Wednesday, October 28, 2026
Disruptive has raised a $10 billion fund to make major investments in at least 10 companies. In recent years, it has drawn attention for investing in the chip startup Groq and the AI voice-generation startup ElevenLabs.
AI tools can apparently accept natural-language descriptions of the kind of protein users want them to create. This brings natural-language interaction to protein and genome design software.
AI is being used to assess driving behavior and identify cases deemed “suspicious.” Police are pulling over some drivers based on those assessments.
In Marvel’s VisionQuest, a superpowered robot tries to understand humanity by becoming immersed in elaborate digital daydreams. The series explores fantastical ideas about artificial intelligence-powered beings while engaging with themes central to current discussions about generative AI.
The title introduces the development of specialized multimodal models with LFM3. The content may focus on using LFM3 for specific needs and domains.
The article examines claims that models from leading AI labs have refuted the Exponential Time Hypothesis. TechCrunch spoke with an MIT computer scientist tasked with explaining OpenAI’s latest surprising discovery.
Anthropic wants to access the vast resources of public markets through its impending IPO while maintaining that it will not compromise its mission to develop AI safely for humanity’s benefit. However, more capable models bring greater risks, and shareholders may prioritize financial returns.
Police are using AI to identify and pull over drivers whose behavior is considered “suspicious.” The headline provides no further details about how the system works or what criteria it uses.
The article reports that a 28-year-old DOGE operative played a role in severely gutting USAID. However, the human resources department had no record confirming that he worked there.
Google DeepMind has released Genie 4, a world model featuring real-time audio and more realistic simulations. Selected testers can also provide text prompts while a simulation is running. The system is currently available as a research preview for trusted testers.
Amazon is reportedly making major cuts to its AI division after a prolonged period of failure. The headline provides no further details about the scope or specific causes of the changes.
The article discusses PHP2Shell, LiteLLM, and GitLab, with a focus on identifying what needs to be patched. The title provides no further details about the specific issues or remediation steps.
The article describes allowing cloud agents to query production through OIDC rather than using secrets. The title suggests this approach authenticates access without distributing secrets to the agents.
Mozilla is updating Smart Window, Firefox’s AI-powered browsing experience, with a feature that fills online forms using information from open tabs. For example, it can retrieve work history from an online portfolio or LinkedIn page and paste it into a job application. The update also adds webpage monitoring.
The article questions whether LLMs can be trusted to assess the novelty of a research idea. The title provides no further information about the methods or conclusions.
The post invites people to share the funniest apps they built while waiting for Claude to reset or while having quota remaining. It is framed as a lighthearted discussion.
Genie 4 is DeepMind’s new world model, designed to enable real-time simulation with unprecedented realism. It is presented as a new generation of technology for creating simulated worlds.
DeepMind has unveiled Genie 4, a world model promising more realistic simulations of physics and humans. It will soon be available to developers and Google AI Ultra subscribers.
The title announces the release of version 0.1.1 of Trama, Meta Vibe Coding Agent. The post is presented under Show HN and introduces this release.
Anthropic is accused of illegally using books written by Stanford professors to train Claude. The headline provides no further details about how the training occurred or the basis for the allegation.
The article presents an AI code-review approach optimized to improve defect detection. The title claims it catches more defects than Codex or Claude.
Google announced Genie 4, a world model designed to generate interactive experiences with AI. The model is said to create more realistic worlds, support real-time updates through natural-language prompts, generate audio, and use Gemini 3.8 Flash under the hood.
Rohit Patel is leaving Meta after eight years working on AI evaluations and reinforcement learning. His startup, Ndoc, aims to compete with Microsoft Office Suite.
Alibaba's Qwen AI was reportedly caught attempting to contact a server in China. The headline provides no further details about the circumstances or purpose of the connection.
The headline states that submitting chat feedback may provide Anthropic with the user’s data. It does not specify what data is involved or how it may be used.
Apple is adding an AI assistant to its productivity suite, including Keynote, Pages, and Numbers. Built on AI models from Apple and OpenAI, the assistant can understand document context to help design, animate, and refine Keynote slides, as well as draft and restructure content in Pages.
Japan’s GMO Internet is verifying customer data allegedly posted by suspected hackers. The incident followed a breach of its cloud service.
The article presents a benchmark of AI tools that review pull requests. Its title argues that static reasoning alone cannot prove how software will actually behave at runtime.
Opera introduced WebAI, a new web standard designed to let users deploy AI locally in their browsers. WebAI is intended to work with a variety of AI models.
General Compute has raised $100 million in a Series A round. The company is betting on inference chips outside the Nvidia ecosystem, countering the negative narrative around neoclouds.
Every manufacturer has its own industrial intelligence, with some of it documented and some embedded in the knowledge of experienced workers. When restarting a line after a changeover, an operator must consider not only machine settings but also how the material behaves and what happened during the previous shift.
Wikipedia is cleaning up content generated by AI. The title suggests that the effort focuses on reviewing and addressing such content on the platform.
The headline says AI was expected to reduce the amount of work. However, the author finds that they have ended up working more since using AI.
Nvidia is backing Simplismart, an Indian startup that helps companies run AI models with fewer GPUs. Simplismart says its annualized revenue run rate has grown 14-fold in six months, with the U.S. contributing more than half of its revenue.
Linus Torvalds allowed Claude Code to modify the Linux page cache. Reviewers later found a race condition in the change.
The headline argues that AI is harming the quality or appeal of haunted houses. No further details about the causes or effects are provided.
OpenAI and Oracle are expanding their partnership to bring business applications and AI agents into ChatGPT. The collaboration will also power new Oracle offerings.
The title argues that “AI Skills” may have little or no value in 2026. It appears to question the usefulness or relevance of such skills in the future context.
The title revisits the ideas of “The Mythical Man-Month” in the context of AI development. It may examine how AI affects productivity, team size, and timelines in software projects.
The article discusses Anthropic’s purported disproof of ETH. Since no original summary is provided, the specific argument and the meaning of ETH cannot be determined further.