跳到主要內容
AI News HubLIVE
公開文章 158採集文章 169可信度 84刷新頻率 720 分鐘
健康狀態 健康來源類型 研究原文權限 站內改寫最近入庫 2026-09-29ID latent-space運行狀態 已啟用

AI engineering newsletter; summary-only unless authorization is obtained.

最新公開文章

待翻譯:Claude Code’s Next Era — Thariq Shihipar, Anthropic

AI 服務暫時不可用,以下為來源摘要,待恢復後補全翻譯:Shipping Opus/Sonnet 5.5, Mods, Plugins, Projects, Tag while Pacing the Frontier

Latent Space站內正文待翻譯:Claude Code’s Next Era — Thariq Shihipar, Anthropic

待翻譯:[AINews] The Future of Latent Space

AI 服務暫時不可用,以下為來源摘要,待恢復後補全翻譯:A quiet day lets us discuss the work behind the scenes - now open for business!

Latent Space站內正文待翻譯:[AINews] The Future of Latent Space

待翻譯:Runway’s WorldPrompt and the Engineering of Real-Time Worlds

AI 服務暫時不可用,以下為來源摘要,待恢復後補全翻譯:GWM Worlds 2 uses persistent context and timed actions to steer a world model generating video and audio in real time.

Latent Space站內正文待翻譯:Runway’s WorldPrompt and the Engineering of Real-Time Worlds

待翻譯:Foundries vs Navigators: Lowering the Cost of Science

AI 服務暫時不可用,以下為來源摘要,待恢復後補全翻譯:Guest Post: In science, thinking has gotten cheap but doing has not. This asymmetry is reshaping how research companies operate, largely inconspicuously.

Latent Space站內正文待翻譯:Foundries vs Navigators: Lowering the Cost of Science

待翻譯:🔬Bio-security is an AI Arms Race - Eric Nguyen (CEO, Radical Numerics)

AI 服務暫時不可用,以下為來源摘要,待恢復後補全翻譯:Radical Numerics is using biological chain-of-thought and multimodal perception to keep up with the bio-defense arms race, design new genomes and gain insights into biology itself.

Latent Space站內正文待翻譯:🔬Bio-security is an AI Arms Race - Eric Nguyen (CEO, Radical Numerics)

待翻譯:🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science

AI 服務暫時不可用,以下為來源摘要,待恢復後補全翻譯:We talked to Google’s Oscar winning “Giganerd” about automating science, solving climate change, and how future generations can contribute to science in the age of superintelligent AI

Latent Space站內正文待翻譯:🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science

待翻譯:[AINews] Here are 6 Clones of Jev in 2 days

AI 服務暫時不可用,以下為來源摘要,待恢復後補全翻譯:Imitation is the sincerest form of Flattery

Latent Space站內正文待翻譯:[AINews] Here are 6 Clones of Jev in 2 days

待翻譯:[AINews] not much happened today

AI 服務暫時不可用,以下為來源摘要,待恢復後補全翻譯:a quiet day

Latent Space站內正文待翻譯:[AINews] not much happened today

待翻譯:Can Skills Learned in Games Transfer to Real-World Work?

AI 服務暫時不可用,以下為來源摘要,待恢復後補全翻譯:Good Start Labs trained an AI on a railroad game — and one version improved at financial research. The difference was the training design.

Latent Space站內正文待翻譯:Can Skills Learned in Games Transfer to Real-World Work?

待翻譯:Humanity’s Last Invention — Richard Socher of Recursive

AI 服務暫時不可用,以下為來源摘要,待恢復後補全翻譯:Richard Socher is an NLP OG and CEO of You.com, who has now spun out an even more ambitious startup focused on RSI — already worth $5B!

Latent Space站內正文待翻譯:Humanity’s Last Invention — Richard Socher of Recursive

待翻譯:The Rise of the Forward Deployed Engineer — and How To Do the Job Right

AI 服務暫時不可用,以下為來源摘要,待恢復後補全翻譯:Before co-founding Kepler, Vinoo Ganesh led Spark at Palantir and built Project Frontline — a pioneering program for Forward Deployed Engineers. He takes us through the best practices of FDEs.

Latent Space站內正文待翻譯:The Rise of the Forward Deployed Engineer — and How To Do the Job Right

待翻譯:[AINews] not much happened today

AI 服務暫時不可用,以下為來源摘要,待恢復後補全翻譯:a quiet day

Latent Space站內正文待翻譯:[AINews] not much happened today

待翻譯:[AINews] OpenAI reports Navier-Stokes singularity find in 88 hours using Astra-next, roughly 10,000 agents and 130B tokens (>$40M), a contender for second ever Millennium Prize awarded

AI 服務暫時不可用,以下為來源摘要,待恢復後補全翻譯:Overshadowing Cognition's $48B Series E, Mistral's $24B Series D, Meta's Muse agent, and GPT Image 2.5. The most jam packed, feel the AGI day in the history of AI.

Latent Space站內正文待翻譯:[AINews] OpenAI reports Navier-Stokes singularity find in 88 hours using Astra-next, roughly 10,000 agents and 130B tokens (>$40M), a contender for second ever Millennium Prize awarded

前沿 AEO 追蹤器:Astra 會選什麼(以及其他前沿模型,以及你能做什麼)

Latent Space 首次發佈 Astra 項目“前沿 AEO 追蹤器”,用 6 種提示變體在 161 個類別中測試 7 個前沿模型,量化 AI 產品推薦。結果顯示模型存在系統偏見、也有 GPT 推薦 Claude 等“無偏”反例,並發現 Opus→Fable、Sol→Astra 兩代之間的選擇反轉和搜索行為差異。

Latent Space站內正文前沿 AEO 追蹤器:Astra 會選什麼(以及其他前沿模型,以及你能做什麼)

[AINews] GPT-6 Astra:OpenAI史上最大規模的LLM發佈

OpenAI正式發佈新一代旗艦模型GPT-6 Astra,宣稱為“最智能且最對齊的模型”,其計算機使用、軟件工程和數學推理能力被標榜為重大突破。不過,第三方評測顯示其能力提升不均衡,而且每token價格更高、可監測性下降,引發了關於評測飽和、部署治理和監管難度的廣泛爭論。

Latent Space站內正文[AINews] GPT-6 Astra:OpenAI史上最大規模的LLM發佈

GPT-6 Astra:不到6美元一小時即可僱到的自動化AI工程師

OpenAI今日發佈GPT-6 Astra,這是Stargate計劃中的首款輕量循環超級模型,多項基準大幅超越Fable 5.1。作者在消耗超過200億token進行實測後發現,Astra已成為真正自主的AI工程師:能幫你選擇和訓練模型、標註數據、部署調試系統、管理大量子代理,並能閲讀日誌、處理超長上下文。按實測速度與定價折算,有效運行成本可低於每小時6美元。

Latent Space站內正文GPT-6 Astra:不到6美元一小時即可僱到的自動化AI工程師

不歡迎PR:頂級AI開源項目如何管理數千名貢獻者

GitHub的拉取請求(PR)傳統上開放,但頂級AI開源項目如Vercel的AI SDK、Astro、Flue和tldraw正轉向“軟件工廠”模式,用內部代理管理PR和問題,減少對外部貢獻的依賴,以提高效率並應對AI生成的垃圾PR。

Latent Space站內正文不歡迎PR:頂級AI開源項目如何管理數千名貢獻者

全部來源