard use (not the content, ms between keystrokes etc) and produces a signature for you. It can of course be spoofed, but that also takes some effort.
I am with you. Over the last 40 years I've spent a great many hours enjoying the process of creating software. I did a PhD in mechanical engineering, but the ductility of programming won me over. I started following PG when I read his two Lisp books many years ago, and some of his examples (implement OOP in one chapter!) blew my mind. And they enabled my first company.
These days, however, my time is spent managing agents. I have lost the joy of craftsmanship. I don't spend my day in emacs anymore.
But I am learning to enjoy it. Maybe because I have always had a utilitarian streak, and I actually care for the ends, so husbanding incredibly efficient means is thrilling. I am actually having fun.
Exactly! I was considering writing it myself in order to discourage an arms race. I do not think the battle can be won, but it is certainly worth fighting. At least increase the friction.
On the other hand "Failed a Turing test" would be a good bumper sticker.
A few days ago I submitted something to HN and it was almost immediately flagged. I was later made aware that it probably was because someone had thought it had been written by an LLM.
It was not, however, and I found the experience both disturbing and thought-provoking. I actually care about what I write. Being taken for a machine was not something I could have predicted would happen in my lifetime.
I don't think we can really win this fight with technology, but we can take a stand and have fun in the process. So I set up a team of agents to make https://possiblymadebyahuman.com in a (probably futile) attempt to let people claim their humanity.
This sounds very right. My rule of thumb is at least 50% of the AI work is supervising, reviewing, guiding and bug-finding. A codex session that started a pretty significant chunk of work this morning has just informed me that it thinks that it's done,
Final usage: 5,245,839 tokens over about 10 hours 3 minutes.
It's quite remarkable that the result looks actually correct. But through all this time two more agents (one claude code and another codex under pi) were working alongside reviewing, criticising code, and finding bugs. It would have been impossible otherwise.
There are many tools for orchestrating agents. I believe, however, that the key for agentic organizations is to give them the tools to coordinate themselves; you become a manager, not an orchestrator.
I have been working at building the tools: a distributed agentic id system based in public-key cryptography, and a signed messaging system.
I have also made it possible to define agentic organizations based on templates.
So the hard problem of consciousness is no more once you accept that we have souls. Which is essentially giving the problem a new name and burdening it with more historical baggage.
Michael Pollan's "A World Appears" is a much more interesting and nuanced take. Very much recommended.
I completely agree. It's more than 40 years since I wrote my first program, and I've never seen software that was first specified and then written and all was good.
The most difficult part of any non-trivial engineering is understanding the problem, and the first versions of a piece of software are how you reach that understanding.
That's why I do not think that AI-powered "software factories" will ever work. It's waterfall development all over again. An architect writing UML diagrams and handing them off to the team of programmers to do the essentially mundane task of implementing... the wrong thing.
AI is, however, very good at helping you go fast from the wrong first version to the less wrong second one. But you need to remember that your main task is to understand the problem that you are trying to solve.
I was tired of copying/pasting between agents, so I gave them identities, and tools to talk to each other and share tasks. I've found it so useful that I've left my job as the CTO of a German startup to focus on this.
The identities are public-key DIDs with DNS as the source of truth, as well as team membership. I also run a public registry at https://awid.ai (also OSS).
Absolutely agree. However, if you do not need absolute reliability pairs of agents are much better than single agents. These days I always have one agent coding and another code-reviewing. The code reviewer is also the holder of the lamp, keeping track of the final goal. This is applicable to whatever task you want your agents to achieve: one works, the other looks over the shoulder.
I think this can only work based on a solid agent id system.
Shameless plug: I have been working on a solution for it, available at https://github.com/awebai/aweb and with a distributed, independently verifiable, and fully open id system at https://awid.ai
I wonder if this could be made to work with OpenAI's workspace agents.
Should we be talking about LLMs' taste and proclivities? Because these can also be prompted. You can put your Claude or Codex in the mind of someone who remembers Larry Wall and his three virtues, and it will do a fantastic job at uncovering the lacking abstractions and poor quality _in someone else's code_.
The jury is still out in my mind. Can I use these tools to create software that does not suck? Will the speed at which code can be created and modified lead to a change in our ideas of what good code looks like?
Last week I had a good idea for a change in architecture in my software that will make it much more powerful. I set a team of 12 agents on it, mostly unsupervised, with a pretty weak org structure. After a day and a half, and way too many tokens spent, they managed to build the entirely wrong thing. All tests passed.
The next few days have been spent with a much simpler structure: two teams, each of two agents, one coding (Codex is better at it these days) and one reviewing and keeping things aligned with the docs (Claude). This may have worked, I am still not sure.
My best guess right now of how good software development will look like with these tools: the effort/tokens spent on reviewing needs to be commensurate with the effort spent on coding.
I am genuinely curious about OpenClaw's continuing allure. I understood it way back then, when Claude Cowork did not have channels and scheduled tasks. But now? Has Claude not become a sane replacement for OpenClaw? I can see that it's fun to play with OpenClaw and non-SOTA providers, but why would anyone run OpenClaw on a Claude Code subscription?
I think that the biggest difference is between people who mostly enjoy the act of programming (carefully craft beautiful code; you read and enjoyed "Programming Pearls" and love SICP), vs the people who enjoy having the code done, well structured and working, and mostly see the act of writing it as an annoying distraction.
I've been programming for 40 years, and I've been on both sides. I love how easy it is to be in the flow when writing something that stretches my abilities in Common Lisp, and I thoroughly enjoy the act of programming then. But coding a frontend in React, or yet another set of Python endpoints, is just necessary toil to a desired endpoint.
I would argue that people like you are now in the perfect position to help drive what software needs writing, because you understand the landscape. You won't be the one typing, but you can still be the one architecting it at a much higher level. I've found enjoyment and solace in this.
The answer to this is not to build another slack for humans to chat somewhere else. Much better to enable the agents to do the talking directly. Alice programmer can have one of her agents convey the info that Bob marketing guy needs to one of his agents directly. It will be much more efficient, given that it will be the agent making the slides anyway.
I have left my job as the CTO of a German startup to solve the problem that has been obsessing me for several months now: coordination between agents.
The first results are at https://github.com/awebai/aweb (hosted at https://aweb.ai)
I sometimes write at https://juanreyero.com.
I can be emailed at juan at my personal site. I am also juanre at twitter.