Filters
- Alibaba Qwen releases Qwen-Audio-3.1-Realtime full-duplex voice model, per MarkTechPost
MarkTechPost reports Alibaba's Qwen team released Qwen-Audio-3.1-Realtime, a full-duplex voice model trained to reason, call tools and decide when to speak; on a tau-Voice adaptation, task success rises to 82.0% from 78.4%.
- X tests a unified subscription bundling Cursor, Grok and X perks
X is testing a new unified subscription plan: an early Android build exposes an 'X Subscription' upgrade card saying one plan will bundle Cursor, Grok and X benefits, per leaked screenshots.
- Elon Musk unifies membership perks across X, Grok and Cursor under SpaceX
Elon Musk said SpaceX's membership program is being unified, covering X, Grok and Cursor, following SpaceX's acquisition of Anysphere.
- Cline SDK 0.0.87 fixes hub startup behind system proxies on Windows
Cline SDK v0.0.87 fixes the hub failing to start reliably on Windows machines with a system proxy, because Bun's fetch sends loopback requests through HTTP(S)_PROXY with no localhost bypass, per the GitHub release notes.
- VoxelCraft: Claude 5.5 one-shots a full-parity Minecraft 1.16 survival clone
A Reddit post showcases VoxelCraft, a Minecraft 1.16 survival clone with full parity, generated in one shot by Claude 5.5.
- Anthropic tightens Sonnet 5.5 thinking-block restrictions
Anthropic added another layer of restrictions to Sonnet 5.5's thinking blocks: thinking blocks generated by the model are now more constrained in how they can be used.
- Anthropic's claude-api skill for Claude Code adds an auto-tuning workflow
Anthropic's official claude-api skill for Claude Code gained an automatic tuning workflow, per the skill's update.
- Baidu AI Cloud is teaming up with Finch, and we've got plans. 🤝
Baidu AI Cloud is teaming up with Finch, and we've got plans. 🤝 https://t.co/4bTO9heuUE
- Trump and Mike Johnson to host six tech CEOs at White House AI lunch
President Trump and House Speaker Mike Johnson will host executives from six tech companies for lunch at the White House to discuss AI risks and regulation; expected attendees include Meta CEO Mark Zuckerberg, Anthropic CEO Dario Amodei and OpenAI President Greg Brockman.
- Anthropic's IPO filing shows soaring revenue, mounting costs, and "existential" risks
Anthropic warns in its own IPO prospectus that its AI technology could pose "existential risks to humanity." Revenue grew twelvefold in 2025 to $4.6 billion, but the operating loss widened to $8.06 billion. Backers are aiming for a valuation above $2 trillion, and the IPO is expected to set the bar for the entire AI industry. The article Anthropic's IPO filing shows soaring revenue, mounting costs, and "existential" risks appeared first on The Decoder .
- Anthropic prospectus details losses, growth and existential-risk warning, per TechCrunch
TechCrunch reports Anthropic told investors in its prospectus that it loses tens of billions of dollars a year while growing rapidly, and warned its own AI might pose an existential risk to humanity; Reuters reported the IPO filing at a $2 trillion valuation after a $42 billion net loss in 2025.
- GitHub Security Lab says its open-source Taskflow Agent found 24 vulnerabilities in Android apps, including a location leak in OsmAnd
GitHub Security Lab researcher Kevin Stubbings says Android-specific audit taskflows for the lab's open-source seclab Taskflow Agent have found and reported 24 vulnerabilities in Android apps. One lets any app, even with no permissions, silently import settings into the OsmAnd navigation app, which has over 10 million downloads, and leak the user's map location and routes to an attacker's server. Two chained bugs in the Wikipedia Android app allow an account takeover through a deeplink. He says LLMs are good at finding vulnerabilities but often misjudge severity, so every finding needs review by a researcher. The taskflows are open source, need a GitHub Copilot license and use premium model requests.
OfficialSource: GitHub Blog - OpenAI details textGrain watermarking for EU AI Act, detector access starts with researchers
OpenAI published its approach to text watermarking under the EU AI Act, built on textGrain — an invisible statistical signal in the model's word choices. Starting immediately, API customers globally can opt in for select models (off by default), eligible ChatGPT and Codex output in the EU gets the watermark over the coming weeks, and detector access opens to approved researchers and expert organizations only. At a 1% target false-positive rate, detection worked on about 80% of 200-token passages and 95% of 400-token passages, but replacing 25% of words cut detection to 17%. OpenAI says textGrain matched or exceeded SynthID for text in its tests and plans to open-source the technology.
- Hugging Face's Trackio 0.39 adds agent-embeddable dashboards
Hugging Face's Trackio 0.39 adds a way for agent applications to embed training dashboards, per a post retweeted by the official account.
- Manus shares a ranked list of games built with its agent
Manus's official account shared a ranked list of games built with its agent, including Prism, Paint Grab, Helix Drop, Peak Flight and Ember Survivors.
- NVIDIA highlights Korean team's top-5 open model score on Artificial Analysis
NVIDIA's AI account highlighted how a small Korean team hit a top-5 AA II score in an open Nemotron Labs model.
- Google demos Gemini's Dial-An-Artist and Chrono Look experiments
Google's official X account promoted a series of Gemini experiments, including Dial-An-Artist for talking with history's iconic creators, Chrono Look for trying on history, and visual-first conversational search.
OfficialSource: x-Google - Replit Agent returns charts, pipelines and automations from a question
Replit's official account demonstrated its agent returning a chart, a data pipeline or a working automation from a single question.
- Anthropic's Mike Krieger: Haiku 5.5 will round out the family in the coming weeks
Anthropic CPO Mike Krieger said on X that Haiku 5.5 will round out the Claude 5.5 family in the coming weeks.
- Steven Pinker calls for sober AI safety engineering over doomsday rhetoric, per The Decoder
The Decoder reports Harvard psychologist Steven Pinker thinks fears of AI-driven extinction are overblown and turned down a public debate with blogger Scott Alexander, calling such events a spectator sport.