Tag: artificial intelligence

  • AWS Director Bill Foy on Using AI to Overcome Supply Chain Disruptions

    AWS Director Bill Foy on Using AI to Overcome Supply Chain Disruptions

    In a recent discussion, Bill Foy, Director and APJ Automotive Sales Leader at Amazon Web Services (AWS), highlighted the critical role of artificial intelligence in modernizing supply chains. According to Foy, 94% of companies have experienced supply chain disruptions, often due to legacy systems that lack real-time visibility.

    “Whether it’s a port strike, an earthquake, or a natural disaster, most companies don’t even know a problem exists until it’s downstream,” Foy explained. This fragmented data landscape poses a particular challenge for original equipment manufacturers (OEMs), who must stitch together information from disparate sources to maintain operations.

    AWS is addressing these issues by offering AI-driven solutions that provide end-to-end visibility and predictive analytics. By integrating machine learning models with existing enterprise systems, companies can detect disruptions early, optimize inventory, and automate decision-making. Foy emphasized that moving away from legacy infrastructure is not just an IT upgrade but a strategic imperative for resilience.

    “Modernizing with AI isn’t about replacing people—it’s about empowering them with data they can trust and act on in real time,” he added. As supply chains grow more complex, AWS’s focus on AI-powered modernization aims to reduce downtime and improve agility across industries, especially in automotive manufacturing.

  • Jinhua Zhao Appointed Lead of MIT’s Urban Studies and Planning Department

    Jinhua Zhao Appointed Lead of MIT’s Urban Studies and Planning Department

    Jinhua Zhao MCP ’04, SM ’04, PhD ’09, a renowned transportation planner and educator, has been appointed head of the Department of Urban Studies and Planning (DUSP) at MIT, effective July 1. Zhao, the Class of 1941 Professor of Cities and Transportation, brings deep expertise in behavioral science and transportation, combined with artificial intelligence and public policy, to tackle urban challenges.

    MIT School of Architecture and Planning Dean Hashim Sarkis praised Zhao as a rare scholar who moves fluidly between cutting-edge research and real-world policy. ‘His work with governments and transportation agencies around the world is a model for what MIT’s impact can look like beyond our campus,’ Sarkis said. Zhao succeeds Professor Christopher Zegras, who served as department head since 2020.

    Zhao earned all three of his advanced degrees at MIT before joining the faculty. He noted that the Institute’s unconventional culture and interdisciplinary connectivity were key attractions. ‘We have fewer boundaries than other universities — intellectually and physically. Our infinite corridor literally connects us to so many disciplines,’ he said.

    Globally, Zhao has shaped mobility systems through collaborations with Transport for London, Hong Kong’s Mass Transit Railway, Japan Railways, and U.S. agencies like the MBTA, Chicago Transit Authority, and Washington Metropolitan Area Transit Authority. He has also guided autonomous vehicle deployment strategy in Singapore and the Middle East. ‘Every city I’ve worked with faces the same tension: The technology is moving faster than the institutions designed to govern it,’ Zhao explained. ‘My work has been about closing that gap.’

    At MIT, Zhao founded the MIT Mobility Initiative and hosts the weekly MIT Mobility Forum, which has grown from a small internal list to a global platform drawing over 200 practitioners, policymakers, and researchers weekly. He also directs the JTL Urban Mobility Lab, which unites behavioral science and transportation technology, and is a lead principal investigator with the Mens, Manus, and Machina initiative exploring AI, work, and learning.

    As department head, Zhao aims to accelerate the transfer of DUSP research to city leaders, planners, and engineers who make decisions on issues like aging infrastructure, AI integration, and congestion. ‘We know a great deal about how cities grow, how people move, and how that will change. The question is whether the people responsible for making these changes can access what we know, when they need it,’ he said.

    Zhao’s appointment underscores MIT’s commitment to addressing urgent urban challenges through interdisciplinary research and practice.

  • AI News Roundup: Investment Hurdles, Strategic Scaling, Chip Deals, and Global Regulation Efforts

    AI News Roundup: Investment Hurdles, Strategic Scaling, Chip Deals, and Global Regulation Efforts

    Weekly AI News Briefing

    This week’s AI landscape highlights critical shifts from experimentation to enterprise execution, major chip industry deals, and new regulatory frameworks. Below are the key stories.

    • AWS Co-Founder Matt Domo on Why AI Investments Are Stalling – Domo explains how organizations can move beyond pilot projects to full-scale AI deployment and unlock real value.
    • Micron and Rivals Secure US$22bn AI Deals with NVIDIA – To break the semiconductor boom-bust cycle, chipmakers like Micron are locking in multi-year contracts with NVIDIA, ensuring steady revenue streams.
    • McKinsey: Scaling AI Beats Fragmented Business Pilots – Partner Rahul Shahani reports that embedding AI across multiple functions yields double the profit margins compared to isolated experiments.
    • AIB Overhauls Mobile Banking App With Advanced AI Insights – The Irish lender uses machine learning to turn transaction data into personalized financial guidance for customers.
    • Pangaea Data and Sanofi Partner to Tackle Disease Underdiagnosis – Their AI scans electronic health records to help clinicians identify patients with Alpha-1 Antitrypsin Deficiency earlier.
    • UN’s New Environmental Initiative for AI – Secretary-General António Guterres launches a program to track the power and water consumption of AI systems during London Climate Action Week.
    • The Global Awards 2026: Tech and AI Categories – The upcoming awards will recognize AI-led innovation in sustainability, procurement, and supply chain.
    • Tech CEOs Push for US-Led AI Coalition at G7 Summit – Global executives propose an international framework to address national security risks and regulate advanced frontier models.
  • The 10 Most Influential Open-Source AI Platforms of 2026

    The 10 Most Influential Open-Source AI Platforms of 2026

    The open-source AI landscape is evolving at a breathtaking pace, democratizing access to cutting-edge technology and shifting power away from closed-door tech giants. From powerhouse models transforming natural language processing to advanced frameworks orchestrating complex autonomous agents, open-source innovation is the true engine driving the AI revolution. Whether you’re a developer looking to deploy efficient local models or an enterprise building scalable AI infrastructure, these tools offer unprecedented flexibility and control.

    1. Hugging Face (Founded 2016, CEO Clément Delangue) – Hugging Face occupies a unique position at the center of the modern AI ecosystem. While many focus on individual models, Hugging Face builds the infrastructure, community, and distribution network that powers much of the open-source AI movement. It hosts hundreds of thousands of models, datasets, and applications, making it the default destination for discovering, evaluating, and deploying AI technologies.
    2. Llama (Meta AI) (Launched 2023, CEO Mark Zuckerberg) – The Llama family of models has done more than any other project to legitimize open-weight AI at global scale. Successive generations push performance boundaries while remaining accessible to researchers, developers, and businesses. Llama has become the foundation for countless derivative projects and commercial applications worldwide.
    3. DeepSeek (Launched 2023, CEO Liang Wenfeng) – DeepSeek has become one of the most disruptive forces in AI, releasing highly-capable open models that rival established competitors while reportedly requiring far fewer resources. Its achievements challenge assumptions about the cost and complexity of building state-of-the-art AI systems.
    4. Mistral AI (Launched 2023, CEO Arthur Mensch) – Founded by former researchers from Google DeepMind and Meta, Mistral AI has become Europe’s most prominent AI challenger. Its open-weight philosophy combined with strong performance and efficiency attracts developers, enterprises, and governments alike.
    5. Qwen (Alibaba Cloud) (Founded 2023, CEO Eddie Wu) – Alibaba Cloud’s Qwen family has emerged as a strong challenger to Western AI platforms. With multilingual capabilities and impressive performance across reasoning, coding, and enterprise tasks, Qwen has gained substantial adoption throughout Asia and beyond.
    6. OpenClaw (Launched 2025, Creator Peter Steinberger) – OpenClaw represents the growing momentum behind truly open AI development. Built around transparency and accessibility, it attracts developers seeking alternatives to closed commercial systems. Its focus on open weights and community-led innovation reflects a broader movement toward decentralized development.
    7. Microsoft AutoGen (Launched 2023, CEO Satya Nadella) – AutoGen has become a leading framework for multi-agent AI systems, enabling multiple language models and software tools to collaborate on complex tasks. Its integration with the Microsoft ecosystem strengthens its appeal among enterprise users as agentic AI becomes a major trend.
    8. LangChain (Launched 2022, CEO Harrison Chase) – LangChain is one of the most widely used frameworks for building applications powered by large language models. It provides tools to connect AI systems with external data sources, workflows, and software environments, making it easier to build chatbots, autonomous agents, and enterprise AI solutions.
    9. Ollama (Launched 2023, CEO Jeffrey Morgan) – Ollama has democratized access to large language models by making local AI deployment straightforward. Users can run powerful open-source models on personal computers, removing dependence on cloud infrastructure for many use cases, and lowering technical barriers for a broad range of users.
    10. vLLM (Launched 2023, CEO N/A) – vLLM has become one of the most influential open-source inference engines in the AI ecosystem. Originally developed at UC Berkeley, it makes deployment of large language models faster and more cost-effective through innovative memory management, enabling organizations to serve AI applications at scale without excessive infrastructure costs.
  • Top 10 AI Platforms Revolutionizing Media and Content Creation in 2026

    Top 10 AI Platforms Revolutionizing Media and Content Creation in 2026

    Artificial intelligence is fundamentally reshaping the media landscape. From voice synthesis and video generation to enterprise-grade cloud infrastructure, AI platforms are now essential tools for studios, newsrooms, and creators. Here are the top 10 AI platforms driving this transformation.

    1. ElevenLabs – Founded in 2022 and based in New York, ElevenLabs specializes in context-aware voice synthesis. Its technology captures human emotion, pacing, and intonation, enabling creators to produce lifelike audiobooks, game dialogue, and multilingual dubbing at scale.
    2. Midjourney – A self-funded research lab from San Francisco (founded 2021), Midjourney democratizes visual concept design. Its high-fidelity text-to-image generation is widely used for pre-visualization in film, TV, and graphic design, compressing weeks of sketching into minutes.
    3. Runway – Founded in 2018 in New York, Runway leads generative video with powerful video-to-video and text-to-video models. It partners with studios like Lionsgate to provide tools for automated rotoscoping, VFX, and marketing asset production.
    4. Anthropic (Claude) – San Francisco-based Anthropic (founded 2021) offers the Claude family of models. With large context windows and strong safety features, Claude helps news networks and screenwriters process research, organize plots, and edit transcripts without compromising brand safety.
    5. Amazon Web Services (Amazon Bedrock) – AWS (founded 2006, Bedrock in 2023) provides the industrial backend for media distribution. Bedrock offers a unified API to access top foundation models, powering personalized recommendations and metadata management for streaming giants.
    6. Adobe (Firefly) – Adobe (founded 1982, Firefly in 2023) takes a responsible approach by training Firefly on licensed content. Integrated into Photoshop, Premiere, and Illustrator, Firefly allows commercial creators to generate and edit images, video, and audio safely.
    7. Microsoft (Copilot and Azure AI) – Microsoft (founded 1975) weaves AI into media workflows through Azure AI and Copilot. Azure handles real-time translation and video indexing, while Copilot assists journalists with summaries and breaking news drafts.
    8. NVIDIA (AI Enterprise/Omniverse) – NVIDIA (founded 1993) provides the hardware and software foundation for AI media. Its AI Enterprise suite accelerates live audio cleanup and video upscaling, while Omniverse enables collaborative 3D environment building for VFX teams.
    9. Google DeepMind – London-based DeepMind (founded 2010, acquired by Google in 2014) drives multimodal AI breakthroughs. Its research powers video generation, audio synthesis, and predictive analytics for YouTube and archival search.
    10. OpenAI – Founded in 2015 in San Francisco, OpenAI remains the catalyst of generative AI. It has publisher licensing deals (e.g., Getty Images) and provides LLMs for script brainstorming and news synthesis. OpenAI is also expanding into ad delivery tools within ChatGPT.
  • Generalists Outperform Specialists in Certain Game Theory Scenarios, MIT Study Finds

    Generalists Outperform Specialists in Certain Game Theory Scenarios, MIT Study Finds

    In a surprising twist that challenges long-held assumptions in game theory, a new MIT study shows that general-purpose algorithms called policy gradient methods can outperform specialized game-theoretic algorithms in certain imperfect-information games. The findings, presented at the International Conference on Learning Representations in Rio De Janeiro, could reshape how artificial intelligence agents are trained to make decisions in competitive, real-world scenarios.

    Imperfect-information games—where players don’t know everything about their opponents—are common in life, from poker and bidding wars to military operations and financial negotiations. For decades, the prevailing belief was that algorithms specifically designed for these games, grounded in game theory, would always outshine general-purpose alternatives. However, the MIT-led team discovered that policy gradient methods, originally developed in the 1990s for single-agent decision-making, often perform better and with greater efficiency.

    The researchers created a benchmark to fairly evaluate different algorithms, measuring performance through a concept called exploitability—how well a player does against a worst-case adversary. In experiments involving five games, including Phantom Tic-Tac-Toe, imperfect-information Hex, and Liar’s Dice, neural networks trained with policy gradient algorithms consistently achieved lower exploitability scores than those trained with game-theoretic algorithms.

    “Our study showed that policy gradient methods can work better than these specialized algorithms, and that the specialized algorithms may not work as well as people thought,” said Samuel Sokota, a co-author from Carnegie Mellon University. The team’s benchmarking software, which they have made freely available, allows others to test and compare algorithms with just a single line of code added to the OpenSpiel library.

    The implications extend far beyond board games. “Hidden information is a very important property of the world,” said Eugene Vinitsky of New York University, another co-author. “It pervades military operations, trading scenarios, and negotiations—all of which are carried out under conditions of hidden information. The idea that we can improve on these games suggests that we can also do better in these other settings as well.”

    Ian Gemp, a computer scientist and game theory expert at Google DeepMind not involved in the study, called the results encouraging: “This work serves as a compelling reminder that modernizing classical tools remains a highly productive path for solving complex strategic problems.”

  • MIT Develops Long-Term Memory System That Lets Robots Answer Where You Left Your Keys

    MIT Develops Long-Term Memory System That Lets Robots Answer Where You Left Your Keys

    Imagine asking a robot, “Where did I leave my keys?” and getting an accurate, real-time answer. MIT researchers have created a new spatial memory framework called DAAAM (Describe Anything, Anywhere, Anytime, at Any Moment) that gives robots the ability to form and recall detailed mental models of large-scale environments. This breakthrough could transform how robots assist humans in factories, homes, and beyond.

    DAAAM combines advanced map representations with rich, language-based descriptions of objects a robot encounters as it explores. The system runs fast enough for mobile robots to use in real-time, answering complex queries in plain English with 21% to 53% higher accuracy than existing methods.

    “If we want robots to work side-by-side with humans and interact better with humans, they must speak the same language,” says Luca Carlone, associate professor in MIT’s Department of Aeronautics and Astronautics and lead researcher on the project. “The robot must be able to reason about time and space the same way humans do.”

    The framework bridges computer vision and robotic mapping. As a robot moves through an environment, DAAAM attaches detailed descriptions to objects—like noting that a red bicycle with a flat tire is in the bike rack outside the Stata Center. It stores this information in a 3D map-based representation arranged spatially, grouping objects into regions for efficient retrieval.

    To overcome the speed limitations of existing annotation techniques, DAAAM aggregates nearby objects and uses an optimization method to select key frames—images with the clearest view of multiple objects—allowing the system to describe several items in parallel. This speeds up computation tenfold, making real-time performance possible.

    “We annotate every object only once, so our framework can run in very large-scale environments in real time,” explains lead author Nicolas Gorlo, an MIT graduate student. “And by clustering objects into regions, it can answer a wide range of queries about objects and locations.”

    The researchers used a large language model (LLM) that calls on various tools to retrieve specific information quickly, reducing hallucinations. For example, if asked about a sculpture near an MIT campus building, DAAAM can use a semantic search tool to retrieve information based on the word “sculpture” or a location-based tool to find the building.

    Future work aims to expand DAAAM to capture significant events and incorporate confidence levels into responses. “Ultimately, we want to have robots that can help with any sort of tasks,” Gorlo says. “With this framework, we are trying to create the foundations to enable a generalist agent that can do anything you ask.”

    The research was presented at the Conference on Computer Vision and Pattern Recognition (CVPR) and funded by the U.S. Army Research Laboratory and the Office of Naval Research.

  • MIT Leaders and Faculty Show How Massachusetts Can Drive the Next Wave of Tech Innovation

    MIT Leaders and Faculty Show How Massachusetts Can Drive the Next Wave of Tech Innovation

    On June 9, The Boston Globe released its 2026 “Tech Power Players” list, recognizing 50 influential local leaders in technology and business across Massachusetts. The list includes eight MIT affiliates: President Sally Kornbluth, Prof. Daniela Rus (director of CSAIL), Prof. Regina Barzilay, Prof. Yet-Ming Chiang, Prof. Max Tegmark, Ana Bakshi (executive director of the Martin Trust Center for MIT Entrepreneurship), Katie Rae (CEO and Managing Partner of Engine Ventures), and Senior Lecturer Brian Halligan, along with numerous MIT alumni.

    The Power Players coverage highlights MIT’s research labs, its culture of innovation and entrepreneurship, industry connections, new AI initiatives, and the Institute’s commitment to maintaining Massachusetts’ technological leadership. “Massachusetts can absolutely lead in this next wave,” says President Kornbluth, pointing to opportunities in manufacturing, life and health sciences, quantum technologies, and energy that can serve Americans nationwide.

    MIT is “working to drive artificial intelligence forward in sectors where the region is strongest, from biotechnology and robotics to defense and clean energy. It’s also trying to broaden entrepreneurship through a ‘dorm-to-startup’ push, creating a pipeline of support services—from hack-a-thons to venture funding—to help students start companies between classes,” writes Robert Weisman for The Globe. President Kornbluth is reinvigorating the school’s support of the local innovation ecosystem, unveiling new online AI classes—with free entry-level courses for anyone—and encouraging more on-campus entrepreneurship.

    MIT is carving out a specialty in applied AI, sometimes called “AI+X,” deploying the technology to help businesses, hospitals, and research institutions supercharge productivity, innovation, and scientific breakthroughs. Aman Narang ’04, CEO of Toast, notes: “The superpower has always been the university system. The best thing Boston can do is keep these people around.”

    MIT startups are a key driver of the region’s entrepreneurial ecosystem. To maintain Boston as a hub for innovators, MIT is building on its existing entrepreneurship resources, including more than 150 courses and 85 centers and programs. President Kornbluth and Provost Anantha Chandrakasan recently formed the Committee on Accelerating Translation and Entrepreneurship (CATE) to explore how the Institute can better support and accelerate the movement of ideas from research into new ventures.

    Applications for The Martin Trust Center’s startup accelerator program have doubled from last year, and nearly one-fifth of MIT undergraduates—about 800 students—attended a recent startup career fair.

    The Globe also highlights MIT startup Liquid AI, which develops AI models inspired by the brain structure of a simple worm. These models require far less electricity to operate than large language models, saving energy and water. Liquid AI recently signed a deal with Mercedes-Benz to incorporate its technology into onboard systems of cars sold in North America.

    Researchers are also focusing on energy innovation. Prof. Yet-Ming Chiang’s lab is developing batteries that store more electricity over longer periods, creating more opportunities for wind, solar, and other clean energy sources. Chiang’s lab and other MIT research centers work on microchips, critical minerals, fusion technology, and defense tech—examples of “tough tech” that combine science and engineering, which Chiang says “are in the sweet spot of the Boston ecosystem.”

    Soon, 80 MIT students will work as summer interns at GE Vernova, thanks to the MIT-GE Vernova Climate and Energy Alliance, a collaboration aimed at accelerating the global energy transition. GE Vernova CEO Scott Strazik dedicated $50 million over five years to fund internships and research projects.

    When asked about the most promising thing about the Greater Boston tech scene, Prof. Rus answers: “talent. Boston has the best AI researchers in the world, and they’re producing genuinely new ideas, not incremental ones.” Bob Mumgaard SM ’15, co-founder and CEO of Commonwealth Fusion Systems, says the region’s expertise in engineering, designing, and manufacturing hardware and access to university researchers made it possible to build the company only in Massachusetts. “The ecosystem has the building blocks… Massachusetts is the strongest in the nation in innovation in energy.”

    President Kornbluth points to quantum: “There isn’t a more important technological field right now than quantum science and technology, and the Boston area has the greatest concentration of quantum talent anywhere in the world.”

  • MIT Forum Explores How AI Reshapes Jobs, Democracy, and Society

    MIT Forum Explores How AI Reshapes Jobs, Democracy, and Society

    At the AI and Society Forum held at MIT, leading researchers from across the Institute gathered to examine the profound ways artificial intelligence is influencing employment, democratic processes, and the very fabric of society. Co-organized by the School of Humanities, Arts, and Social Sciences (SHASS) and the Social and Ethical Responsibilities of Computing (SERC), the event featured keynote talks, panel discussions, and even a musical performance blending generative AI with live instrumentation.

    AI and the Future of Work

    Economist David Autor challenged the widespread fear that AI will simply eliminate jobs. Instead, he argued that the technology’s impact hinges on how it alters the scarcity and value of human expertise. “When I think about how technology interacts with the value of labor, I think about it in terms of how it changes the scarcity of expertise, whether it makes it more valuable or whether it makes it more of a commodity,” Autor said. He emphasized that proactive policies—like worker training, wage insurance, and broader capital ownership—are essential to navigate the coming changes.

    During a panel moderated by Rob Loughlin of McKinsey & Company, MIT experts explored the changing nature of work. Daniela Rus, director of CSAIL, envisioned AI as a collaborative assistant: “I’d like to imagine the robot as your friend and assistant… but the role of the human as the decider, as the person with good judgment, remains super important.” David Mindell added that history shows work constantly evolves, and the key is to “constantly be creating the new work.” Sendhil Mullainathan cautioned that while AI offers productivity gains, long-term growth requires careful differentiation, and we are entering a period of high variance in workforce restructuring.

    Democracy and AI

    The second session turned to AI’s impact on democratic institutions. Chara Podimata of MIT Sloan presented research auditing large language models for bias in election information. Her study of 12 major models during the 2024 U.S. presidential election revealed that responses varied dramatically based on users’ stated demographics and political leanings. A new audit is planned for the 2026 midterms.

    In a panel moderated by Songyee Yoon, experts voiced both concerns and cautious optimism. Bailey Flanigan warned that automating decision-making could strip away the procedural rituals essential to democracy. Charles Stewart III highlighted the risk of AI-induced chaos in elections, noting that “if an election is called into question, that can lead to violence.” Lily Tsai argued that AI designers must embed democratic values such as agency, equality, and mutual respect. She shared a promising example: a Socratic dialogue chatbot that helps people articulate their beliefs, which actually moderated their policy positions.

    A Call for Interdisciplinary Action

    In his opening remarks, SHASS Dean Agustín Rayo stressed that “paying attention to the societal consequences of AI is not a departure from MIT’s mission; it’s a way of ensuring that our technical leadership has maximum impact.” Dan Huttenlocher, dean of the MIT Schwarzman College of Computing, echoed the need for interdisciplinary research to avoid overreliance and unintended consequences. The forum made clear that as AI continues to advance, understanding its societal impacts is as critical as the technology itself.

  • Two LLMs Team Up to Help Robots Interpret Vague Instructions and Prioritize What Matters

    Two LLMs Team Up to Help Robots Interpret Vague Instructions and Prioritize What Matters

    Researchers at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) have developed a new method that uses two large language models (LLMs) to help robots understand ambiguous instructions and focus on key details. The approach, called Masked Inverse Reinforcement Learning (Masked IRL), reduces the amount of demonstration data needed to teach a robot by nearly five times, while improving the robot’s ability to infer unspoken user preferences.

    Traditional robot training often requires either extensive physical demonstrations or detailed written instructions. Masked IRL automates the process: first, one LLM clarifies ambiguous prompts (e.g., turning “stay close” into “stay close to the surface of the table”) by comparing a user’s demonstration trajectory to the shortest possible path. Then a second LLM evaluates the environment and “masks” irrelevant details – such as a person leaning on a table – while highlighting critical ones like obstacles to avoid. The robot then uses these prioritized details to generate a safe motion plan.

    In experiments, the system correctly identified unstated user preferences up to 15 percent more often than comparable baselines. Real-world tests showed a robotic arm successfully moving a coffee mug around a laptop, wiping a table while “staying close” to it, and handing a user a bag of chips while “staying away” from both the person and the table – all after fewer than 50 kinesthetic demonstrations.

    The team plans to enhance Masked IRL with camera input, allowing robots to visually focus on relevant objects in dynamic environments. The work was supported by the Tata Group via the MIT Generative AI Impact Consortium Award and the Department of Defense, and will be presented at the 2026 IEEE International Conference on Robotics and Automation.