<!-- mobian-agent-page publisher="dailydev" canonical="https://daily.dev/posts/1XMQ9zlpA" -->

---
title: An ex-OpenAI researcher just deleted language from the...
description: A satirical, tongue-in-cheek video introduces 'Jev', a supposed new 'system one' AI model from ex-OpenAI researcher Diego Almeida's startup Typesafe AI (which...
canonical: https://daily.dev/posts/an-ex-openai-researcher-just-deleted-language-from-the-llm--1xmq9zlpa
twitter:card: summary_large_image
twitter:site: @dailydotdev
og:type: website
og:site_name: daily.dev
og:title: An ex-OpenAI researcher just deleted language from the LLM... | daily.dev
og:description: A satirical, tongue-in-cheek video introduces 'Jev', a supposed new 'system one' AI model from ex-OpenAI researcher Diego Almeida's startup Typesafe AI (which...
og:url: https://daily.dev/posts/an-ex-openai-researcher-just-deleted-language-from-the-llm--1xmq9zlpa
og:image: https://api.daily.dev/og/posts/1XMQ9zlpA.png
og:image:alt: An ex-OpenAI researcher just deleted language from the LLM...
og:image:width: 1200
og:image:height: 630
og:locale: en
---

> ## Documentation Index
> Fetch the complete documentation index at: https://daily.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# An ex-OpenAI researcher just deleted language from the LLM...

**[Fireship](https://daily.dev/sources/fireship)** · 5 min read · 68 upvotes · 9 comments

## Summary

A satirical, tongue-in-cheek video introduces 'Jev', a supposed new 'system one' AI model from ex-OpenAI researcher Diego Almeida's startup Typesafe AI (which raised $40 million). Jev is pitched as a fast, cheap, type-safe classifier that returns structured outputs (choice, score, or yes/no) instead of free-form text, with a 'calibrated confidence' score instead of typical LLM overconfidence. The host humorously questions its legitimacy, noting the architecture is undisclosed, critics say it's just a rebranded zero-shot classifier, and an open-source clone (OpenJV) already replicates its behavior using a frozen Qwen 4B model without new training. The video ends with a sponsored segment for Mux (video API).

## Full article

daily.dev links to this article rather than hosting it. Read it at the original source: <https://www.youtube.com/watch?v=TbkUKCm3CHQ>

## Questions this post answers

### What is Jev, the AI model from Typesafe AI that people are talking about?

Jev is a 'system one' AI model built by Typesafe AI, a startup founded by ex-OpenAI researcher Diego Almeida that raised $40 million. Unlike typical LLMs, it does not generate free text; instead it returns strongly typed outputs in one of three shapes (choice, score, or yes/no) along with a calibrated confidence value, and is claimed to be roughly 200 times faster and 400 times cheaper than standard large language models.

_Developers evaluating fast, cheap classification models can follow the Jev debate and similar tool comparisons on daily.dev._

### Is Jev just a rebranded zero-shot classifier?

Critics argue yes, saying Jev is functionally similar to zero-shot classifiers that have existed for over a decade, pioneered by researchers like Jiny Yang. One developer claims Jev matches a paper they published a year earlier, and another built an open-source clone called OpenJV that reproduces Jev's interface by reading option probabilities off a frozen Qwen 4B model in a single forward pass, requiring no new training and running on a 3090 GPU.

_Anyone weighing hyped AI startups against established techniques can track independent reproductions like this on daily.dev._

## Community discussion

Top comments from developers on daily.dev.

**@byteoutlaw** · 6 upvotes

> Another week in AI. 🙃

**@gibrancastillo** · 1 upvotes

> ![GIF](https://static.klipy.com/ii/935d7ab9d8c6202580a668421940ec81/c2/4a/3ROhIYTQ.gif)

**@starfallcodes** · 1 upvotes

> You are absolutely right! Jeb is awesome!

**@ristotoldsep** · 0 upvotes

> A classifier in a blazer is still a classifier. This was fun.

**@anmolbaranwal** · 0 upvotes

> amaze amaze amaze. statement

## Similar posts on daily.dev

- [What is Jev? How System One models could change AI agents : fireup.pro](https://daily.dev/posts/what-is-jev-how-system-one-models-could-change-ai-agents-fireup-pro-eqcqub8on) · FireUp · 1 upvotes · 0 comments

---

Tags: [#machine-learning](https://daily.dev/tags/machine-learning), [#llm](https://daily.dev/tags/llm), [#ai-agents](https://daily.dev/tags/ai-agents)

[View this post on daily.dev](https://daily.dev/posts/an-ex-openai-researcher-just-deleted-language-from-the-llm--1xmq9zlpa)

```json
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]}
{"@context":"https://schema.org","@type":"TechArticle","headline":"An ex-OpenAI researcher just deleted language from the LLM...","url":"https://daily.dev/posts/an-ex-openai-researcher-just-deleted-language-from-the-llm--1xmq9zlpa","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/an-ex-openai-researcher-just-deleted-language-from-the-llm--1xmq9zlpa"},"datePublished":"2026-09-21T17:55:39.265Z","dateModified":"2026-09-21T19:03:11.640Z","description":"A satirical, tongue-in-cheek video introduces 'Jev', a supposed new 'system one' AI model from ex-OpenAI researcher Diego Almeida's startup Typesafe AI (which...","image":"https://i.ytimg.com/vi/TbkUKCm3CHQ/sddefault.jpg","thumbnailUrl":"https://i.ytimg.com/vi/TbkUKCm3CHQ/sddefault.jpg","isAccessibleForFree":true,"articleSection":"Fireship","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Fireship","logo":"https://media.daily.dev/image/upload/s--75ndqrkr--/f_auto,t_logo/v1702882094/logos/fireship.jpg","url":"https://daily.dev/sources/fireship"},"commentCount":9,"discussionUrl":"https://daily.dev/posts/an-ex-openai-researcher-just-deleted-language-from-the-llm--1xmq9zlpa","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":68},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":9}],"keywords":"machine-learning,llm,ai-agents","timeRequired":"PT5M","video":{"@type":"VideoObject","name":"An ex-OpenAI researcher just deleted language from the LLM...","description":"A satirical, tongue-in-cheek video introduces 'Jev', a supposed new 'system one' AI model from ex-OpenAI researcher Diego Almeida's startup Typesafe AI (which...","thumbnailUrl":"https://i.ytimg.com/vi/TbkUKCm3CHQ/sddefault.jpg","uploadDate":"2026-09-21T17:55:39.265Z","duration":"PT5M","url":"https://api.daily.dev/r/1XMQ9zlpA","embedUrl":"https://www.youtube.com/embed/TbkUKCm3CHQ"}}
{"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Fireship","item":"https://daily.dev/sources/fireship"},{"@type":"ListItem","position":3,"name":"An ex-OpenAI researcher just deleted language from the LLM..."}]}
{"@context":"https://schema.org","@type":"WebPage","@id":"https://daily.dev/posts/an-ex-openai-researcher-just-deleted-language-from-the-llm--1xmq9zlpa","comment":[{"@type":"Comment","text":"Another week in AI. 🙃","datePublished":"2026-09-21T18:22:48.608Z","url":"https://daily.dev/posts/1XMQ9zlpA#c-9CjgDTtcp","author":{"@type":"Person","name":"Jacob B. Bonde","url":"https://daily.dev/byteoutlaw","image":"https://media.daily.dev/image/upload/s--veTChHK7--/f_auto/v1733220070/avatars/avatar_5cQvIZKr5tDFDotVDjulg"},"interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":6}},{"@type":"Comment","text":"","datePublished":"2026-09-22T01:53:59.852Z","url":"https://daily.dev/posts/1XMQ9zlpA#c-AKAGRWRfT","author":{"@type":"Person","name":"Gibran Castillo","url":"https://daily.dev/gibrancastillo","image":"https://lh3.googleusercontent.com/a/ACg8ocLPBhMkOmbFL6Q-M2KXWBhF9iqnoRHBqiiWZjlEdLAMHywsYGyoFA=s96-c"},"interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1}},{"@type":"Comment","text":"You are absolutely right! Jeb is awesome!","datePublished":"2026-09-22T08:59:36.580Z","url":"https://daily.dev/posts/1XMQ9zlpA#c-w2tpVY0wU","author":{"@type":"Person","name":"starfall","url":"https://daily.dev/starfallcodes","image":"https://media.daily.dev/image/upload/s--4xoTqFbk--/f_auto/v1786806197/avatars/avatar_ecPDW57nUVRidLxwsxruz?_a=BAMAMicg0"},"interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1}},{"@type":"Comment","text":"A classifier in a blazer is still a classifier. This was fun.","datePublished":"2026-09-25T08:13:38.213Z","url":"https://daily.dev/posts/1XMQ9zlpA#c-3WCjmxW3o","author":{"@type":"Person","name":"Risto Tõldsep","url":"https://daily.dev/ristotoldsep","image":"https://media.daily.dev/image/upload/s--i6RRoxB7--/f_auto,pg_1/v1794383659/avatars/avatar_ryDlbifZ1EuRuh3g51ryC?_a=BAMAMiiw0"}},{"@type":"Comment","text":"amaze amaze amaze. statement","datePublished":"2026-09-30T16:39:28.644Z","url":"https://daily.dev/posts/1XMQ9zlpA#c-dLbZEgdyR","author":{"@type":"Person","name":"Anmol Baranwal","url":"https://daily.dev/anmolbaranwal","image":"https://avatars.githubusercontent.com/u/74038190?v=4"}}]}
{"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/an-ex-openai-researcher-just-deleted-language-from-the-llm--1xmq9zlpa#faq","mainEntity":[{"@type":"Question","name":"What is Jev, the AI model from Typesafe AI that people are talking about?","acceptedAnswer":{"@type":"Answer","text":"Jev is a 'system one' AI model built by Typesafe AI, a startup founded by ex-OpenAI researcher Diego Almeida that raised $40 million. Unlike typical LLMs, it does not generate free text; instead it returns strongly typed outputs in one of three shapes (choice, score, or yes/no) along with a calibrated confidence value, and is claimed to be roughly 200 times faster and 400 times cheaper than standard large language models. Developers evaluating fast, cheap classification models can follow the Jev debate and similar tool comparisons on daily.dev."}},{"@type":"Question","name":"Is Jev just a rebranded zero-shot classifier?","acceptedAnswer":{"@type":"Answer","text":"Critics argue yes, saying Jev is functionally similar to zero-shot classifiers that have existed for over a decade, pioneered by researchers like Jiny Yang. One developer claims Jev matches a paper they published a year earlier, and another built an open-source clone called OpenJV that reproduces Jev's interface by reading option probabilities off a frozen Qwen 4B model in a single forward pass, requiring no new training and running on a 3090 GPU. Anyone weighing hyped AI startups against established techniques can track independent reproductions like this on daily.dev."}}]}
```

