Founding engineer @schematikio | Building agents that build hardware | Teardowns of new models & coding agents | ex-Head of Product & Tech @thenextweb, @Vinted
Amsterdam, The Netherlands
Joined August 2019
- Tweets835
- Following246
- Followers440
- Likes4.4K
So don't sleep on Deepseek!
Their harness + Flash model together is a great value combo if you want a good level of work for a low price
If you want to learn more about Deepseek harness, I covered it here:
xitter.cf/itsjack/status/2088989…
this is the way to do it
something that keeps bothering me is how some people see slop as something that is happening to them, like they're some kind of helpless bystander powerless to do anything about it
excuse me but you are a software engineer, you literally learned how to make sand think and make computers do your bidding?! and now you can't do anything about it? please
the 2 things you should know to cut through all the noise:
1. you need to give your agent a way to verify its own work (i just posted about this so won't rehash)
2. your codebase is memory
the 2nd is not new either, we have always had lint rules, compilers with good diagnostics and strong type systems, LSPs, etc to help us write better code. there is no difference when agents write the code instead, you just need more of them, and stronger ones.
forget rules and skills. you can add those later.
yes, your codebase is a form of memory to agents! agents love extending what already exists. so every nit and every annoyance you've ever seen, every mistake and bad pattern you observe from agents, needs to turn into a constraint, or the agents will just keep repeating those same mistakes like a virus in your code. then you keep observing and repeating this process until your codebase is locked down enough that even agents with little context can one shot a high quality PR.
an even better direction is to refactor/rewrite your code such that agent anti-patterns become categorically impossible. use type systems to your advantage, use libraries/frameworks that reduce the architectural search space, or heck, create your own framework like we did for grok bot and design high quality primitives and well thought out APIs
The cool thing is that this outputs straight looking images because it flashed these calibration cards on the screens to position them
Can confirm that @bot is a whole lot better now
I quit using it - while the product experience is fantastic, the model was really showing its age
Now it can handle long running tasks + loses coherence much less than it did before
Smart move!
I made 𝚌𝚕𝚎𝚏 a night-club bouncer
You can set your admission rules and clef from @Cloudflare decides who gets in
introducing 𝚌𝚕𝚎𝚏: our first models trained by @cloudflare's workers ai team.
today, we're releasing two fast and accurate decision models that top the benchmarks for quality and latency. use them hosted on workers ai or grab the weights from @huggingface, because we open-sourced it too.
blog.cloudflare.com/clef-dec…
You can try it out here: bot-club.jadu.workers.dev/
