The day's AI news, briefed
The Daily Prompt
Our AI editor reads each day's AI and robotics stories, pulls out why they matter, and links every original source — so you can keep up in minutes.
Join over 250,000 subscribers getting the latest AI news
Latest AI & Robotics News Today
EmbodMocap: In-the-Wild 4D Human-Scene Reconstruction for Embodied Agents
The development of EmbodMocap could significantly enhance the capabilities of embodied AI, enabling better human-scene interactions in everyday environments without expensive setups.
- EmbodMocap utilizes a portable dual-iPhone system for metric-scale human-scene reconstruction.
- The method supports embodied AI tasks such as physics-based animation and robot motion control.
- It addresses limitations of existing capture systems which often rely on costly setups.
EmbodMocap is a novel pipeline designed for human-scene reconstruction that enables embodied AI, which refers to artificial agents that interact with their environment. The technology seeks to improve the collection of motion data in real-world settings, allowing for more effective training of AI agents. Its adoption could lead to advancements in diverse AI applications like robotics and animation.
Read the original ↗.jpg)
Uncanny Valley: Pentagon vs. ‘Woke’ Anthropic, Agentic vs. Mimetic, and Trump vs. State of the Union
The mounting conflict between Anthropic and the Pentagon underscores the complexity of government technology contracts and the debate over the ethical use of AI in military applications.
- Anthropic is in a growing dispute with the Pentagon regarding the use of its AI technologies due to differing views on military applications.
- The company has imposed restrictions on its AI, disallowing its use in domestic surveillance and fully autonomous weapons.
- Defense Secretary Pete Hegseth has issued an ultimatum to Anthropic, demanding compliance with Pentagon requirements or risk losing their contract.
Anthropic is an AI research company that recently secured an up to $200 million contract with the Defense Department. Tensions have arisen over Anthropic's ethical positions on AI deployment, which differ from the Pentagon's expectations. Observers are closely watching how this dispute influences future tech collaborations with the government.
Read the original ↗
OpenAI Fires an Employee for Prediction Market Insider Trading
This incident highlights ongoing concerns regarding insider trading in prediction markets, which could undermine the integrity of financial and technological ecosystems.
- OpenAI fired an employee for using confidential company information in connection with prediction markets like Polymarket.
- Suspicious trading activities related to OpenAI events were flagged by Unusual Whales, indicating potential insider trading.
- Prediction markets have gained popularity, allowing users to bet on various future outcomes, raising concerns about insider knowledge exploitation.
Prediction markets like Polymarket and Kalshi allow users to speculate on future events, but they have raised concerns about insider trading among tech employees. This incident is a significant moment in addressing these ethical violations in tech companies.
Read the original ↗
Anthropic vs. the Pentagon: What’s actually at stake?
The dispute between Anthropic and the Pentagon has significant implications for the future of military AI use and the ethical guidelines surrounding its deployment.
- Anthropic opposes the use of its AI models in mass surveillance of Americans and autonomous weapons lacking human oversight.
- Defense Secretary Pete Hegseth asserts that the Pentagon should use Anthropic’s technology for any lawful application without restrictions from the company.
- Anthropic faces potential designation as a supply chain risk, which could limit its ability to conduct business with the government.
Anthropic is a company focused on developing AI technology with ethical safeguards. The Pentagon's current policies allow for limited use of autonomous weapons, leading to concerns about military transparency and control over AI development.
Read the original ↗
Musk bashes OpenAI in deposition, saying ‘nobody committed suicide because of Grok’
Musk's comments regarding OpenAI's safety record could impact ongoing lawsuits and highlight contrasting approaches to AI safety amidst rising concerns about mental health linked to AI interactions.
- Elon Musk criticized OpenAI's safety record during his deposition in a lawsuit against the company, claiming his AI firm xAI prioritizes safety better than OpenAI.
- Musk stated that nobody has taken their own life due to the xAI system Grok, in contrast to alleged incidents linked to ChatGPT.
- His deposition discusses broader concerns about AI safety and a public letter he co-signed calling for a pause in AI development beyond GPT-4, citing insufficient planning in the field.
The lawsuit stems from Musk's claim that OpenAI transitioned from a nonprofit to a for-profit entity in violation of its founding agreements. Concerns over AI safety have gained prominence, particularly with rising allegations surrounding ChatGPT's impact on mental health. Following Musk's deposition, xAI has also faced scrutiny related to nonconsensual content being spread through its Grok system.
Read the original ↗
ChatGPT reaches 900M weekly active users
Reaching 900 million weekly active users demonstrates ChatGPT's significant growth and increasing relevance in the AI chatbot market, highlighting its widespread adoption for various tasks.
- ChatGPT now has 900 million weekly active users, up from 800 million reported in October 2025.
- OpenAI also announced it has 50 million paying subscribers, with significant growth in early 2026.
- The company raised $110 billion in private funding, including major investments from Amazon, Nvidia, and SoftBank.
ChatGPT is an AI chatbot developed by OpenAI, used for tasks like learning and planning. The company previously reported 800 million active users in late 2025. The latest funding round is among the largest in history, indicating strong investor interest in AI technologies.
Read the original ↗
MediX-R1: Open Ended Medical Reinforcement Learning
The development of MediX-R1 could significantly enhance the reliability of medical reasoning in AI systems, improving clinical decision-making in healthcare settings.
- MediX-R1 is an open-ended reinforcement learning framework for medical multimodal large language models (MLLMs), allowing for free-form answers beyond multiple-choice formats.
- The framework employs a composite reward system, including LLM-based accuracy and semantic rewards, to enhance medical reasoning.
- MediX-R1 outperformed existing large open-source models, achieving a score of 73.6% with its 30B model on medical benchmarks despite a smaller training dataset of approximately 51,000 examples.
- The evaluation framework uses LLM-based metrics to gauge semantic correctness and contextual alignment, moving away from traditional metrics that may not accurately reflect reasoning quality.
MediX-R1 aims to address limitations in existing medical AI by utilizing a diverse set of reward signals tailored to clinical scenarios. Previous models primarily relied on traditional question formats, which restricted the scope of AI's reasoning capabilities. Researchers will continue to refine such frameworks to further improve the effectiveness of AI in healthcare applications.
Read the original ↗.jpg)
Wall Street Has AI Psychosis
The stock market's volatility in response to AI-related predictions highlights ongoing investor anxiety and uncertainty about the economic impacts of artificial intelligence.
- Alap Shah and Citrini's blog predicted an AI-induced economic downturn, projecting over 10% unemployment by June 2028 and significant stock market declines.
- Following the blog's publication, the Dow Jones Industrial Average fell by 800 points, reflecting investor concerns about AI's future impact on jobs and the economy.
- Critics of the report argue that its predictions lack substance and that AI's economic effects have been minimal to date.
When Alap Shah coauthored a blog discussing AI's potential negative impact on the economy, it triggered a significant sell-off in financial markets, demonstrating widespread anxiety about technological change. Economic predictions regarding AI remain contentious, with many analysts disputing gloomy forecasts and cautioning against undue panic.
Read the original ↗
Mercury 2: The AI Model That Feels Instant
The introduction of the Mercury 2 reasoning model could redefine user experience by significantly reducing response times and costs for AI applications, facilitating real-time interactions across industries.
- Mercury 2, developed by Inception Labs, uses a diffusion approach to generate responses much faster than traditional AI models.
- It achieves a throughput of approximately 1,000 tokens per second, making it over ten times faster than comparable models like GPT-5 mini and Claude 4.5.
- The model is designed to refine responses in parallel, enhancing both speed and quality while remaining cost-effective at $0.25 per million input tokens.
- Mercury 2 scores competitively on various benchmarks, indicating that its rapid processing does not compromise answer quality.
- This innovation could pave the way for new applications in real-time AI interactions, such as voice assistants and search tools.
Mercury 2 is a new large language model from Inception Labs that employs a diffusion-based method for generating text. This approach allows for high-speed responses without sacrificing quality, addressing latency issues faced by existing AI models. Observers will be keen to see how this model performs in practical applications amidst growing demand for responsive AI solutions.
Read the original ↗
AI music generator Suno hits 2M paid subscribers and $300M in annual recurring revenue
The significant growth in Suno's subscriber base and revenue highlights the increasing demand for AI-generated music, which may reshape the music industry and its copyright challenges.
- Suno, an AI music generator, has grown to 2 million paid subscribers and achieved $300 million in annual recurring revenue.
- Just three months prior, Suno reported an annual revenue of $200 million, indicating rapid growth.
- The company enables users to create music from natural language prompts, catering to individuals with minimal experience in music production.
Suno allows users to generate music easily and has recently faced copyright lawsuits from musicians and record labels. It successfully settled a suit with Warner Music Group, enabling the use of licensed music in its models.
Read the original ↗
Sakana AI Introduces Doc-to-LoRA and Text-to-LoRA: Hypernetworks that Instantly Internalize Long Contexts and Adapt LLMs via Zero-Shot Natural Language
Sakana AI's new technologies could enhance the adaptability and efficiency of large language models (LLMs) by allowing them to process longer contexts with zero-shot capabilities.
- Sakana AI has introduced two technologies: Doc-to-LoRA and Text-to-LoRA.
- These hypernetworks can instantly internalize long contexts and adapt LLMs without prior training.
Sakana AI focuses on developing advanced AI technologies, particularly in natural language processing. The introduction of these hypernetworks marks a significant step in enhancing the capabilities of LLMs. Watch for how these developments may influence future AI applications.
Based on the headline only — the full article could not be fetched.
Read the original ↗
OpenAI Just Raised $110 Billion – Here’s Why That Number Should Blow Your Mind
The unprecedented $110 billion funding round for OpenAI signals a significant shift in AI infrastructure development and competition among tech giants, influencing future tech applications and tools.
- OpenAI raised $110 billion in a private funding round, the largest in history, valuing the company at $730 billion.
- New funding includes major partnerships with Amazon and Nvidia, with significant investments in computing services and infrastructure.
- Amazon's $50 billion investment includes $35 billion contingent on specific conditions such as achieving artificial general intelligence or going public.
OpenAI is a leading artificial intelligence research organization known for its developments in AI technology, including ChatGPT. This funding marks a shift from research to practical AI applications, likely affecting the tech landscape significantly. Companies like Google, Microsoft, and Meta are expected to respond, intensifying competition in AI infrastructure and product development.
Read the original ↗
Employees at Google and OpenAI support Anthropic’s Pentagon stand in open letter
This situation raises critical considerations about the ethical implications of AI technology deployment in military contexts, potentially shaping future corporate governance and public policy.
- Over 300 Google employees and 60 OpenAI employees signed an open letter supporting Anthropic's stance against Pentagon requests for unrestricted AI access.
- The letter urges both companies to maintain Anthropic’s position against the military's use of AI for domestic mass surveillance and fully autonomous weaponry.
- Anthropic faces pressure from the Pentagon, which threatens to label the company a security risk if it doesn't concede to military demands.
Anthropic is an AI company with a partnership with the Pentagon, yet it resists the military's use of its technology for surveillance and autonomous weaponry. The ongoing conflict highlights the tensions between technological advancement and ethical considerations in defense applications.
Read the original ↗
Perplexity’s new Computer is another bet that users need many AI models
Perplexity's new computer illustrates a significant shift towards leveraging multiple AI models to improve task execution, which may influence how businesses approach AI integration and decision-making.
- Perplexity Computer integrates 19 different AI models to create a user agent capable of executing complex tasks independently.
- The tool is available exclusively on the $200/month Perplexity Max subscription and runs in the cloud.
- Perplexity aims to cater to a smaller, targeted user base making significant economic decisions, rather than pursuing mass user growth.
- The company has developed its own AI-optimized search API, moving away from reliance on third-party APIs to enhance its offerings.
- Upcoming features include an iOS launch for the Perplexity Comet browser and a developers conference scheduled for March 11.
Perplexity is known for its AI tools that provide user-friendly interfaces for complex queries. The new computer represents a strategic evolution following its shift from advertising and is part of a trend toward multi-model AI integration. Observers are watching how this approach affects Perplexity's market position against larger competitors.
Read the original ↗
Gemini Seizes the Lead, Investors Panic Over Agentic AI, Optimism at Global AI Summit, Local Versus Cloud
The evolving landscape of AI models and technologies is impacting investor confidence and prompting a reevaluation of how AI systems are deployed across different sectors.
- Google released Gemini 3.1 Pro, achieving top rankings in the Intelligence Index while maintaining competitive pricing.
- Investors sold off shares in SaaS companies due to concerns that agentic AI could disrupt their business models, before a slight recovery following Anthropic's partnerships.
- The Global AI Summit highlighted India's potential as an emerging leader in AI, contrasting with the dominance of the U.S. and China.
- Researchers explored the effectiveness of edge models, which run AI locally, as a possible solution to the increasing demand for large language models.
- The rise in demand for AI capabilities is leading to significant investments in data centers to support larger operations.
Gemini is Google's flagship AI model, and its latest version is designed to outperform competitors while being cost-effective. The Global AI Summit, featuring key nations, shifts focus from hazards of AI to its benefits. The growing interest in AI also draws attention to the operational differences of local versus cloud-based AI systems.
Read the original ↗
OpenAI raises $110B in one of the largest private funding rounds in history
This significant funding round accelerates OpenAI's capacity to develop AI technologies, influencing the competitive landscape and infrastructure of artificial intelligence globally.
- OpenAI has raised $110 billion in private funding, marking one of the largest rounds in history.
- The funding includes a $50 billion investment from Amazon, alongside $30 billion each from Nvidia and SoftBank.
- OpenAI plans to launch new infrastructure partnerships with Amazon and Nvidia as part of this funding round.
- The funding round remains open, with expectations of additional investors joining soon.
- The firm committed to using significant AWS computing resources as part of its partnerships with Amazon.
OpenAI specializes in artificial intelligence research and technology development. This funding round follows a previous $40 billion raise in March 2025 and underscores the importance of scaling AI infrastructure to meet global demand. Future developments should focus on product integrations and potential additional investments as conditions are met.
Read the original ↗
Block lays off 40% of workforce as it goes all-in on AI tools
The job cuts at Block highlight the broader impact of AI on employment across various industries, signaling a potential shift in the labor market.
- Block, led by Jack Dorsey, plans to lay off nearly 40% of its 10,000 employees, amounting to over 4,000 job reductions.
- Dorsey links the workforce reduction to the advancements in AI tools that enable smaller teams to operate more efficiently.
- Despite the job cuts, Block reported a strong financial performance in 2025, with revenue reaching nearly $6.3 billion.
Block is a fintech company known for its payment processing service, Square, and has been investing heavily in AI technology. The layoffs reflect a trend among tech companies like Amazon, which have begun to decrease their workforce in response to AI advancements impacting job roles. Observers are closely watching how these trends will affect future employment in the economy.
Read the original ↗
Our Brain’s “Fear Center” May Guide Complex Learning Decisions, New Research Reveals
Understanding the amygdala's broader role in decision-making could improve insights into human cognition and adaptive behavior, with implications for learning and memory research.
- New research from Dartmouth College reveals the amygdala plays a crucial role in complex learning decisions, rather than solely processing fear.
- Damage to the amygdala disrupts learning, suggesting its significant involvement in decision-making processes.
- The research involved monkeys performing tasks to learn how to maximize rewards, illustrating the amygdala's role in switching learning strategies.
The amygdala is commonly known as the brain's fear center, traditionally viewed as a simple emotional system. Recent studies are redefining its function as integral to more complex cognitive tasks, with ongoing research exploring its mechanisms in decision-making.
Read the original ↗
This Tiny Dinosaur Fossil Could Offer a Missing Link That Explains the Spread of an Unusual Class of Dinosaurs
The discovery of Alnashetri cerropoliciensis enhances understanding of Alvarezsaurs' evolutionary history and their spread across ancient landmasses, potentially rewriting significant aspects of dinosaur lineage.
- Paleontologists discovered a complete fossil of Alnashetri cerropoliciensis in Argentina, providing insights into the Alvarezsaurs, a group of bird-like dinosaurs.
- The fossil is approximately 90 million years old, offering a clearer picture of the evolution of a group previously shrouded in mystery due to incomplete records.
- Alnashetri is thought to have adapted features for an ant-eating diet, such as small teeth and short arms that end in a single claw.
Alvarezsaurs are characterized by features that suggest an ant-eating lifestyle, but their fossil record has been incomplete, especially in South America. The La Buitrera region of Patagonia has been crucial for discovering Cretaceous-era fossils, allowing researchers to fill gaps in dinosaur evolutionary history.
Read the original ↗
OmniGAIA: Towards Native Omni-Modal AI Agents
Developing more capable omni-modal AI assistants can significantly enhance the functionality and usability of artificial intelligence in complex real-world tasks.
- OmniGAIA is a benchmark that assesses multi-modal AI agents on reasoning tasks involving video, audio, and images.
- The OmniAtlas agent enhances tool-use abilities through a method called hindsight-guided tree exploration combined with fine-tuning.
- Current AI models primarily operate in bi-modal contexts, highlighting the need for integrated approaches to achieve general AI capabilities.
OmniGAIA aims to address the limitations of existing AI models that usually focus on single modalities such as vision or language. By enabling cross-modal reasoning and interaction with tools, the initiative represents progress toward advanced AI capabilities. The introduction of OmniAtlas reflects evolving strategies for enhancing AI perception and reasoning.
Read the original ↗