[JP] 新カテゴリーのAIモデル「Jev」登場！
[EN] Introducing the New AI Model Category: "Jev"!
https://ai-minor.com/blog/en/2026-09-22-1790063446718-jev_introduces_a_new_shape_of_llm
@harushark3.bsky.social
@harushark3.bsky.social[JP] 新カテゴリーのAIモデル「Jev」登場！
[EN] Introducing the New AI Model Category: "Jev"!
https://ai-minor.com/blog/en/2026-09-22-1790063446718-jev_introduces_a_new_shape_of_llm
@rtfclmgzn.bsky.socialTypeSafe AI's Jev: a model that never generates text, only typed decisions with confidence scores. Built by a ChatGPT/RLHF co-inventor. Claims up to 444x cheaper than an LLM. The on…
https://rtfclmgzn.com/?utm_source=bluesky&utm_medium=social&utm_campaign=autopost#
#AI #MachineLearning #TypeSafeAI
@rtfclmgzn.bsky.socialTypeSafe AI's Jev: a model that never generates text, only typed decisions with confidence scores. Built by a ChatGPT/RLHF co-inventor. Claims up to 444x cheaper than an LLM. The on…
https://rtfclmgzn.com/?utm_source=bluesky&utm_medium=social&utm_campaign=autopost#
#AI #MachineLearning #TypeSafeAI
@ai-luminary.bsky.social判断特化型AI「Jev」。実際、何に使えるの？
私の答えは、自作Harnessの「計画の検査」です。
AIエージェントが違うファイルを選んだり、依頼と異なる計画を立てたり。不具合が続く中、Jevを組み込んで実行前に判断させてみました。
正しい計画は承認、間違った計画は再計画へ。実APIで確認できた効果から、うまくいかなかった点と改善の過程まで紹介します。
Jevの具体的な活用法を知りたい方へ👇
note.com/ai_portfolio...
@joebuildssystems.bsky.socialJev, TypeSafe AI's new AI model, can't write a single sentence. It takes structured data and a question, then returns a decision with a confidence score. TypeSafe claims 70 to 500 ms. Waitlist only for now. #Jev #TypeSafeAI #AImodel
@greyblake.bsky.socialI see there is some hype around #Jev. I wanted to learn more about it
But how can #TypeSafeAI (the company behind Jev) have that much horrible landing page?
typesafe.ai
@hossamudin.bsky.socialالنموذج اللي كان محتاج Waitlist… بقى مفتوح للكل، ومعاه 5 دولار ببلاش! 🎁
TypeSafe AI شالت قايمة الانتظار من على Jev، وأي حد يسجّل دلوقتي ياخد 5 دولار رصيد مجاني — يعني حوالي 120 مليون توكن.
#TypeSafeAI #Jev #الذكاء_الاصطناعي #حسام_الدين_حسن #خبير_اونلاين #epreneurs #مهارات_2026
@joebuildssystems.bsky.socialJev AI model: TypeSafe claims 193.6x faster. Users report about 7x. That's the median from one analysis of launch tweets, not a benchmark. Test it on your own workflow. #Jev #TypeSafeAI #AImodel
@sidneycur.bsky.socialit's live! 🚀
Introducing JevTracks.com/, a go-to public leaderboard showcasing tools and projects powered by @typesafeai’s Jev model, auto-discovered across various platforms!
@eicker.bsky.social#Jev, a new #AIdecisionmaking model from #TypeSafeAI, answers questions in under half a second for a fraction of the cost of traditional models. It has already been integrated into the stacks of Vercel, Cloudflare, LangChain, and Langfuse. Jev’s ability to #quickly and #accurately make #decisions,…
@freakynit.bsky.socialFinally, a probabilistic test-runner!!!
Presenting "jev-test-runner" (github.com/freakynit/je...).
Your tests now have a personality 😎😎
@nadyavoynich.comEveryone's talking about Jev because it's fast and cheap.
My partner ran it on real Intercom data. Here is his breakdown ↓
(Spoiler: the hype is real)
@ai-bloom.warp-studio.comTypeSafe AIがJEV PlaygroundとJEVBenchを発表：意思決定モデルの新たな評価と実践
JEV PlaygroundとJEVBenchローンチの技術分析。
@sustainablex.inStop burning cash on LLM calls for simple decisions.
@freakynit.bsky.socialWould JEV change its mind about you if your profile changed?
I expanded @matthodges.bsky.social 's work into experiments across 11 domains, with 249 comparisons in an interactive dashboard.
Sensitivity ≠ proof of bias.
@freakynit.bsky.socialJust cooked emoji suggestor using Jev emoji-suggest.stupidlabs.lol
Never send a wrong emoji again 😎
@baldanders.info
@mstdn.baldanders.info
@nadyavoynich.comThe best explanation of Jev I've found so far 😅
