Ai · Top stories
Greek Prime Minister Discusses AI and Economic Recovery During US Visit
Greek Prime Minister Kyriakos Mitsotakis visited San Francisco to promote Greece's economic recovery and discuss AI, meeting with tech leaders and investors. He highlighted Greece's improved economic standing, digital infrastructure investments, and policy changes aimed at attracting tech companies.
Waymo Launches Transit Rewards Program in San Francisco Bay Area
Waymo introduced a transit rewards program in the San Francisco Bay Area, offering riders Waymo Cash when they combine Waymo trips with public transit using a linked Visa card. This initiative aims to integrate autonomous vehicle services with public transportation networks, making multimodal travel more convenient and affordable.
China builds numerous data centers in Inner Mongolia to support AI ambitions
China is constructing multiple data centers in Inner Mongolia as part of its strategy to become a global leader in artificial intelligence. This development highlights China's infrastructure investment in AI, contrasting with US attitudes towards the data center industry.
Rabbit releases standalone AI agent OS3, discontinues R1 device manufacturing
Rabbit has launched OS3, a cloud-based AI agent that runs locally on Windows, Mac, and Linux devices, making its functionality available without the R1 hardware. The company has ceased manufacturing the R1 and plans to release a new "cyberdeck" device running OS3 in the coming months.
New N-Operator Correlator (NOC) formulated for measuring quantum information spreading
Researchers introduced N-Operator Correlators (NOC) as a generalized Out-of-Time-Ordered Correlator (OTOC) to measure quantum information spreading. This formulation allows for a broader analysis by incorporating more measurement probes than the two operators used in traditional OTOCs. The development offers a new method for analyzing quantum information dynamics in systems.
Barbara Mazzolai Proposes "Sustainability Robotics" to Reduce Environmental Impact
Roboticist Barbara Mazzolai, from the Italian Institute of Technology, is advocating for a new field called "sustainability robotics." This initiative aims to integrate environmental considerations into robot design to lessen technology's ecological footprint.
MiMo-V2.6-Pro Model Analysis: Intelligence, Performance, and Pricing Detailed
An analysis of the MiMo-V2.6-Pro model shows it scores 46 on the Artificial Analysis Intelligence Index, placing it above the median of 18 for comparable models. The model supports multimodal input, has a 1M token context window, and processes 125 tokens per second, but its input token pricing is somewhat expensive at $0.43 per 1M tokens.
Zelda Williams Condemns AI-Generated Videos of Late Father Robin Williams
Zelda Williams has publicly criticized fans for circulating AI-generated videos of her late father, Robin Williams, stating that such content is disrespectful and manipulative. She reiterated her stance against the use of AI to recreate her father's likeness, calling for an end to these creations.
Ed Davey Calls for Global Treaty to Halt Super-Intelligent AI Development
Ed Davey, leader of the Liberal Democrats, will propose a global non-proliferation treaty for super-intelligent AI, similar to nuclear arms treaties, during his keynote speech at the party conference. He advocates for an international AI safety agency, headquartered in the UK, to regulate AI development due to potential existential risks.
EXL Launches AI-Powered Medical IDP Solution on AWS to Reduce Claims Review Time
EXL introduced an AI-powered Medical Intelligent Document Processing (IDP) solution built on AWS to automate medical claims review. This solution combines IDP with domain-specific large language models to extract, summarize, and query medical information, aiming to reduce the over 100 minutes adjusters spend per case manually reviewing records.
Lossless-memory Project Creates Personal AI Memory System That Avoids Summarization
The Lossless-memory project introduces a personal AI memory system that stores raw conversation logs with timestamps, avoiding summarization and embedding. This approach aims to preserve the exact content and temporal context of interactions for a single user's AI assistant.
Jevchat tool converts Jev into a character-by-character chatbot
A new tool called Jevchat has been developed to turn the Jev model into a chatbot that generates responses one symbol at a time. This experimental project explores different sampling strategies and alphabets for text generation, primarily for fun and demonstration.
ENZO: An open-source, self-hosted AI platform for interacting with 300+ models
ENZO is a newly released open-source AI platform that allows users to chat with over 300 AI models and build agents locally. It emphasizes user control by running on personal infrastructure without accounts or subscriptions, processing requests directly between the user's browser and the chosen AI provider.
AI Actress Tilly Norwood Experiences Malfunctions During Press Tour
Tilly Norwood, an AI-generated actress created by Particle6 Group, exhibited malfunctions during its press tour, including unexpectedly speaking Chinese during an interview. This behavior raises questions about the AI's current capabilities and the production company's strategy.
Rep. Chip Roy Advocates Congressional Oversight for AI, Opposes Regulation
Representative Chip Roy stated that Congress should have an oversight role regarding artificial intelligence, but he opposes direct regulation of the technology. He suggested that AI CEOs should testify before Congress to address safety concerns, reflecting a limited government perspective amidst ongoing debate on AI governance.
Google Flow AI Studio Used by Fashion Designers for Workflow Streamlining
Google's Envisioning Studio and Google Labs collaborated with fashion designers Jane Wade and Sergio Hudson to develop specialized tools using Google Flow, an AI creative studio. These tools helped streamline design and production workflows for New York Fashion Week, allowing designers to virtually style collections and visualize runway staging within budget constraints. This initiative demonstrates a practical application of AI in the fashion industry to address specific production challenges.
Neutrino Flavor Changes May Explain Supernova Discrepancies
A new paper in Physical Review D suggests that neutrino flavor changes could resolve inconsistencies in current supernova models. This research addresses the discrepancy between observed supernova rates and star formation, as well as the 'mass gap' in black hole formation.
Prior Software Launches Predictive Intelligence Platform for Future Event Forecasting
Prior Software Inc. launched "Prior," a predictive intelligence platform that forecasts outcomes for various future events, including sports, markets, and personal decisions. The platform gathers context, runs simulations, and provides weighted outcomes and probabilities to users.
UK Government Pilots AI Bootcamps to Address Youth Unemployment
The UK government launched a pilot program of AI bootcamps in the North-West to train 16- to 24-year-olds not in education, employment, or training (NEETs) in AI skills. This initiative aims to equip young people with relevant skills for the evolving job market and connect them with AI-focused apprenticeships offered by partner companies.
AI tools assist managers in rehearsing difficult workplace conversations
New AI coaching tools are emerging to help managers prepare for difficult workplace conversations, such as performance reviews or job eliminations. These tools allow managers to practice delivering messages and simulate employee reactions, addressing a common struggle among managers in handling such interactions.
Engineer creates Stonkfly, a simulated fruit fly brain that day trades cryptocurrency
A Coinbase engineer developed Stonkfly, an open-source project that simulates a fruit fly brain with 166,700 neurons to day trade cryptocurrency based on candlestick charts. The simulation uses positive reinforcement (dopamine rush) for profitable trades and negative reinforcement for losses, without employing leverage or short selling. This project demonstrates the connection between visual input, neural changes, and trading decisions, though the author cautions against interpreting it as actual trading ability.
Mathematical model explores game theory with evolving strategies and random rewards
Researchers developed a mathematical model to study game theory scenarios where strategies evolve and rewards vary randomly. This approach addresses limitations of traditional game theory, which typically assumes constant rewards, by better reflecting real-world strategic choices.
ElevenLabs Releases Music v2.5 Model and Introduces Track Downloads
ElevenLabs has released Music v2.5, an updated music generation model, which is now the default in ElevenMusic. The update also introduces the ability for users to download generated tracks, with daily or monthly limits depending on the subscription plan.
Altimeter CEO Brad Gerstner Criticizes AI Extinction Warnings as Hyperbolic
Altimeter Capital CEO Brad Gerstner stated that warnings from AI researchers about human extinction are detached from reality and hyperbolic. Gerstner, whose firm invests in major AI companies, argued that significant safety measures are already in place for AI development, unlike previous technologies.
Claude AI now requires users to be 18 or older, implements age verification
Anthropic's AI assistant, Claude, now restricts access to users aged 18 and above. The platform will use third-party age verification via Yoti for accounts flagged as potentially belonging to minors, a measure implemented to ensure compliance with age restrictions.
Airbnb CEO Brian Chesky Discusses AI's Potential at Communacopia Conference
Airbnb CEO Brian Chesky stated at the Goldman Sachs Communacopia + Technology Conference that AI is a "cognitive revolution" that humans can surf rather than be swallowed by. He compared AI's potential to nuclear power, emphasizing that its outcome depends on human direction.
Researcher Accuses OpenAI of Training on Conversations and Claiming Breakthrough
A researcher has accused OpenAI of using conversational data for training its models and subsequently presenting the results as a breakthrough. This accusation raises questions about the originality and ethical sourcing of data used in AI development.
CoreWeave CEO attributes public AI pushback to rapid technological change
CoreWeave CEO Mike Intrator stated that public opposition to AI and data centers stems from the rapid pace of change, not the technology itself. He suggested that companies have not adequately communicated the benefits of AI to the public. This perspective was shared at the Goldman Sachs Communacopia + Technology Conference, where AI adoption and public pushback were key discussion points.
Google DeepMind and Primordial Soup Use AI to Recreate Couple's Unrecorded Memory
Google DeepMind and Darren Aronofsky's Primordial Soup collaborated on "Love, Rendered," a documentary short that used AI to recreate a couple's unrecorded memory of how they met. The project applied image restoration and video animation models to help a couple with memory loss visualize a significant past event. This demonstrates a novel application of AI in reminiscence therapy, focusing on memories without existing visual cues.
New AI code comment detector rebuilt with public data and improved foundation
A new AI code comment classifier has been developed using public data, offering a balanced accuracy of 77% in distinguishing human-written from AI-generated comments. This tool provides calibrated probability percentages for its verdicts and allows users to inspect feature activations for individual classifications, addressing limitations of a previous version built on private data.
Lyft integrates Waymo driverless car service in Nashville
Lyft now offers Waymo driverless car rides in Nashville, making it the first market where Waymo vehicles are available through both the Lyft app and Waymo's own application. This expansion allows users to opt into autonomous rides within a specific central Nashville service area at no extra cost.
Deltafin fork runs Kimi K3 (2.8T) at 1 token/s on MacBook Pro using four SSDs
A fork of Deltafin, named gavamedia/deltafin, has demonstrated running the Kimi K3 large language model (2.8T parameters) on an Apple Silicon MacBook Pro, achieving approximately 1 token per second by streaming data from four SSDs. This experiment explores the limits of consumer hardware for running large models without compromising model quality.
Trailer Released for Sam Altman Biopic "Artificial" Starring Andrew Garfield
The first trailer for "Artificial," a biopic about OpenAI CEO Sam Altman, has been released, featuring Andrew Garfield in the lead role. Directed by Luca Guadagnino, the film covers Altman's 2023 firing and reinstatement at OpenAI, and is set to premiere on Christmas Day.
New visualization tool shows how LLMs use attention to generate text
A new tool visualizes the attention mechanism in transformer-based large language models, showing which previous tokens influence the generation of subsequent tokens. This visualization helps explain how LLMs can accurately copy information and generate coherent text by selectively drawing from past context.
TradingAgents Multi-Agent LLM Financial Trading Framework Open-Sourced
The TradingAgents multi-agent LLM financial trading framework has been open-sourced, allowing community access to its specialized LLM-powered agents for financial analysis and trading. This open-sourcing provides developers and researchers with a tool to experiment with and build upon LLM-driven trading strategies.
OpenAI reinstates 5-hour usage limit for Plus and Business Standard users
OpenAI has brought back a 5-hour usage limit for its Plus and Business Standard subscribers. This change means that limit resets are less valuable for users.
Trump Posts AI-Generated Images and Unverified Claims on Truth Social
Donald Trump posted numerous AI-generated images and unverified claims on Truth Social, including a statement about making "Hundreds of Billions of Dollars on Stocks" for the U.S. These posts included depictions of Iran's defeat and other political imagery, alongside claims lacking supporting evidence.
Dolly Parton's Sister Pleads for End to 'AI Garbage' Posts After Singer's Death
Dolly Parton's sister, Stella Parton, has publicly requested an end to "fake AI garbage" posts circulating on social media following the singer's death. This plea highlights a growing concern among celebrity families regarding AI-generated content created in the wake of a loved one's passing, emphasizing the insensitivity and pressure it creates.
Fire at Lake Mariner AI Data Center Reveals Safety Lapses and Complex Ownership
A fire at the unfinished Lake Mariner AI data center in New York exposed significant safety deficiencies, including non-functional alarms, suppression systems, and hydrants, according to local fire officials. The incident highlights potential risks in large-scale AI infrastructure projects and the complexities of their corporate structures.
Fairground Launches AI Creator TV Channel on Roku and Other Platforms
Fairground Entertainment launched AI Creator TV, a free ad-supported streaming television (FAST) channel featuring films made with generative AI, now available on Roku, Xumo Play, and Canela. The channel curates and distributes content from over 100 AI filmmakers, demonstrating a new application for AI-generated media in traditional TV formats.