@the-decoder.com.feed.birb@rss-parrot.net
I'm an automated parrot! I relay a website's RSS feed to the Fediverse. Every time a new post appears in the feed, I toot about it. Follow me to get all new posts in your Mastodon timeline! Brought to you by the RSS Parrot.
---
Artificial Intelligence is changing the world. THE DECODER brings you all the news about AI.
Site URL: the-decoder.com/feed/@birb
Feed URL: the-decoder.com/feed
Posts: 69
Followers: 2
AI spam websites flood the web with false information, and the number is growing fast
Published: March 14, 2026 14:20
Newsguard and AI detector Pangram Labs have launched a real-time system to spot so-called "AI content farms." More than 3,000 of these sites have already been flagged, with hundreds more popping up every month, and the number keeps growing. The article AI…
China pushes OpenClaw "one-person companies" with millions in AI agent subsidies
Published: March 14, 2026 12:19
At least seven Chinese local governments have launched million-dollar funding programs for OpenClaw projects within days. The goal: "one-person companies" where a single founder runs a business with AI agents as employees. The article China pushes OpenClaw…
Hume AI open-sources TADA, a speech model five times faster than rivals with zero hallucinated words
Published: March 14, 2026 11:12
Hume AI has open-sourced TADA under the MIT license, a fast speech generation model that processes text and audio in sync and produced zero hallucinations in testing. The article Hume AI open-sources TADA, a speech model five times faster than rivals with…
Ai2 releases new robotics models trained entirely in simulation to skip real-world data collection
Published: March 14, 2026 10:25
Robots trained entirely in virtual worlds working in the real world without any real training data. Ai2 says it can be done. The article Ai2 releases new robotics models trained entirely in simulation to skip real-world data collection appeared first on…
Meta reportedly plans to cut up to 20 percent of its workforce as $600 billion AI bet drives need to offset costs
Published: March 14, 2026 09:55
Meta is reportedly planning massive layoffs to finance its expensive AI investments. The article Meta reportedly plans to cut up to 20 percent of its workforce as $600 billion AI bet drives need to offset costs appeared first on The Decoder.
New Top 100 AI list shows a maturing market where ChatGPT leads but users are shopping around
https://the-decoder.com/new-top-100-ai-list-shows-a-market-where-chatgpt-leads-but-no-one-is-loyal/
Published: March 14, 2026 09:22
Andreessen Horowitz's latest Top 100 ranking of AI consumer products shows a market in flux: ChatGPT still dominates, competitors are growing fast, and global usage is fracturing along geopolitical lines. The article New Top 100 AI list shows a maturing…
Ex-Anthropic researchers launch AI startup Mirendil to tackle scientific research
Published: March 14, 2026 08:47
Former Anthropic researchers launched Mirendil, an AI startup focused on scientific research in biology and materials science. The company is reportedly seeking a $1 billion valuation. The article Ex-Anthropic researchers launch AI startup Mirendil to…
Anthropic drops the surcharge for million-token context windows, making Opus 4.6 and Sonnet 4.6 far cheaper
Published: March 13, 2026 19:50
Anthropic removes the surcharge for long contexts in Claude Opus 4.6 and Sonnet 4.6: requests with more than 200,000 tokens no longer cost up to twice as much. The article Anthropic drops the surcharge for million-token context windows, making Opus 4.6 and…
Elon Musk admits xAI "was not built right first time around," launches full restructuring
Published: March 13, 2026 19:15
Elon Musk's AI company xAI is going through a major restructuring. Musk acknowledged on X that the company "was not built right first time around" and is now being rebuilt from the foundations up. The article Elon Musk admits xAI "was not built right first…
Google explains the differences between its three Nano Banana image generation models
Published: March 13, 2026 17:42
A new guide from Google breaks down the three Nano Banana image models and when to use each one. The cheaper Nano Banana 2 reportedly delivers 95 percent of Pro's capabilities and can search the web for reference images on its own before generating output.…
AI chips are pushing everything else off TSMC's most advanced production lines
Published: March 13, 2026 13:31
By 2027, 86 percent of TSMC's N3 capacity could go to AI accelerators, according to SemiAnalysis. Smartphones are becoming a buffer for overflow demand. The article AI chips are pushing everything else off TSMC's most advanced production lines appeared…
Ukraine opens its battlefield data to allies to train AI models for autonomous drones
https://the-decoder.com/ukraine-provides-allies-with-a-platform-with-combat-data-for-ai-training/
Published: March 13, 2026 13:24
Ukraine opens its battlefield data to allies to train AI models for autonomous drones. The article Ukraine opens its battlefield data to allies to train AI models for autonomous drones appeared first on The Decoder.
Bytedance secures access to Nvidia Blackwell cluster in Malaysia, circumventing US export ban on China
Published: March 13, 2026 13:20
TikTok's parent company plans to use around 36,000 Nvidia Blackwell chips in Malaysia, according to a Wall Street Journal report. US export controls prevent direct access in China. Even the Trump administration's recent relaxations explicitly exclude these…
Perplexity's "Personal Computer" promises a tireless AI agent for $200 a month
https://the-decoder.com/perplexitys-personal-computer-promises-a-tireless-ai-agent-for-200-a-month/
Published: March 13, 2026 12:34
Perplexity AI's "Personal Computer" is an AI assistant that works around the clock - handling emails, presentations, and app control. The article Perplexity's "Personal Computer" promises a tireless AI agent for $200 a month appeared first on The Decoder.
Meta delays its next AI model Avocado after internal tests show it can't keep up with Google and OpenAI
Published: March 13, 2026 09:06
Meta is apparently postponing its new AI model "Avocado" because it is falling behind Google, OpenAI and Anthropic in tests. The article Meta delays its next AI model Avocado after internal tests show it can't keep up with Google and OpenAI appeared first…
Grok 4.20 trails Gemini and GPT-5.4 by a wide margin but sets a new record for not hallucinating
Published: March 12, 2026 19:25
xAI's Grok 4.20 is cheap, fast, and hallucinates less than any other tested model, but it can't keep up with the top tier in benchmarks. The article Grok 4.20 trails Gemini and GPT-5.4 by a wide margin but sets a new record for not hallucinating appeared…
US War Department CTO says Anthropic's AI models "pollute" the supply chain with built-in ethics
Published: March 12, 2026 18:52
The US Department of War wants to ban Anthropic's Claude from its supply chain because the AI models are too ethical. The approach echoes China's political control of AI. The article US War Department CTO says Anthropic's AI models "pollute" the supply…
Copilot Health marks Microsoft's entry into the AI health race alongside OpenAI and Anthropic
Published: March 12, 2026 17:31
Microsoft is launching Copilot Health, an AI health assistant that pulls data from wearables, medical records, and lab results to deliver personalized health advice. Long term, the company says it's working toward "medical superintelligence." The article…
Claude can now create interactive charts and visualizations directly in chat
Published: March 12, 2026 17:12
Anthropic has launched a new beta feature for its AI chatbot Claude: the ability to generate interactive diagrams, charts, and visualizations directly within the conversation. The article Claude can now create interactive charts and visualizations directly…
ChatGPT still leads the chatbot market but its dominance is slipping as Google's Gemini gains ground
Published: March 12, 2026 16:30
ChatGPT continues to dominate the chatbot market, but its lead is shrinking. According to new Similarweb figures, OpenAI's chatbot has fallen from 75.7 to 61.7 percent market share in just twelve months. The big winner: Google Gemini, which has quadrupled…
Iranian propaganda images made with AI end up in major German media outlet
https://the-decoder.com/iranian-propaganda-images-made-with-ai-end-up-in-major-german-media-outlet/
Published: March 12, 2026 15:08
Der Spiegel has removed several images from its Iran coverage because they were most likely generated or altered by AI. The article Iranian propaganda images made with AI end up in major German media outlet appeared first on The Decoder.
Meta's JEPA architecture outperforms standard AI methods in noisy medical imaging
Published: March 12, 2026 15:01
Researchers have presented an AI model for cardiac ultrasound based on Meta's JEPA architecture that outperforms common methods such as masked autoencoder or contrastive learning, according to their benchmarks. The article Meta's JEPA architecture…
Nvidia steps into the open-source AI gap that OpenAI, Meta, and Anthropic left behind
Published: March 12, 2026 14:55
An SEC filing reveals that Nvidia plans to spend $26 billion on open-weight AI models over the next five years. The move doubles as a strategic response to the growing dominance of Chinese open-source models - and a way to keep developers locked into…
Google's new Ask Maps lets you search for places in plain language using Gemini AI
Published: March 12, 2026 13:08
Google Maps is turning into an AI travel advisor. The new "Ask Maps" feature answers complex questions in plain language and shows personalized results on a customized map. Google is also shipping a completely revamped 3D navigation system. The article…
OpenAI is reportedly planning to integrate its video AI Sora into ChatGPT
https://the-decoder.com/openai-is-reportedly-planning-to-integrate-its-video-ai-sora-into-chatgpt/
Published: March 12, 2026 12:30
OpenAI's video AI Sora has dropped from No. 1 to No. 165 in the App Store. Now the company wants to fold it into ChatGPT to reach its 920 million users. The article OpenAI is reportedly planning to integrate its video AI Sora into ChatGPT appeared first on…
Meta unveils four generations of custom AI chips to cut inference costs for billions of users
Published: March 12, 2026 11:54
Meta unveils four new custom AI chips focused on inference, pushing to reduce its dependence on GPU makers like Nvidia and AMD. The article Meta unveils four generations of custom AI chips to cut inference costs for billions of users appeared first on The…
Google unifies text, image, video, and audio in a single vector space with Gemini Embedding 2
Published: March 11, 2026 18:41
Google's first native multimodal embedding model brings text, images, video, audio, and documents into one vector space, cutting out the need for separate models in AI pipelines. The article Google unifies text, image, video, and audio in a single vector…
Half of AI-written code that passes industry test would get rejected by real developers, new study finds
Published: March 11, 2026 17:49
About half of the AI code solutions that pass the popular SWE-bench benchmark would get rejected by actual project maintainers, according to a new study by the research organization METR. The article Half of AI-written code that passes industry test would…
Claude's Excel and PowerPoint add-ins now share context across apps
https://the-decoder.com/claudes-excel-and-powerpoint-add-ins-now-share-context-across-apps/
Published: March 11, 2026 17:00
Anthropic is updating its Claude add-ins for Excel and PowerPoint with shared context, reusable workflows, and broader cloud support. The article Claude's Excel and PowerPoint add-ins now share context across apps appeared first on The Decoder.
Grammarly's AI writing tips claim inspiration from experts who never agreed to participate
Published: March 11, 2026 16:40
Grammarly is apparently using the names of journalists and authors without permission for an AI feature called "Expert Review." The article Grammarly's AI writing tips claim inspiration from experts who never agreed to participate appeared first on The…
OpenAI's new training dataset teaches AI models which instructions to trust
https://the-decoder.com/openais-new-training-dataset-teaches-ai-models-which-instructions-to-trust/
Published: March 11, 2026 15:31
OpenAI has released IH-Challenge, a training dataset designed to teach AI models to reliably prioritize trusted instructions over untrusted ones. Early results show significant improvements in both security and prompt injection defense. The article…
Anthropic launches internal think tank to study AI's impact on society and security
Published: March 11, 2026 15:26
Anthropic has launched the "Anthropic Institute," an internal think tank dedicated to studying how powerful AI affects society, the economy, and security. The article Anthropic launches internal think tank to study AI's impact on society and security…
An AI agent hacked McKinsey's internal AI platform in two hours using a decades-old technique
Published: March 11, 2026 15:21
Security firm Codewall turned an offensive AI agent loose on McKinsey's internal AI platform Lilli, a system used by over 43,000 employees for strategy work, client research, and document analysis. No credentials, no insider knowledge, no human assistance.…
Microsoft and rival AI researchers unite to back Anthropic in its escalating legal battle against the Pentagon
Published: March 11, 2026 11:43
A broad coalition is backing Anthropic in its legal fight with the US Department of Defense: Microsoft, dozens of OpenAI and Google employees, former military leaders, and civil rights organizations have all filed amicus curiae briefs supporting the AI…
Amazon gets court order blocking Perplexity's AI shopping agent
https://the-decoder.com/amazon-gets-court-order-blocking-perplexitys-ai-shopping-agent/
Published: March 10, 2026 19:11
Amazon stops Perplexity's shopping bot by court order: a ruling that could shape the future of AI agents in e-commerce. The article Amazon gets court order blocking Perplexity's AI shopping agent appeared first on The Decoder.
ChatGPT now explains math and physics with interactive visualizations
https://the-decoder.com/chatgpt-now-explains-math-and-physics-with-interactive-visualizations/
Published: March 10, 2026 18:05
ChatGPT now displays interactive visualizations for math and physics topics where users can tweak variables and watch graphs update in real time. More than 70 concepts are available at launch. The article ChatGPT now explains math and physics with…
German court says "It's AI" isn't enough to void copyright
https://the-decoder.com/german-court-says-its-ai-isnt-enough-to-void-copyright/
Published: March 10, 2026 17:12
A German regional court has ruled that song lyrics written by a human are still protected by copyright, even if the music was made with AI tools like SunoAI. Simply claiming a work is AI-generated isn't enough to strip that protection, you need proof. The…
Amazon makes senior engineers the human filter for AI-generated code after a series of outages
Published: March 10, 2026 16:10
Following a series of allegedly AI-caused outages, Amazon is turning its senior engineers into human filters for AI-generated code. The article Amazon makes senior engineers the human filter for AI-generated code after a series of outages appeared first on…
Nvidia and Mira Murati's Thinking Machines Lab announce long-term AI partnership
Published: March 10, 2026 15:14
Nvidia and Thinking Machines Lab, the AI startup founded by former OpenAI executive Mira Murati, are entering a long-term partnership. The article Nvidia and Mira Murati's Thinking Machines Lab announce long-term AI partnership appeared first on The…
Meta acquires Moltbook, the Reddit-style platform built for AI agents
https://the-decoder.com/meta-acquires-moltbook-the-reddit-style-platform-built-for-ai-agents/
Published: March 10, 2026 14:25
Meta has acquired Moltbook, a platform designed as a kind of Reddit specifically for AI agents. The article Meta acquires Moltbook, the Reddit-style platform built for AI agents appeared first on The Decoder.
Startup claims first full brain emulation of a fruit fly in a simulated body
Published: March 10, 2026 13:56
Eon Systems says it has connected a complete fruit fly brain emulation to a virtual body, producing multiple behaviors for the first time. The emulation covers over 125,000 neurons and 50 million synapses. The article Startup claims first full brain…
Philosopher David Chalmers: Current AI interpretability methods miss what matters most
Published: March 10, 2026 12:03
Philosopher David J. Chalmers proposes interpreting AI systems through their attitudes toward propositions - much like we interpret humans. His concept of "propositional interpretability" aims to put mechanistic AI explanation on new footing, drawing on…
Investors bet $1 billion on Yann LeCun's vision for AI beyond LLMs
https://the-decoder.com/investors-bet-1-billion-on-yann-lecuns-vision-for-ai-beyond-llms/
Published: March 10, 2026 10:47
Yann LeCun, former chief AI scientist at Meta and Turing Award winner, has raised over $1 billion for his new startup Advanced Machine Intelligence Labs (AMI Labs) - making it Europe's largest seed funding round ever. The article Investors bet $1 billion…
Claude Code gets parallel AI agents that review code for bugs and security gaps
Published: March 10, 2026 09:12
Anthropic has released a code review feature for Claude Code that automatically checks changes for errors before they're merged. The article Claude Code gets parallel AI agents that review code for bugs and security gaps appeared first on The Decoder.
OpenAI plans to acquire Promptfoo and bake AI security testing directly into its Frontier enterprise platform
Published: March 9, 2026 19:10
OpenAI is acquiring AI security platform Promptfoo to build automated vulnerability testing, covering jailbreaks, prompt injections, and data leaks, directly into its Frontier enterprise platform. The article OpenAI plans to acquire Promptfoo and bake AI…
Microsoft brings Anthropic's Claude Cowork into Copilot to run tasks across Outlook, Teams, and Excel
Published: March 9, 2026 17:51
With Copilot Cowork, Microsoft taps Anthropic's Claude instead of OpenAI to let AI handle tasks across Outlook, Teams, and Excel autonomously. The article Microsoft brings Anthropic's Claude Cowork into Copilot to run tasks across Outlook, Teams, and Excel…
Anthropic's groundbreaking lawsuit challenges the government's power to punish AI safety decisions
Published: March 9, 2026 17:15
Anthropic is suing 17 US federal agencies. The 48-page complaint reveals how deeply Claude is already embedded in classified Pentagon systems and how the government pressured the company with contradictory threats when it refused to drop its AI safety…
Millions already use AI chatbots for financial advice, but experts warn of clear limits
Published: March 9, 2026 14:01
Millions of people are already using chatbots like ChatGPT for retirement planning, the Financial Times reports. The article Millions already use AI chatbots for financial advice, but experts warn of clear limits appeared first on The Decoder.
Study warns of "AI Brain Fry" as workers hit cognitive limits overseeing AI agents
Published: March 9, 2026 13:51
A BCG study of nearly 1,500 workers shows that simultaneously overseeing too many AI tools triggers cognitive exhaustion. The consequences are measurable, from higher error rates to increased intent to quit. The article Study warns of "AI Brain Fry" as…
U.S. Military strikes 3,000 targets in Iran with AI support, but oversight remains "underinvested"
Published: March 9, 2026 13:48
The Wall Street Journal confirms and expands on previous reports about the massive use of generative AI in the U.S. military campaign against Iran. New details reveal how deeply AI is already embedded in intelligence, targeting, and logistics. The article…
OpenAI employees hint at a new omni model
https://the-decoder.com/openai-employees-hint-at-a-new-omni-model/
Published: March 9, 2026 13:41
A new omni model from OpenAI? Employee posts and a leaked audio project called "BiDi" suggest the company is working on its next big multimodal upgrade. The article OpenAI employees hint at a new omni model appeared first on The Decoder.
Anthropic's Claude Opus 4.6 saw through an AI test, cracked the encryption, and grabbed the answers itself
Published: March 9, 2026 11:11
Anthropic's Claude Opus 4.6 independently figured out it was being tested during a benchmark, identified the specific test, and cracked its encrypted answer key. According to Anthropic, this is the first documented case of its kind. The article Anthropic's…
Luma AI's new Uni-1 image model tops Nano Banana 2 and GPT Image 1.5 on logic-based benchmarks
Published: March 8, 2026 17:08
Luma AI takes on OpenAI and Google with Uni-1, a model that combines image understanding and generation in a single architecture and reasons through prompts as it creates. The article Luma AI's new Uni-1 image model tops Nano Banana 2 and GPT Image 1.5 on…
LLM text data is drying up, but Meta points to unlabeled video as the next massive training frontier
Published: March 8, 2026 14:21
A research team from Meta FAIR and New York University trained a multimodal AI model from scratch and found that several common assumptions about how these models should be built don't hold up. The article LLM text data is drying up, but Meta points to…
Hallucinated references are passing peer review at top AI conferences and a new open tool wants to fix that
Published: March 8, 2026 09:00
Accepted papers at leading AI conferences contain hallucinated references: citations that don't point to any real publication. A new tool called CiteAudit aims to tackle the problem systematically for the first time. The article Hallucinated references are…
OpenAI's hardware and robotics chief quits over military deal she says lacked enough deliberation
Published: March 8, 2026 08:11
OpenAI's head of robotics, Caitlin Kalinowski, is leaving the company over its Pentagon deal, citing concerns about mass surveillance and lethal autonomy. The article OpenAI's hardware and robotics chief quits over military deal she says lacked enough…
AI agent benchmarks obsess over coding while ignoring 92% of the US labor market, study finds
Published: March 8, 2026 07:15
A large-scale study shows that AI agent development focuses almost entirely on programming tasks, ignoring the vast majority of the labor market. The article AI agent benchmarks obsess over coding while ignoring 92% of the US labor market, study finds…
Trump administration drafts AI contract rules requiring companies to license systems for "all lawful use"
Published: March 7, 2026 18:09
New US guidelines for AI contracts would force companies to grant the government an irrevocable license for "all lawful use" and ban ideological bias in AI outputs, which is itself an ideological requirement with striking parallels to China's approach. The…
Anthropic's Claude AI uncovers over 100 security vulnerabilities in Firefox
https://the-decoder.com/anthropics-claude-ai-uncovers-over-100-security-vulnerabilities-in-firefox/
Published: March 7, 2026 12:39
Anthropic's AI model Claude has found over 100 bugs in Firefox, including vulnerabilities that decades of testing had missed. The article Anthropic's Claude AI uncovers over 100 security vulnerabilities in Firefox appeared first on The Decoder.
When language models hallucinate, they leave "spilled energy" in their own math
Published: March 7, 2026 12:20
When large language models hallucinate, they leave measurable traces in their own computations. Researchers at the Sapienza University of Rome have developed a training-free method that picks up on these traces and generalizes better than previous…
OpenAI offers open-source maintainers six months of free ChatGPT Pro and Codex access
Published: March 7, 2026 11:44
OpenAI supports open-source developers with free access to ChatGPT Pro, Codex and security tools. The article OpenAI offers open-source maintainers six months of free ChatGPT Pro and Codex access appeared first on The Decoder.
Anthropic's Claude Code subscription may consume up to $5,000 in compute per month while charging the user just $200
Published: March 7, 2026 11:34
Anthropic's $200 Claude Code subscription could consume up to $5,000 in compute per user, according to Cursor's internal analysis reported by Forbes. The numbers reveal just how aggressively AI companies are subsidizing their coding tools and what that…
Bytedance's open-weight Helios model brings minute-long AI video generation close to real time
Published: March 7, 2026 11:00
Bytedance researchers say Helios is the first 14-billion-parameter video model to hit 19.5 FPS on a single GPU while generating minute-long clips. Code and weights are publicly available. The article Bytedance's open-weight Helios model brings minute-long…
OpenAI and Oracle stop expanding their flagship data center in Texas over power supply delays
Published: March 7, 2026 10:15
OpenAI and Oracle are halting the expansion of their Stargate data center in Texas. Instead, OpenAI plans to invest in Nvidia's next-gen Vera Rubin chips at new locations. The article OpenAI and Oracle stop expanding their flagship data center in Texas…
Anthropic turns Claude Code into a background worker with local scheduled tasks
Published: March 7, 2026 09:42
Claude Code Desktop now lets users schedule recurring tasks that run automatically, like checking error logs and creating pull requests for fixable bugs every few hours. The article Anthropic turns Claude Code into a background worker with local scheduled…
Despite Pentagon ban, Google, AWS, and Microsoft stick with Anthropic's AI models
Published: March 7, 2026 09:21
Anthropic in a clinch with the Pentagon: Google, Amazon and Microsoft are backing their AI partner, at least outside the military. The article Despite Pentagon ban, Google, AWS, and Microsoft stick with Anthropic's AI models appeared first on The Decoder.
Anthropic's new marketplace lets enterprise customers spend their existing AI budget on third-party tools
Published: March 7, 2026 08:56
Anthropic is launching the Anthropic Marketplace, a storefront where corporate customers can buy third-party software built on Anthropic's AI models. The article Anthropic's new marketplace lets enterprise customers spend their existing AI budget on…
Video AI models hit a reasoning ceiling that more training data alone won't fix, researchers say
Published: March 7, 2026 08:32
An international research team has released the largest dataset for video reasoning to date, roughly a thousand times bigger than previous alternatives. The results show that even Sora 2 and Veo 3.1 fall far behind humans when it comes to reasoning tasks.…
OpenAI launches Codex Security, an AI agent designed to detect vulnerabilities in software projects
Published: March 6, 2026 19:15
OpenAI's new Codex Security agent automatically hunts for vulnerabilities in code and has already found gaps in OpenSSH and Chromium. The article OpenAI launches Codex Security, an AI agent designed to detect vulnerabilities in software projects appeared…