7 articles
Everything the newsroom has published on artificial intelligence, from the latest model release to the use cases changing how people work.
Featured
A cryptic leak has reignited rumors of RSI at Google DeepMind. What we know, what remains unclear, and how self-improvement could reshape the race against OpenAI and Anthropic.
Rohan Nayak says Pocket FM has reached roughly $500M in ARR, an annualized revenue run rate, by combining highly industrialized video acquisition, serialized audio fiction, and AI production. Replies confirm an LTV/CAC ratio of about 3x, its ElevenLabs partnership, and the launch of Pocket Saga, while raising questions about AI content quality, the durability of its CTR advantage, and figures added by @VadimStrizheus that do not appear in Rohan’s post.
@kimmonismus presents DeepSeek V4.1 Flash as a rapid leap beyond July’s release, featuring a causal encoder-decoder architecture, a 552-billion-parameter MoE, and a far smaller KV cache. @deepseek_ai confirms the key figures, the phaseout of V4 Pro, API pricing, cache costs, and native multimodal support. While readers praise local inference and the release pace, mixed Terminal Bench results challenge its frontier-model status. Self-hosting is feasible for well-equipped deployments, though the official announcement also mentions 2,000-GPU installations with storage.
An AI-designed drug reduced the estimated biological age of pulmonary fibrosis patients by 3 to 4 years. Here’s what that figure leaves out.
In seven posts, @hilbertspaess announces his resignation from Anthropic and calls the race toward self-improving superintelligence irresponsible. He says AI leaders privately fear an existential risk, while critics question how AI could destroy humanity, whether slowing down against China is feasible, and what the public, researchers, and governments should do.
OpenAI launches ChatGPT Images 2.5 with faster generation, greater fidelity, consistent details across edits, and targeted comment-based editing. The thread also introduces Sketch for drawing directly in ChatGPT and explores its impact on creative workflows and the questions left unanswered.
OpenAI says 10,000 agents solved Navier-Stokes in 88 hours using an internal model billed as more capable than GPT-6 Astra. Here’s what it means and why credit is already disputed.