Models, products and AI startups, near-duplicates collapsed to the most credible source
CreditSights' Cisar sees cracks in AI credit boomWinnie Cisar, global head of strategy at CreditSights, says the surge in AI financing is making credit analysis increasingly difficult as investors weigh uncertain returns, potential data center overbuilding, and more complex financing structures.
JetBrains Releases Mellum2.1 Small-Scale MoE ModelJetBrains released its Mellum2.1 model collection on the LocalLLaMA community, including a Thinking-GGUF version with 12B total parameters and 2.5B active parameters in a MoE architecture.
Anthropic's Claude-led CRISPR-like discovery sparks controversyAnthropic said its AI agent Claude led the discovery of a novel enzyme system similar to CRISPR, potentially a new gene-editing mechanism, but outside experts criticized the finding as early and incremental, not yet demonstrating gene editing. Gene-editing company stocks fell after the announcement.
Physician edits to AI messages tied to time burdenA study found that physician edits to AI-generated messages to patients were associated with different response time burdens, suggesting AI-assisted communication may affect physician efficiency.
Ex-OpenAI, Cognition Staff Launch Hone With $60M Seed RoundFormer OpenAI and Cognition AI employees have founded Hone, which builds AI agents that act as professional staffers handling tasks lasting weeks. The company announced a $60 million seed round, entering the crowded AI workplace automation field.
TechCrunch Disrupt 2026 to spotlight AI infrastructure boomAt TechCrunch Disrupt 2026, Ambrosia Energy CEO Ben Longmier and Bloom Energy SVP Bill Thayer will speak on 'Where the AI Infrastructure Boom Is Creating Winners,' discussing how physical constraints like power and data centers create durable business opportunities.
How AI impacts the graduate experienceCareer strategist Elizabeth Lenihan said AI is reshaping the graduate job market: Ireland's knowledge-intensive economy is more exposed to AI, batch applications are common, and graduates need to be more selective to stand out, as entry-level roles demand more judgment from day one.
AI decision models play Pac-Man: jev 1.13 tops leaderboardA Reddit user benchmarked six decision models by having them play Pac-Man in real time and published an open-source leaderboard. jev 1.13 led with an average score of 2,750, followed by GPT-6 Luna and Clef Flash. Users can also play as Pac-Man against model-controlled ghosts.
Only 10% of IT Chiefs Use Agentic AI for ModernizationA Kyndryl survey of 2,000 senior IT decision-makers found only 10% of IT chiefs apply agentic AI to modernization. About half of projects improved operations, under half saw innovation gains, 18% found limited value, and nearly half ran over budget or schedule. A former IBM strategist says the economics don't hold up.
Alphabet seen as next $5 trillion club member after NvidiaThe Motley Fool argues Alphabet, with its full-stack AI position across infrastructure, frontier models, cloud, and distribution, could become the next $5 trillion company by end-2028. Gemini's app has 950M monthly users, and TPUs are now being offered to select customers.
I Put an AI Agent on a Nokia 110A user who used a Nokia 110 feature phone to cut screen time tried to run an AI agent on it. Installing Android failed due to only 48MB of RAM, so after days of reverse engineering he succeeded in running an AI agent and automating some actions.
Council of Europe Presents HUDERIA to Data Protection AuthoritiesThe Council of Europe presented its HUDERIA methodology for AI risk and impact assessment at the 24th meeting of Central and Eastern European data protection authorities, advancing regional AI governance cooperation.
TIME100 AI honoree Angle Bush on building Houston's AI futureAngle Bush, founder and CEO of Black Women in Artificial Intelligence, was named to the 2026 TIME100 AI list. She urges moving beyond AI literacy and adoption to creation and economic opportunity, with her organization now spanning five continents.
Big banks seen keeping up borrowing pace to fund AI boomBig banks are expected to maintain their torrid borrowing pace to fund the AI boom, according to Bloomberg, underscoring how AI investment continues to drive credit markets.
Alexa Plus: great for smart home, not ready for lifeA year-long review of Alexa Plus finds it impressive for smart home tasks and memory-based photo montages, but frustrating due to intrusive full-screen ads and failures to recognize the user, showing it is not yet a reliable life assistant.
Can China match Meta's Muse in AI agent race?South China Morning Post examines whether China can compete with Meta's Muse in the race to harness AI agents, analyzing opportunities and challenges for Chinese companies.
Guadagnino's Sam Altman satire 'Artificial' premieres at NYFFLuca Guadagnino's satirical film 'Artificial' premiered at the New York Film Festival, dramatizing OpenAI CEO Sam Altman's rise and brief ouster while sticking closely to factual events, exploring themes of power and playing God.
Cal AI co-founder raises $10M for new AI agent startup PersonaCal AI co-founder Zach Yadegari launched Persona, a personal AI agent startup, raising $10M led by Vine Ventures. MyFitnessPal previously acquired Cal AI, which had topped $30M in annual revenue. Persona will offer a $179 wearable band that uses cloud models, button activation, and privacy-focused design.
Meta Muse and OpenAI Dots compete in consumer AI agentsMeta's Muse, OpenAI's Dots and xAI's Grok have brought easy-to-use AI agents to mainstream users, with Muse free and Dots paid. The debate centers on whether consumers will hand sensitive data like credit cards and inboxes to agents, and what AI can and cannot do today.
AI Math Breakthroughs Spark Ethereum Wallet Security DebateEthereum researcher Justin Drake urges the industry to prepare a 'bunker mode' for a scenario where AI-powered math could break wallet signatures within months. Vitalik Buterin broadly agrees but warns against rushed migrations, noting he's lost more to botched transitions than hacks. No one has actually broken ECDSA yet.
Singapore PM Wong Warns AI Tech Rally Faces Eventual CorrectionSingapore Prime Minister Lawrence Wong warned that the global tech rally driving regional growth will eventually face a market correction. He said his government aims to capitalize on the current window before trade momentum cools.
AMD Unveils Full-Stack Agentic AI PlatformAMD introduced a full-stack platform for agentic AI spanning EPYC CPUs, Instinct accelerators, Pensando networking and ROCm software. Agentic AI shifts the CPU-to-GPU ratio from 1:4-1:8 for chatbots toward 1:1 or higher, alongside new embedded and PC partnerships.
OpenAI Funding Round Boosts Tiger Global's Hedge FundOpenAI is in talks to raise $30 billion at a $1.4 trillion valuation, with Tiger Global poised for a $5 billion paper profit.
Bain: AI infrastructure buildout needs $6T annual revenueBain projects annual AI infrastructure spending near $1.5 trillion by 2031 and $5-6.5 trillion cumulative data center spending through 2030, requiring AI market revenue approaching $6 trillion a year to justify. Power, electrical labor and community opposition are slowing projects.
OpenAI Used AI to Help Write Warning Email to AustraliaThe Guardian reports OpenAI used AI to help draft an email warning the Australian government that its AI had hacked its websites, highlighting frontier AI security risks and corporate disclosure practices.
LittleBit Achieves Sub-1-Bit LLM CompressionLittleBit (NeurIPS 2025) factorizes weight matrices into low-rank latent factors, binarizes them, and restores magnitude via lightweight learned scales, enabling 1.0 to 0.1 bits per weight while keeping the original inference architecture. The follow-up LittleBit-2 (ICML 2026) adds Joint-ITQ initialization, improving initialization without adding inference overhead.
Google launches local-first AI meeting note-taker ForesightGoogle released AI Edge Foresight, a fully offline Mac meeting note-taker powered by the 740M-parameter on-device EmbeddingGemma 2 model. It transcribes meetings, generates notes, and answers questions using uploaded documents, positioning it as a rival to Granola.
Google Launches 'Universal' Gemini AI Agent for WorkplaceGoogle announced a 'universal' Gemini agent for the workplace, intensifying competition with OpenAI, Anthropic and Microsoft for corporate customers.
Single Prompt Hijacked All AI Agents in an AWS Account, Zenity Researchers FindZenity Labs researchers found that a single publicly accessible AI agent on Amazon's Bedrock AgentCore could hijack all agents in the same AWS account and region. The attack exploited an internal AWS interface for temporary credentials. AWS has patched the issue and tightened default permissions.
OpenAI says Russian operation used ChatGPT to trick schools, politicians, mediaOpenAI reported that a Russian influence operation used ChatGPT to trick schools in Latin America, stoke Ukraine-Poland tensions and draw responses from politicians in several countries.
The A.I. Daily