4,338 stories
TL;DR Your support agent demo runs on one 2,000-word system prompt. Three months later, five teams have copied it, every copy has drifted, and nobody...
Duvo today announced a $32 million Series A led by Balderton Capital, with participation from existing investors Index Ventures, Northzone, and Credo Ventures. The round brings Duvo's total funding to...
Uvi, a startup that feeds company knowledge to AI agents, has emerged from stealth with $7.5 million in seed funding. CT Innovations and Cardumen led the funding round, with Mana VC, 91 Ventures, and Gaingles VC also participating.
The complete history of Alibaba's Qwen models, from the 2023 Tongyi Qianwen beta to Qwen 3, Qwen 3.7 Max, and the Qwen 3.8 open weights. Updated October 2026.
The complete history of DeepSeek, from Liang Wenfeng's High-Flyer quant fund to R1, the January 2025 market shock, V4 Pro, and V4.1 Flash. Updated October 2026.
Datalab released OmniParseBench, an open OCR benchmark with 16,288 yes/no tests on 2,937 pages in 92 languages. Every score is a pass rate you can slice by tag, script or language.
Fish Audio released Drama 3, a 35B mixture-of-experts text-to-speech model with about 3B active parameters. It takes plain-language direction, covers 83 languages and ships open weights under a revenue-capped community license.
Coworker AI's enterprise AI security checklist: test permission-aware access, retention, SSO, audit logs, agent approvals and MCP risk before you buy.
Marktechpost ranks the 9 best AI gateways for 2026. We compare Bifrost, Portkey, LiteLLM, Kong, TrueFoundry, Cloudflare, Vercel, OpenRouter and Orq.ai on latency, pricing, licensing, governance and deployment.
Liquid AI released LFM2.5-24B-A2B, an open-weight 24B Mixture-of-Experts reasoning model with 2B active parameters. It targets local agents, scoring 48.09 on LiveClaw and decoding 184 tokens/s on an Apple M5 Max.
Google has made Gemini 4 Argon generally available on the Gemini API, AI Studio and the Gemini app. The sparse MoE model writes up to 1M output tokens and leads the Vals Index at 68.90%. The post Google Releases Gemini 4 Argon API With 1M Token Output Limit appeared first on MarkTechPost.
TL;DR: Increasingly when you interact with frontier models, your prompt doesn't only contain the words on the page, but also a lot of what you've ever publicly written online.This is not intended to be a scary post.
Tsinghua researchers released SRHarness, an open-source runtime for agentic symbolic regression, and MechBench, a benchmark that tests whether AI can recover the mechanisms behind formulas.
By Gabriel Unger and Arvind Narayanan. Unger is an economist at Stanford’s Digital Economy Lab.
How we approach working with AI companies on projects that the public might rely on to understand risks.
The key questions about internal AI development the public needs more evidence about, and a concrete proposal for how developers could work with METR to produce and share that evidence.
Our proposal is to analyze a developer’s internal incidents and whether training may suppress evidence of worse misalignment.
Our proposal is to analyze whether a developer’s internal agents could get away with serious harm without being caught.
Showing 324 of 4,338 stories
Load more storiesGet the top AI news and tools straight to your inbox every week.
Weekly · AI news
Join 15,000+ AI enthusiasts
