Introducing System One Models and Jev • Hacker News Board

Re: Introducing System One Models and Jev

#261

Post by InsideOutSanta » Wed, Sep 16, 2026, 1:26 AM UTC

CompleteSkeptic wrote: post #141

WhitneyLand wrote: post #136

Earlier quoted context omitted.

His claim was that the title is misleading, not sure how it's relevant to that claim that you use "string models" (full LLMs). The original title before it changed less than an hour ago was: "Jev: New frontier model 40-400x cheaper and 20-200x faster" I'm going to agree that was misleading. And on the second point: >>Also "can't hallucinate" seems wrong? Sure, it can't emit an invalid type, but it can still emit a co…

> Type safety is not factual correctness. I very much agree with this and want to hone in on where do actually disagree. Would you say a linear classifier hallucinates?

A hallucination in the context of LLMs is generally understood as an incorrect answer presented as factual. If you claim that "x can't hallucinate" in the context of LLMs, you're saying that x always gives accurate answers. It does not matter whether the answer is type safe. If its value is incorrect, it's a hallucination.

Re: Introducing System One Models and Jev

#262

Post by mmastrac » Wed, Sep 16, 2026, 1:27 AM UTC

Is this a Markov/Diffusion model with some sort of external Engram memory? If so, this could be extremely interesting.

Re: Introducing System One Models and Jev

#263

Post by dthedavid » Wed, Sep 16, 2026, 1:28 AM UTC

Looks promising. I'm building an AI video editor and multi tool calls take >30s using Gemini. This would be a a game changer if Jev can take that down to single digits at p95.

Re: Introducing System One Models and Jev

#264

Post by janalsncm » Wed, Sep 16, 2026, 1:30 AM UTC

bigglebear wrote: post #227

alfalfasprout wrote: post #173

Earlier quoted context omitted.

How is this not a frontier model? It's bleeding edge in its own niche. It's not a frontier LLM; however, applicable to many of the things people use LLMs for.

It's nothing like a traditional LLM and so should not be compared to one. It's a heavily constrained, tiny model that can only produce a probability score or a yes/no answer over pre-defined selections. It has no long-context capacity. I mean, imagine comparing this thing to Astra, it's hilarious. They don't even tell you what the max input size is, and they only allow 10 possible answers to choose from for the Choic…

> It's nothing like a traditional LLM and so should not be compared to one.

Frontier LLMs are expensive jack of all trades. You can absolutely compare them to purpose-built tools on any domain they touch. Engineering is all about assessing tradeoffs.

Re: Introducing System One Models and Jev

#265

Post by Wazzymandias » Wed, Sep 16, 2026, 1:30 AM UTC

This looks and feels a lot like productionized conformal prediction

Re: Introducing System One Models and Jev

#266

Post by activehuman » Wed, Sep 16, 2026, 1:30 AM UTC

I can see the value in this but looks like there's going to be trouble in communicating the difference between this and a regular LLM, and also proving the potential cost savings in using this to replace existing systems that are using LLMs with frameworks like langgraph, as this can't be a drop in replacement and would require a significant amount of re-architecting/reengineering of systems to get the type system to work

Re: Introducing System One Models and Jev

#267

Post by strich » Wed, Sep 16, 2026, 1:32 AM UTC

Huh this looks fantastic. The Doom demo really sold for me that this could be a great tool for accelerating QA at my gamedev studio. Signed up for early access.

Re: Introducing System One Models and Jev

#268

Post by cfowles » Wed, Sep 16, 2026, 1:38 AM UTC

Wasn't really till seeing this home assistant demo they have (https://www.loom.com/share/18c4dbcf8db546dfb2d7f2ef018e78e4) that the value really clicked for me.

Seems really cool.

Re: Introducing System One Models and Jev

#269

Post by caspar » Wed, Sep 16, 2026, 1:42 AM UTC

albelfio wrote: post #2

https://x.com/completeskeptic/status/2099925682726002904?s=4... The doom demo is quite cool

I'm not sure the authors realize this is way more than "just a cool demo": if this holds up, it's going to be huge for game QA work.

Instrument your game to output properties of entities near the player and the output is the various control inputs - moment to moment gameplay gets solved. Maybe augment with a tick-by-tick controlled stepping mode if particularly twitchy - an LLM can take care of the higher level reasoning then.

Re: Introducing System One Models and Jev

#270

Post by bqsile » Wed, Sep 16, 2026, 1:56 AM UTC

kypro wrote: post #56

> Outputs > LLMS > Strings / generated text. Strings are flexible and can be anything: chat responses, code, hallucinations, refusals, or even type-safe structured values. To be used by software, responses need to be parsed + validated. There is also always some risk that the AI goes off the rails. > Jev > Type-safe structured values. Possible outputs and structure are defined in advance. The model never makes type e…

Creativity is not required for AGI, that's maybe the only thing that is not required for AGI actually.

What a sad world would you live in if you don't keep creativity for the humans.