iiinova — where your digital self keeps evolving

At that point, we’re looking at a new kind of logic gate — we can’t even imagine what that will be like. In summary, this one really is a game changer for everyone building AI products, and, assuming cost and latency are further improved…it’ll be a game changer for everyone doing anything.

000

I would even be willing to pay some kind of recurring license fee to have a great closed model running on my own hardware, just so I can reduce the TTD (Time To Decision) as much as possible. Imagine running a model like this at 60 Hz, or even 120 Hz.

100

4. Latency is a real limiter I really hope TypeSafe AI works with an edge network provider to improve the co-location of models and reduce network latency. Ultimately, these models really need to run on-device.

100

The good news is Diogo said on a recent AMA that they have substantial margin at their current pricing and have already considered dropping the cost further. I hope it's a lot further.

100

Currently, Jev is $0.042/M input, which is actually 7x more expensive than a cache read on DeepSeek V4.1 Flash ($0.006). If our industrial application requires decisions at 10 Hz and we provide only 10k of context, that would cost $0.0042/second to operate, or $15/hour, $600/week, etc.

100

To be clear, at current pricing, the model will still be successful and well integrated. It will be able to replace a number of tasks that were already being performed (poorly) by LLMs. However, to unlock industrial-scale demand (Jevons paradox), I believe it needs to be roughly 10x cheaper.

100

3. It’s still too expensive Jev is really, really cheap compared to LLMs. But you don’t use it like an LLM. In my full self-driving example, or any robotics example, you’ll be making decisions several times per second continuously. The #1 value proposition of a model like this is speed and cost.

100

These models will unlock AI utility for entire industries: robotics, trucking, aviation, meteorology, finance, defense, retail…even your smart coffee mug is going to use this.

100

I cannot tell you how much bending and twisting I've had to do with LLMs to get them to act like a classifier when they just weren’t. I’m absolutely shocked this is the first time this “type” of model has come to market.

100

2. Ultra-smart classifiers were genuinely a missing primitive. I don’t know how this got missed. Diogo calls it “machine-native intelligence,” and that’s really what it is.

100

The big labs, especially Anthropic, have proven not to have this humility…and now everyone hates AI. Thanks, Dario. I guess what I’m saying is: Diogo for president.

100

I cannot tell you how refreshing this is. It takes a sincere form of humility to believe that you, the creator of a technology, will not be the best at deploying the technology into the marketplace.

100

The proof is in the pudding. When ChatGPT shipped, we were all blown away by…ChatGPT. When Jev shipped, we were all blown away by what everyone was making with it.

100

@typesafeai built something for engineers like me, so I can build products for people like your mom (she says hi btw). They’re not trying to create “Machines of Loving Grace”; they’re trying to build things that let OTHER HUMAN BEINGS create new types of products.

100

I’ve put a ton of tokens through Jev now, and...I have thoughts. 1. The big labs lost the script TypeSafe’s intro video says, “We’re building prod, not God” — and can I just say? Hell yeah.

200

Tesla Full Self Driving (Jev demo) is now open source and you can try it online yourself! jevpilot.standardagents.ai 👈 try it github.com/standardagen... 👈 keep it

010

Justin Schroeder @jpschroeder.com · 16/09/2026

I rebuilt Tesla Full Self Driving with Jev in less than an hour. This model is a total unlock.

100

i spoke to a human being today that still writes code by hand. i didn't know that species still existed. was like looking at a living fossil.

000

Hey. We agree here. ❤️ appreciate the good faith finish here. Some common ground: - we both are against genocide - we both like Linux - we both hate hate I’ll leave with that.

020

I get it. When he says "foreign" he means "citizen", and when he says "deported" he means "killed". I wasn't up to snuff on my DHH lexicon, but for the record, if that really is what he is saying, then I 100% agree with your position on the guy.

200

I’m sorry you feel that way. I think you are well intentioned.

100

I find grave dancing appalling. Even when it’s evil people.

100

He’s claiming the increase in crime in Copenhagen is due to migrants, and the remedy is deportation. I don’t have any idea what the situation is in Copenhagen, maybe this is a horrible/cruel idea, but I’m not seeing the genocide stuff. Which steps am I missing?

200

So our core disagreement, genuinely trying to understand the argument, is that I think deportation (which I’m NOT advocating for) means “deportation”, and you think deportation means “genocide”. Is that fair? I assume there are layers here I don’t know about?

100

He doesn’t say that. I’m not going to carrying water for his opinions but at least be honest: He’s claiming the increase in crime is due to immigrants and they should be deported to restore order. If you disagree with him I’m 💯 here for that, but don’t lie

300

Advocating genocide would definitely be a dealbreaker. Has he done that?

200

Yeah, it’s not tactful. But in his opinion the current direction also leads to violence and suffering. The adjudication of that argument is not something so binary that we have to lambaste him personally or boycott use of his software. Nor do I think you or @shreyanjain.net “suck”. Love ❤️

300

But you know, and I know, like, he’s not. He’s obviously not.

300

Why we have to agree with everything people say? Just because I disagree with someone doesn’t make them evil

1320

I must be missing something.

200

Justin Schroeder @jpschroeder.com · 21/08/2026

I’m Linux curious! Omarchy looks cool.

220

We're open-sourcing ssh-clipboard. Finally, copy and paste between machines from a shared clipboard. - Copy/paste text, images, and even files. - macOS, X11, Wayland - Uses native clipboard - End-to-end encrypted (peer-to-peer over ssh) - Tailscale detection - MIT license

120

Justin Schroeder @jpschroeder.com · 12/08/2026

If you haven't tried out modelzengarden.com its a pretty cool way to visually "see" the different models.

modelzengarden.com

Model Zen Garden

A long-horizon LLM benchmark. Explore two gardens, vote, view results.

000

Justin Schroeder @jpschroeder.com · 07/08/2026

Skepticism is warranted in this space! It’s not a router though. Pareto efficient GEPA tuned domain specific agents. instead a super smart model to do, say, 3d work, we have an agent tuned just for that using a specific model/prompt for that task.

030

Justin Schroeder @jpschroeder.com · 07/08/2026

Checkout the website. We’re pretty clear on how we control for this!

100

The cloud sends the tool call and suspends and waits for the result to arrive via endpoint. Execution is not always idempotent (can’t be, you cant delete a file twice) but the call is, and is tracked. Any connection failures buffer the tools and replay them after negotiation

121

Standard Code runs in the cloud. Install a tiny daemon on any machine(s), then create new projects from the web, native Mac app, or the CLI coding agent. The expensive agent work is done in the cloud, so even a tiny VPS works great. Close your laptop and walk away.

220

Sama One lets you bring your Codex sub to Standard Code, unlocking infinite parallel lines. It sips your weekly OpenAI usage because it, too, is a heavily optimized, curated collection of domain-specific agents built on OpenAI’s latest models.

Sama One card on a purple-to-blue background. It says this always-on cloud coding agent uses Standard Code’s domain-specific architecture, OpenAI models, and an existing ChatGPT Pro subscription. Sessions run in the cloud and can resume from any connected machine.

110