If more and more services have stopped accepting tokens in URLs because they keep leaking into logs. Then why do we still accept RPC providers slapping your API key in the URL by default?
Nathan retweeted
Own your intelligence.
Anyone that opposes this has something to sell you (perhaps your own intelligence back to you.)
Local models need to continue proliferating. The “absolute unit” massive models may be out of reach for now locally, so your best option is to, indeed, keep your options open. Own your harness, data, connections, runtime: your agents. Embrace open source.
Nathan retweeted
What’s happening in the AI-powered reverse engineering and decompilation is absolutely insane. Pretty soon all software will be de facto open-source.
AI is coming for everything and everyone.
AI PCB design is currently not great because the decision making process of designing them can't be simplified easily. The source data for training a model specifically for PCB design is not easily available, as many designs are proprietary and their creators don't fancy giving them away.
If @heypcbai get enough human assisted PCB repos or work out how to collect training data on the nuance and creativity that goes into PCB design we could get a model that makes it accessible for everyone.
GitHub wasn’t built for hardware.
So we built Circuit World.
An AI-native EDA + MCAD platform for designing, iterating, collaborating, versioning, and manufacturing hardware in one browser.
You don’t have to start every design from scratch.
Start with one of 8,800+ existing boards, build on it, and iterate with AI from PCB to mechanical assembly. Then send it straight to manufacturing.
heypcb.ai/world
Update your person vocabulary again!
slop shop, n.
/ˈslɒp ʃɒp/
Computing, slang, derogatory.
A software company or team that ships code rapidly and in large volumes, but whose output is of little or no value. It is often associated with overreliance on AI code generation tools.
"Four hundred releases this quarter and not one of them fixed the login bug. Absolute slop shop."
Derivatives: slop-shopper n.; slop-shopping n.
can I try and coin a new term:
Slop shop
Companies shipping code at breakneck speed but it's all worthless crap with bad UX
2
99
Add 'Slop Grenade' to your personal vocabulary alongside 'Meat Proxy'.
"Slop Grenade: Letting AI produce the work and pass it on without adding any value (including checking it)."
Lazy work used to mean too little output.
Now, with AI, it often means too much and more work for everyone else.
@tobi call it "slop grenades."
A "Slop Grenade" is when you let AI produce the work and pass it on without adding any value (including checking it).
Someone else has to wade through it, catch the mistakes, and clean up the mess.
You save time and look productive but someone else pays for it.
4
104
Nathan retweeted
astonishing blogpost from Shengyu Liu (刘胜与, also known as interestingLSY/intlsy), kernel engineer at DeepSeek. The first part is his personal struggle with the fact that his work is about to render his beloved craft obsolete. The second is…
well. let's just say we agree.
As everyone slowly becomes a meat proxy, all communication is going to sound the same.
There was already a downward trend, but it seems to be accelerating.
I believe writing is thinking. So just because generating words is commoditised, doesn't mean writing has been. LLMs have made writing more valuable because people are simply thinking less.
1
6
109
The paper is worth a read. It suggests that AI tool adoption is associated with convergence toward stylistic norms and reduced variability in scientific communication.
It also suggests llm adoption may diffuse at different speeds. arXiv users adopt it faster, while Reddit is slower, which might be because of Reddit’s broader user base using llms more gradually than academic authors which have more exposure to tech.
1
1
34
Full paper: arxiv.org/abs/2502.11266
17
Good take
I finally looked into Omarchy and I can see why dev twitter is suddenly obsessed with it.
It’s basically Arch with somebody else making all the hard/endless decisions for you.
Hyprland. Terminal first. Keyboard driven. Neovim. Plain config files. Very little “click through six settings panels and hope for the best”
This style of OS makes a lot of sense now with ai. An agent can read the config, edit the config, install the package , restart he service, inspect the logs and change the workflow.
Which is weird because the future of ai computing might end up looking a lot like the old way of using computers: files, terminals, scripts, text and a machine you can actually understand.
2
105
Today is a very historical moment for AI video generation
You can now generate AI video faster than you can watch it
Before it'd take let's say 2-5 minutes to generate 15 seconds of video
@fal made a post-trained Minimax H3 variant called Max which is 50x faster than the original but still maintains quality
It generates 15 seconds of video in 9 seconds!
That means you can now do new things like build a perpetual livestream with it that never ends!
Nathan retweeted
the biggest downside of coding with AI is how easy it is to blow past the MVP of a product and just keep on building






















