Analyst warns OpenAI could quickly replicate TypeSafe's Jev classifier
A technical blog argues OpenAI is positioned to fast-follow Jev's classification approach, sparking debate over whether the startup has real defensibility.
What to know
- TypeSafe's Jev achieved record adoption speed but now faces analysis suggesting OpenAI can quickly replicate its classification approach.
- The core debate: whether Jev's speed and ease-of-use represent a defensible innovation or whether it rediscovers pre-transformer classifier ideas in a new package.
- Commenters dispute whether OpenAI's stated strategy (reasoning-focused RL training) even aligns with copying a fast, non-reasoning classifier.
- The conversation reveals uncertainty about whether Jev is a genuine paradigm shift or successful marketing of a narrower technical niche.
The dispute Whether Jev represents a defensible innovation that OpenAI will fast-follow, or a re-packaging of well-known classifier techniques whose success is primarily a product/marketing phenomenon unrelated to technical novelty. · positions read across 32 posts and comments
OpenAI will likely replicate or acquire Jev's approach because classifiers embedded in larger models offer real efficiency and cost gains.
-
“I'd be surprised if they weren't aqui-hired by one of the big labs as a unicorn. The ability to use classifiers under the hood for the larger models has the potential to dramatically improve cost and throughput.”
60secs · Hacker News ↗
OpenAI's stated focus on reasoning-based RL models makes fast, non-reasoning classifiers strategically misaligned with their roadmap.
-
“OpenAI is all in training models with reasoning with RL, and Jev-like models are the total opposite. They are made to not reason at all to be fast. If you want to add reasoning on top, you might as well use a conventional LLM.”
andy12_ · Hacker News ↗
Jev's success is hype; classifiers are not new, and the real innovation is in interface design and execution, not the technical moat.
-
“Everyone was obsessed with classification prior to transformers, then we had 5+ years of everyone (rightfully) obsessed with next-token prediction. What's this sudden resurgence of interest in classifiers?”
halyconWays · Hacker News ↗
Jev offers practical benefits over existing embeddings and classification tools, regardless of whether it's technically novel.
-
“I've been classifying a lot of email and while OpenAI `text-embedding-3-small` has been very helpful for fast and cheap embeddings, initial tests with Jev are very promising and much more ergonomic.”
nzoschke · Hacker News ↗
John Berryman Arcturus Labs analystTypeSafe Creator of Jev classifier modelOpenAI Potential fast-follower
How it unfolded 2 developments, newest first · click a bar or a number to jump articlespostscomments
-
2
Commenters question the broader framing and whether this is hype or substance
Several voices expressed skepticism about the entire premise—some dismissing the article as LLM-written and outdated, others questioning whether Jev represents rediscovery of pre-transformer classifier ideas, and one suggesting the viral success is marketing rather than genuine innovation. The conversation shifted from OpenAI risk to whether classifiers are actually the future or a step backward.
“Everyone was obsessed with classification prior to transformers, then we had 5+ years of everyone (rightfully) obsessed with next-token prediction. What's this sudden resurgence of interest in classifiers?”
— halyconWays -
People are desperately trying to cope themselves into thinking that there are alternatives to scaling up transformers to AGI/actual competition to OpenAI or Anthropic. Jev, continual learning, linear attention, local models, non-transformer architectures etc. Imo these are just random technologies that nerdsnipe your average twitter or hackernews…
2 more of the top 3 · 26 posts in this stretch
-
The next step is to rediscover Eliza :)I did so a month ago, I developed a deterministic agent framework that works with a set of predefined intents, it is instantaneous and fully deterministic.It works thanks to FlintParser (https://github.com/gioblu/NPC-Forge/blob/main/src/FlintParse...) which can transpile plain English to any Programming…
-
People in my lab (sklearn people) developped something that I feel close to jev but focused on tabular data : https://tabicl.readthedocs.io/en/latest/This is a transformer based classifier with massive pretraining on synthetic datasets and it outperforms boosting classifiers on many benchmarks without the need of more gradient descent steps (the…
-
-
1
Commenters push back on the fast-follow thesis with competing technical arguments
Multiple Hacker News commenters disputed Berryman's premise. Some argued OpenAI's focus on reasoning-based RL training makes Jev-style fast classifiers incompatible with their strategy; others contended that every major AI shop already has internal classifiers and the real question is interface design, not moat dynamics.
“OpenAI is all in training models with reasoning with RL, and Jev-like models are the total opposite. They are made to not reason at all to be fast.”
— andy12_ -
first by HN Best, 1d ago
-
I find it unlikely. OpenAI is all in training models with reasoning with RL, and Jev-like models are the total opposite. They are made to not reason at all to be fast. If you want to add reasoning on top, you might as well use a conventional LLM because you lose the price and speed benefits when you output auto-regressive tokens. I don't think…
2 more of the top 3 · 6 posts in this stretch
-
This article is extraordinarily hard to read. It’s tummelvisioned on OpenAI and things like tool calling which are only relevant to the extent that llms have been tuned to make relative choices, but this applies to all LLMs. Also, some really outdated references. LLM written, perhaps?Also, moat discussion is the lowest form of discussion. I don’t…
-
comments are pretty weird here, there's no real moat to what jev is doing, it is certain that frontier labs are going to release their own jev and there are even open source alternatives (although nowhere near as accurate as jev).so maybe typesafe's real plan is to front run and releasing their own new models for some time until they can get…
-
-
background
Arcturus Labs analyst warns OpenAI is well-positioned to fast-follow Jev — John Berryman published a technical analysis arguing that TypeSafe's Jev, though rapidly adopted, lacks a defensible moat against OpenAI replication. He argues OpenAI has years of implicit classification experience through tool-calling and could quickly replicate Jev's training approach, then embed it into their own models for efficiency gains.
What people are saying 17 voices from 1 site · best of 32 · verbatim
- Will OpenAI actually build a Jev-like product, or is their strategic bet so heavily on reasoning models that classifiers fall outside their roadmap?
- What technical advantage does Jev have over traditional ML classifiers or OpenAI's existing embedding models?
- Is Jev's rapid adoption genuine product-market fit or successful viral marketing in a hype cycle?
- Yesterday
-
All discussions around Jev and its moat and whether it is new or not all focus on the architecture part, but what Diogo always states is that he thinks their secret sauce is in the data and the post training with RLCD … that is also why I think these clones using RLHF trained LLMs completely miss the point.
- Sep 22
-
This. It’s machine learning vs. “AI” for the uninitiated. Soon there will be a new ground breaking model that does k-means clustering and will get a billon dollar funding (but only if you're young and live in SF)The good news is it’s fun to see people discover and get excited about things that I like as well.
-
There are already open-weight versions of Jev that are equal or better performance, and can run on regular hardware. It took people 24 hours to create them. OAI could create Jev in a day if they wanted to.
-
Entertainingly, OpenAI had a general purpose zero-shot classifier API built on GPT-3! Just no one ever cared that much about it, so I guess it got dropped somewhere along the way since 2020/2021.
-
"I don't see much substance to this buzz..."Agreed. This isn't new. I led a research team at a Fortune 500 that used a transformer based classifier approach in a commercial product as far back as 2022 and we didn't come up with it. It was already common enough that we found the inspiration for our implementation on some web forum. Models like…
-
Correct me if I'm wrong, but a zero-shot classifier like Jev is fundamentally different to a classifier with a fixed task (e.g. for safeguards), unless they trained a general purpose system to complete the safeguard task, which seems unlikely.
-
> Every major AI shop has a ton of in-house classifiers already, big, small, generalistI think building generalist classifier is some open ended research task, where frontier labs can contribute: different internal reasoning, instruction tuning, building datasets and benchmarks, building and distilling super large models.
-
Why would OAI need to follow Jev? I really think this is paid by Jev. Jev itself won't have lunch money in a shortwhile because there are literally 10s of free alternatives available which can be run locally on basic consumer hardware. Terrible utility aside, there is no sensible business proposition in Jev.
-
If your "master AI" is good enough, it should be able to find and learn about and use specialized tech AI like Jev if it suits your goalsand then whatever tech it is will be absorbed/assimilated/Sherlocked into the leading products anyway
-
I'd be surprised if they weren't aqui-hired by one of the big labs as a unicorn.The ability to use classifiers under the hood for the larger models has the potential to dramatically improve cost and throughput, allowing them to increase margin on a very similar service.
-
Yeah, if your "master AI" is good enough, it should be able to find and learn about and use specialized tech AI like Jev if it suits your goals, before you even know about such stuff.
-
I fed into the hype at first. Testing Jev and Laya, they both suffer from the same issues as LLMs that stop them being useful beyond limited classifications.I can't see any benefits that a typical ML classifier would not be better at.
-
Isn't this more and more likely on all shapes of model evolution? The providers will all copy each other.And in this case I hope its true. I've been classifying a lot of email and while OpenAI `text-embedding-3-small` has been very helpful for fast and cheap embeddings, initial tests with Jev are very promising and much more ergonomic.I put more…
-
Every single AI shop has a ton of in-house classifiers already. Some are used in inference pipelines (e.g. safeguards), some are used in data preparation, training, various one-off and intermediate tasks etc. Offering them on a public API doesn't always make business sense. I don't see much substance to this buzz, looks like people that are new to…
-
Rather than focusing on OpenAI in particular, let's just say that there are many smart people at other AI labs and if it seems like it will be popular, this technique will probably be copied. What would prevent them from adding another API?Hopefully there will be some decent benchmarks and gateways for switching between providers easily.
-
If OpenAI releases something similar to what Jev does, then that would be like admitting defeat. Their whole spin is AGI and world ending danger. Why would somebody with an AGI at home make something like Jev which is intended to be a part of some SW the AGI is going to replace anyway.
-
I'm not exactly following through with the claim, can someone explain how the built-in classification would not necessitate more tokens used, or be much different from turning on reasoning? Not that I don't see the difference, I just doing see how OpenAI would do it well.