Article
CommunityJev Guides & Articles
MAD Podcast: Diogo Almeida on AI that's 95% Einstein, 5% Mr. Bean
On The MAD Podcast, Matt Turck talks with TypeSafe CEO Diogo Almeida about why intelligence per dollar matters more than enterprise adoption, and why RLVR benchmark gains still leave models 95% Einstein, 5% Mr. Bean.
- Format
- Article
- Reading time
- 2 min
- Published by
- Community
- Author
- Matt Turck
- Last verified
- Oct 2026
Highlights
- Matt Turck hosts TypeSafe CEO Diogo Almeida, who says intelligence per dollar matters more than enterprise adoption.
- Turck's intro cites the launch numbers: 40 million video views, a million users a day, a trillion tokens a day, a third of the Fortune 500.
- Almeida says intelligence per dollar has dropped two orders of magnitude and expects that to make the Jevolution unstoppable.
- The summary says decomposing email automation into measurable classifiers and code avoids the fragile errors of giant prompts.
- Per the summary, RLVR benchmark gains still produce models that are 95% Einstein, 5% Mr. Bean rather than reliable automation.
Watch out
Copy reviewed against the published episode page and its transcript excerpt on 2026-10-17; the excerpt is truncated, so this entry covers the episode's opening discussion. Launch statistics are TypeSafe's own claims, as cited by the host.
Questions about this entry
Short answers drawn from the linked project or page.
Diogo Almeida, CEO of TypeSafe, speaking with host Matt Turck in an episode dated 10/15/2026 that runs 1 hour 8 minutes across 27 chapters.
Turck's introduction cites a launch video with 40 million views, about a million users a day, a trillion tokens a day, and a third of the Fortune 500, all in the few weeks since launch.
Intelligence per dollar. He says the adoption stats do not really matter for what he calls the Jevolution, and that the future is primarily about intelligence per dollar, which he says has dropped two orders of magnitude.
It is the episode summary's shorthand for the reliability gap: RLVR benchmark gains still produce models that are mostly brilliant but not dependable automation.
By decomposing email automation into measurable classifiers and code instead of relying on giant prompts, which the summary says avoids fragile errors.
Turck raises the two-year stealth period; Almeida says he ignored the noise in the field, did not buy the story that AI was solving all sorts of problems, and saw a large technological gap no one was addressing.
The entry is grounded on the episode page and its truncated transcript excerpt; the launch statistics are TypeSafe's claims cited by the host, and Almeida's arguments about cost and reliability are his own.
Reactions & coverage
Posts, threads, and videos about this entry from around the web.
Hacker News: Simon Willison's overview of Noul, Choice, and Score, and where a decision model fits.
benwerd53 pointsJev introduces a new shape of LLMRead the thread on Hacker NewsHacker News: swyx posts the Latent Space interview with Diogo Almeida.
swyx4 pointsJev: System One Models for Prod, Not God – With Diogo Almeida, CEO, TypeSafe AIRead the thread on Hacker NewsX: Diogo Almeida in five minutes
X: LLMs vs. Jev, clearly explained
More Jev resources like this
TWIML episode 779 features TypeSafe co-founder and CEO Diogo Almeida on Jev, the company's model for bringing fast, reliable intelligence directly into software, and why text-generating models are poorly suited to many real-world automation decisions.
Jev Guides & ArticlesArticle#community#press#podcast
Community
ThursdAI's September 17 episode covers TypeSafe's Jev launch with founding DevRel Allie Laabs: choice, score and null primitives, $42-per-billion input pricing, and a live X timeline classifier running at 12 tweets a second.
Jev Guides & ArticlesArticle#community#press#podcast
Community
An episode of The Merge (by CodeRabbit) recorded at the first Jev hackathon 11 days after launch; TypeSafe DevRel Allie Laabs explains how Jev replaces prompt-parse cycles with typed choice, score and yes/no outputs.
Jev Guides & ArticlesArticle#community#press#podcast
Community
John Lindquist joins Claire Vo to walk through eight Jev use cases, from a real-time voice to-do app to data deduplication with confidence scores, on the episode billed as Jev week at How I AI.
Jev Guides & ArticlesArticle#community#press#podcast
Community
An interactive site by Sarah Drasner that explains System One thinking — fast, automatic judgments versus deliberate checking — and what that distinction means for software design, closing with the Jev × WebMCP Chrome extension as a case study.
Jev Guides & Articles#community#github#javascript
Community
sdras/system-one-explainerJavaScript
sdras/system-one-explainerJavaScriptRogue Security's post uses Jev and System One models to frame agentic security as a decision problem: the same send_file action should be allowed, blocked or reviewed depending on who requested it and where the data goes, illustrated with policy case files rather than product results.
Jev Guides & ArticlesArticle#community#press#security
Community
From the Jev guides
Original write-ups on what this entry is for.
From the Jev AI model community
Posts from builders shipping with Jev right now.