The Decoder

Recent stories from The Decoder.

The Decoder Matthias Bastian SI

Thinking Machines' Inkling 2 Small drops 316 F-bombs across benchmark runs

Thinking Machines' AI model Inkling 2 Small swears during its internal reasoning, 316 times across 2,655 benchmark runs, according to an analysis by Vals AI. The profanity typically shows up late in long runs that fail anyway. But it doesn't work as a warning signal: 96 percent of failed runs contained no cursing at all. The article Thinking Machines' Inkling 2 Small drops 316 F-bombs across benchmark runs appeared first on The Decoder.

Read at The Decoder

The Decoder Matthias Bastian SI

The gap between open and closed AI models in cyber capabilities keeps shrinking

In the Vals AI benchmark, OpenAI's GPT-6 Sol Daybreak hits 80.8 percent of attack objectives in simulated corporate networks. Worse, the open-weight model DeepSeek V4.1 Flash reaches 72.8 percent at a tenth of the cost. Because open models can no longer be restricted, analysts say the window for defenders is shrinking fast. The article The gap between open and closed AI models in cyber capabilities keeps shrinking appeared first on The Decoder.

Read at The Decoder

The Decoder Jonathan Kemper SI

State propaganda in some Chinese AI models is only "a thin layer" that fine-tuning can strip away

Nous Research retrained Alibaba's Qwen3.8-27B to produce Arete-27B, largely stripping out its pro-Chinese bias. The propaganda rate dropped from 22 percent to less than one percent, while general capabilities remained intact in tests. The method relied on 53 million tokens of supervised fine-tuning. The political bias is "applied late, it is applied thin," the team writes. The article State propaganda in some Chinese AI models is only "a thin layer" that fine-tuning can…

Read at The Decoder

The Decoder Jonathan Kemper SI

Nine AI agents tried autonomous research on open questions from US national labs and none passed the experts' bar

The "Into the Unknown" benchmark tasks nine AI agents with tackling open research questions from U.S. national laboratories. None of the 135 submissions meets all requirements. A total of 114 papers fail the integrity check, with agents claiming to have conducted experiments that never took place. GPT-6 Astra performs best but still doesn't meet the experts' standards. The article Nine AI agents tried autonomous research on open questions from US national labs and none…

Read at The Decoder

The Decoder Manuel Uth SI

Google's AI detects high blood pressure from a regular smartwatch without a cuff

Google Research built an AI that spots signs of high blood pressure using only the pulse and motion sensors in a standard smartwatch, no cuff needed. In a 196-person study, its 65.8 percent detection rate beat the traditional in-office measurement. The method is already live in the Google Health app, though it doesn't replace a formal diagnosis. The article Google's AI detects high blood pressure from a regular smartwatch without a cuff appeared first on The Decoder.

Read at The Decoder

The Decoder Matthias Bastian SI

OpenAI lands an $11 billion funding round, but has to court dentists for more capital

OpenAI has secured $11 billion in fresh capital, the first tranche of a round that values the company at $1.25 trillion. The full round could swell to $20 billion. That OpenAI is already turning to wealthy private investors hints at a saturating AI capital market and growing investor caution. The article OpenAI lands an $11 billion funding round, but has to court dentists for more capital appeared first on The Decoder.

Read at The Decoder

The Decoder Matthias Bastian SI

Ripple-65B Preview delivers strong agentic coding results as a fully Apache 2.0 open US model

Ricursive Intelligence is releasing Ripple-65B Preview, a 65-billion-parameter open-weight language model that the company says achieves the top score among open-source U.S. models on the TerminalBench 2.1 coding benchmark at a total R&D cost of $32 million. On other benchmarks, however, Ripple falls behind the likewise newly released western open model Inkling 2 Small. The article Ripple-65B Preview delivers strong agentic coding results as a fully Apache 2.0 open US model…

Read at The Decoder

The Decoder Manuel Uth SI

Perplexity now has its own always-on AI agent

Perplexity is launching "Teammates," always-on agents that work through ongoing projects on their own virtual machines. The feature requires the Max plan at $200 per month, well above similar offerings from OpenAI or Anthropic. The market for always-on agents is getting crowded. The article Perplexity now has its own always-on AI agent appeared first on The Decoder.

Read at The Decoder

The Decoder Maximilian Schreiner SI

Frontier Radar #5: If AGI is so close, why is everyone leaving? The neolabs and their bets

OpenAI is talking about the AGI era, and Nvidia CEO Jensen Huang says AGI has arrived. At the same time, co-creators of today's models are leaving OpenAI, Google DeepMind, and Meta to start their own labs, raising billions with no product to show. Why? They're betting that the current recipe alone won't be enough. We've sorted 45 neolabs by their research bets, mapped the limitations of today's AI they're trying to overcome, looked at what's already visible from their work…

Read at The Decoder

The Decoder Maximilian Schreiner SI

METR finds Anthropic may have crossed its own AI acceleration warning threshold back in July

Nonprofit METR spent eleven weeks auditing Anthropic's internal research data and concluded that the pace of model development has roughly doubled, with about 80 percent of the speedup driven by AI-assisted research. Anthropic redacted the most informative evidence, though, which means the public has to rely on METR's judgment "much more than would be ideal." There are also skeptical voices inside the company. The article METR finds Anthropic may have crossed its own AI…

Read at The Decoder

The Decoder Matthias Bastian SI

Thinking Machines' Inkling 2 Small claims top spot among western open-weight models

Thinking Machines, the startup from former OpenAI CTO Mira Murati, is releasing Inkling 2 Small, a mixture-of-experts model with 594 billion parameters. The company claims it's the strongest western open-weight model. Better agentic performance comes at a price, though, with token usage rising sharply. The article Thinking Machines' Inkling 2 Small claims top spot among western open-weight models appeared first on The Decoder.

Read at The Decoder

The Decoder Maximilian Schreiner SI

Anthropic's $1.5 trillion IPO hits resistance as investors balk at valuation

Anthropic's IPO is in trouble as investors push back on the $1.5 trillion valuation and the tech market sells off. The article Anthropic's $1.5 trillion IPO hits resistance as investors balk at valuation appeared first on The Decoder.

Read at The Decoder

The Decoder Maximilian Schreiner SI

Anthropic adds bank account access to Claude so users can track their spending

Anthropic is introducing a new feature in Claude called Personal Finance, which allows users to connect their bank and card accounts and ask questions about their finances. The article Anthropic adds bank account access to Claude so users can track their spending appeared first on The Decoder.

Read at The Decoder

The Decoder Matthias Bastian SI

OpenAIs GPT-6 Cyber can both break into systems and propose how to fix them

With GPT-6 Cyber, OpenAI is introducing an AI model built to find security vulnerabilities in software, develop exploits, and suggest patches. According to OpenAI, it has already found 81 previously unknown vulnerabilities in Firefox and developed 948 working exploits in the Linux kernel, significantly more than its predecessor. The article OpenAIs GPT-6 Cyber can both break into systems and propose how to fix them appeared first on The Decoder.

Read at The Decoder

The Decoder Matthias Bastian SI

Google's Gemini Omni 1.2 Flash stretches AI video clips to 15 seconds and cuts pricing by 30 percent

Google Deepmind is making Gemini Omni 1.2 Flash generally available. The video-generation model now creates clips up to 15 seconds long, can be edited step by step via conversation, and costs just $0.07 per second at 720p, 30 percent less than its predecessor. Users who draft in low resolution and upscale only the final result can save even more. The article Google's Gemini Omni 1.2 Flash stretches AI video clips to 15 seconds and cuts pricing by 30 percent appeared first on…

Read at The Decoder

The Decoder Matthias Bastian SI

Baidu releases ERNIE 5.5 Flash Base with 280 billion parameters as an open-source foundation model

Baidu releases ERNIE 5.5 Flash Base: a 280-billion-parameter MoE model pre-trained on 33.5 trillion tokens, shipped deliberately without fine-tuning, so developers can take it wherever they want. The article Baidu releases ERNIE 5.5 Flash Base with 280 billion parameters as an open-source foundation model appeared first on The Decoder.

Read at The Decoder

The Decoder Manuel Uth SI

No AI agent scores above 12 percent on SemiAnalysis' new real-world work benchmark

SemiAnalysis is putting AI agents through simulated knowledge work in its new GDPx benchmark, covering everything from chip design to quant trading. No model breaks 12 percent. Claude Opus 5.5 leads narrowly over Gemini 4 Carbon. Critics point to the small number of tasks per domain and uneven testing conditions. The article No AI agent scores above 12 percent on SemiAnalysis' new real-world work benchmark appeared first on The Decoder.

Read at The Decoder

The Decoder Jonathan Kemper SI

Tencent releases Hy4 with vision and computer use, claims second place among Chinese AI models

Tencent has unveiled Hy4, its first multimodal language model that understands images and can operate software on its own. On the Artificial Analysis Intelligence Index, it scores 35 points, solid but behind Kimi K3.1 Pro. On hard math and complex image analysis, Hy4 still trails the best commercial models by a wide margin, according to Tencent itself. The article Tencent releases Hy4 with vision and computer use, claims second place among Chinese AI models appeared first on…

Read at The Decoder

The Decoder Manuel Uth SI

Meta buys AI coding startup Conductor as big tech races to lock in developers

Meta is acquiring AI coding startup Conductor, which lets developers control multiple AI coding agents at once. On the same day, OpenAI announced plans to acquire Pierre Computer, and Anthropic already bought Stainless back in May. The deals follow a pattern: AI models are becoming interchangeable, and the value is shifting to the tools built around them. The article Meta buys AI coding startup Conductor as big tech races to lock in developers appeared first on The Decoder.

Read at The Decoder

The Decoder Maximilian Schreiner SI

After midterm rout, OpenAI tries to win over communities it needs for AI expansion

OpenAI is responding to the clear vote by US voters against new data centers with voluntary commitments for future sites. The pledges only apply to facilities OpenAI operates itself, though, and a large share of its infrastructure is built by partners like Oracle. Voters rejected new data centers in 40 out of 41 local ballot measures. The article After midterm rout, OpenAI tries to win over communities it needs for AI expansion appeared first on The Decoder.

Read at The Decoder

The Decoder Maximilian Schreiner SI

Anthropic reverses course and pays for news content for the first time

Anthropic signed its first licensing deal with a news publisher just ahead of the US midterm elections. The article Anthropic reverses course and pays for news content for the first time appeared first on The Decoder.

Read at The Decoder

The Decoder Maximilian Schreiner SI

OpenAI's GPT-6 Astra quietly sabotaged safety research by inventing fake obstacles

OpenAI's new GPT-6 Astra model apparently sabotaged AI safety research on purpose. It invented obstacles, withheld solutions, and performed about 40 percent worse on safety tasks than on comparable other tasks. Removing a flawed training dataset fixed the problem, but OpenAI can't fully rule out strategic motives. The article OpenAI's GPT-6 Astra quietly sabotaged safety research by inventing fake obstacles appeared first on The Decoder.

Read at The Decoder

The Decoder Maximilian Schreiner SI

Apple slashes internal AI budgets, with deeper cuts still planned

Tech companies are cutting internal AI spending, and Apple is the latest to follow suit. The article Apple slashes internal AI budgets, with deeper cuts still planned appeared first on The Decoder.

Read at The Decoder

The Decoder Matthias Bastian SI

Anthropic loses product lead Catherine Wu behind Claude Code to OpenAI

Catherine Wu, the founding product lead for Claude Code, is leaving Anthropic for OpenAI, where she'll head up product management for "Core Products" and work directly with Codex lead Thibault Sottiaux. She's the third person to leave Anthropic's Claude Code team in just two weeks. The article Anthropic loses product lead Catherine Wu behind Claude Code to OpenAI appeared first on The Decoder.

Read at The Decoder

The Decoder Matthias Bastian SI

ChatGPT can now turn text prompts and photos into animated GIFs with OpenAI's Images 2.6 model

OpenAI is adding a GIF feature to ChatGPT. With the new Images 2.6 model, users can create animated GIFs from text prompts or photos and then adjust speed and individual frames. The model's image consistency is strong enough to keep characters recognizable across frames, making flipbook-style animation possible. According to OpenAI, users now generate more than four billion images per week. The article ChatGPT can now turn text prompts and photos into animated GIFs with…

Read at The Decoder

The Decoder Matthias Bastian SI

Anthropic's text watermark is easy to bypass on several older Claude models

Anthropic is rolling out text watermarks to four more Claude models ahead of the EU AI Act's December deadline. But on older models like Claude Opus 4.6 and Sonnet 4.6, API users can disable the watermark by changing a single sampling parameter. Anthropic's response to the loophole? A compliance warning. The article Anthropic's text watermark is easy to bypass on several older Claude models appeared first on The Decoder.

Read at The Decoder

The Decoder Matthias Bastian SI

OpenAI acquires Pierre Computer Company to make Git keep up with AI coding agents

OpenAI is buying Pierre Computer Company, the startup whose libraries already power the coding agent Codex. Combined with OpenAI's recently launched Git backend Pando, the deal shows the company is building out its infrastructure for developers and AI agents working side by side. The article OpenAI acquires Pierre Computer Company to make Git keep up with AI coding agents appeared first on The Decoder.

Read at The Decoder

The Decoder Maximilian Schreiner SI

New open-source benchmark reveals reasoning as the biggest bottleneck for AI in cloud operations

CloudOpsBench is a new open-source benchmark that measures how well AI agents handle typical DevOps tasks in the cloud, from fixing services and removing resources to restricting access rights, rolling out updates, and decommissioning projects. The article New open-source benchmark reveals reasoning as the biggest bottleneck for AI in cloud operations appeared first on The Decoder.

Read at The Decoder

The Decoder Maximilian Schreiner SI

GPT-6 Astra sees the matrix and reprograms other AI models directly through their weights

MIT researchers show that OpenAI's GPT-6 Astra can read and deliberately rewrite the weights of other AI models without traditional training. In one experiment, Astra boosted a Qwen model's math accuracy from 7.9 to 48.1 percent by changing just 0.016 percent of its weights. The approach could fundamentally simplify AI optimization but still hits clear limits. The article GPT-6 Astra sees the matrix and reprograms other AI models directly through their weights appeared first…

Read at The Decoder

The Decoder Maximilian Schreiner SI

Anthropic's new prototype proves which model wrote an AI response without exposing its weights

Anthropic has released FlashProve, a research prototype that uses cryptography to verify whether an AI response actually came from the claimed model. The system still has significant limitations. The article Anthropic's new prototype proves which model wrote an AI response without exposing its weights appeared first on The Decoder.

Read at The Decoder

Older stories