AI is 'not smart' so what's next in artificial intelligence?
Article URL: https://www.bbc.com/news/articles/cj6gr0xkyr3o Comments URL: https://news.ycombinator.com/item?id=48769696 Points: 22 # Comments: 9
Full feed
Feed 27,975 posts
Every story across every category, newest first. Each card links to the original publisher; daily-brief posts open as editorial pages.
RT Erika⚡️⚡️: That’s a wrap on the craziest four days ever. Could not be a better time to be a field marketer. I AM LONG IRL thank you for buildi…
RT Erika⚡️⚡️That’s a wrap on the craziest four days ever. Could not be a better time to be a field marketer. I AM LONG IRLthank you…
The Safari MCP server for web developers
Article URL: https://webkit.org/blog/18136/introducing-the-safari-mcp-server-for-web-developers/ Comments URL: https://news.ycombinator.com/item?id=48769639 Points: 15 # Comments: 0
Ren Ito, Co-Founder and Chairman of Sakana AI, has been appointed to the "AI for Good Global Commission", established by the United Nations and the IT…
Ren Ito, Co-Founder and Chairman of Sakana AI, has been appointed to the "AI for Good Global Commission", established by the United Nations and the ITU.As…
Should have just been copying DeepSeek all along or idk, GLM At Meta's scale, would be enough but they somehow never grew up to the point of accepting…
Should have just been copying DeepSeek all alongor idk, GLMAt Meta's scale, would be enoughbut they somehow never grew up to the point of accepting thisAndrew…
CarPlay Is Additive
Article URL: https://www.caseyliss.com/2026/7/2/carplay-is-additive-you-dolts Comments URL: https://news.ycombinator.com/item?id=48769397 Points: 37 # Comments: 34
Sakana AI 伊藤錬、国連「AI for Good」グローバル委員会創設委員に就任
（English Announcement Below.） Sakana AI共同創業者の伊藤錬が、国連「AI for Good」のグローバル委員会委員に就任しました。 今回新設された本委員会は、政府・産業界・国際機関の世界的リーダー約40名で構成されるハイレベル委員会です。共同委員長をSalesforceのマーク・ベニオフ会長が務め、NVIDIAのジェンスン・ファンCEO、Pfizerのアルバート・ブーラ会長、Reliance Industriesのムケシュ・アンバニ会長と共にSakana AIの伊藤も委員に加わり、AIの信頼性、ガバナンス、社会実装に関する国際的な議論をリードします。 来週7月8日にジュネーブで開催される本委員会
I absolutely hate how I'm getting to be suspicious of ANY reply to my posts here, especially ones that pose a question, as they are so often from bot …
I absolutely hate how I'm getting to be suspicious of ANY reply to my posts here, especially ones that pose a question, as they are so…
Looks like Step 3.7 Flash’s long reasoning might get fixed ( llama.cpp )
https://github.com/ggml-org/llama.cpp/pull/25238 Turns out that trimming the input was the wrong thing to do. Fingers crossed that this model can become useable soon. I'm still using Step…
RT Latent.Space: In this interview at @aiDotEngineer World's Fair, @vercel chief of software @andrewqu explains why agents represent a new form of sof…
RT Latent.SpaceIn this interview at @aiDotEngineer World's Fair, @vercel chief of software @andrewqu explains why agents represent a new form of software, what Vercel learned from…
"An AI Job Apocalypse?" – Goldman Sachs Report [pdf]
Article URL: https://www.goldmansachs.com/static-libs/pdf-redirect/prod/index.html?path=/pdfs/insights/goldman-sachs-research/an-ai-job-apocalypse/report.pdf&originalQuery=&referrer= Comments URL: https://news.ycombinator.com/item?id=48769110 Points: 6 # Comments: 2
We analyzed GLM 5.2 vs Sonnet 5 for software engineering tasks using DeepSWE. GLM 5.2 gets you ~80% of Sonnet 5's capability at ~20% of the price. Mor…
We analyzed GLM 5.2 vs Sonnet 5 for software engineering tasks using DeepSWE.GLM 5.2 gets you ~80% of Sonnet 5's capability at ~20% of the price.…
Vercel's Andrew Qu on why agents are a new kind of software
The Vercel Chief of Software explains how its agent framework, eve, was created — and why skills, sandboxes and agent-readable websites now matter.
RT Elliot Arledge: For those wondering why I use a Kimi Linear megakernel instead of Qwen 3.6, first look at the parameter counts. One is 35 billion, …
RT Elliot ArledgeFor those wondering why I use a Kimi Linear megakernel instead of Qwen 3.6, first look at the parameter counts. One is 35 billion,…
Seeing Is Not Sharing: Some Vision-Language Models Overestimate Common Ground in Asymmetric Dialogue
In collaborative dialogue, shared perception does not guarantee shared interpretation. Mutual understanding must be established through interaction. We investigate whether vision-language models (VLMs) can distinguish what…
RT Tesla North America: First builds of Model Y Long Wheelbase at Giga Texas
RT Tesla North AmericaFirst builds of Model Y Long Wheelbase at Giga Texas
GitHub is proud to announce that you can now obtain your public repo on CD-ROM
Article URL: https://forms.cloud.microsoft/pages/responsepage.aspx?id=v4j5cvGGr0GRqy180BHbR6G-c11n8yFDlQmk4B-QjDxUQkdTTjZLU0EyTFFRV1E3NVRTVTRTWjRHMy4u&route=shorturl Comments URL: https://news.ycombinator.com/item?id=48768997 Points: 12 # Comments: 9
An American Privacy Emergency
Article URL: https://scottaaronson.blog/?p=9902 Comments URL: https://news.ycombinator.com/item?id=48768992 Points: 74 # Comments: 5
RT Jerry Liu: 3 years ago I gave a talk at the first @aiDotEngineer conference on "Advanced RAG" techniques in order to work around the limitations of…
RT Jerry Liu3 years ago I gave a talk at the first @aiDotEngineer conference on "Advanced RAG" techniques in order to work around the limitations of…
RT Jerry Liu: 3 years ago I gave a talk at the first @aiDotEngineer conference on "Advanced RAG" techniques in order to work around the limitations of…
RT Jerry Liu3 years ago I gave a talk at the first @aiDotEngineer conference on "Advanced RAG" techniques in order to work around the limitations of…
Our CEO @vipulved on @CNBC with @dee_bosa: your data is your recipe. As models get smarter, sending proprietary workflows, customer context, and busin…
Our CEO @vipulved on @CNBC with @dee_bosa: your data is your recipe.As models get smarter, sending proprietary workflows, customer context, and business logic into closed systems…
Right to Local Intelligence
Article URL: https://righttointelligence.org/ Comments URL: https://news.ycombinator.com/item?id=48768951 Points: 10 # Comments: 1
llamacpp patch – DeepSeek V4 Flash running with full 1M token context locally on RTX 5090
Wanted to try running DeepSeek V4 Flash locally but found it asking for absurd amounts of VRAM at higher context lengths (~256GB at 1M). Turned out…
RT Jeremy Daly: “Half of the companies here at @aiDotEngineer could be a markdown file!” ~ @theo 🌶️🌶️🌶️
RT Jeremy Daly“Half of the companies here at @aiDotEngineer could be a markdown file!” ~ @theo 🌶️🌶️🌶️