Top 10 AI News
MIT in the media: 2025 in review
MIT community members made headlines with key research advances and their efforts to tackle pressing challenges ([news.mit.edu](https://news.mit.edu/2025/mit-media-review-1222#:~:text=MIT%20in%20the%20media%3A%202025,in%20review)). This year MIT researchers appeared in news worldwide for projects ranging from climate solutions to cutting-edge AI. Read full story
Guided learning lets “untrainable” neural networks realize their potential
Researchers at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) showed that even neural networks considered “untrainable” can learn effectively when guided by another network’s built-in biases ([news.mit.edu](https://news.mit.edu/2025/guided-learning-lets-untrainable-neural-networks-realize-their-potential-1218#:~:text=Guided%20learning%20lets%20%E2%80%9Cuntrainable%E2%80%9D%20neural,networks%20realize%20their%20potential)). Using a new “guidance” method, the team enabled these networks to succeed at tasks they could not learn on their own. This finding could broaden the capabilities of neural nets in AI applications. Read full story
A new way to increase the capabilities of large language models
MIT-IBM Watson AI Lab researchers developed an expressive neural architecture that improves large language models (LLMs) by providing better state tracking and sequential reasoning over long texts ([news.mit.edu](https://news.mit.edu/2025/new-way-to-increase-large-language-model-capabilities-1217#:~:text=MIT,in%20LLMs%20over%20long%20texts)). This approach allows LLMs to handle longer or more complex inputs (like passages of text) more accurately, potentially making AI language tools more powerful. Read full story
A “scientific sandbox” lets researchers explore the evolution of vision systems
An AI-powered “scientific sandbox” tool was introduced for exploring how vision systems evolve. By simulating virtual organisms with evolving vision sensors, researchers can identify principles for better camera and sensor design. This could inform improvements in robotics and autonomous vehicles, as the tool lets scientists test visual systems under varying conditions and evolutionary pressures. Read full story
“Robot, make me a chair”
An AI-driven system was developed that allows users to simply describe a multicomponent object in words and have a robot build it. Called “Text2Assembly,” the system interprets a text description (e.g., “make me a chair”) and autonomously constructs the corresponding object out of modular parts. This demonstrates how natural language AI can directly interface with robotics to fabricate custom designs on demand. Read full story
3 Questions: Using computation to study the world’s best single-celled chemists
MIT biology professor Yunha Hwang combines computational methods with genetic research to study microbes, the “world’s best single-celled chemists.” Her work uses AI and genomic analysis to uncover biochemical pathways in microorganisms, reflecting MIT’s growing commitment to interdisciplinary projects at the nexus of genetics and AI. Read full story
Working to eliminate barriers to adopting nuclear energy
Doctoral student Dauren Sarsenbayev is using AI-driven modeling to tackle nuclear waste challenges. His research focuses on creating advanced models to improve nuclear fuel cycles, aiming to remove long-standing safety and efficiency bottlenecks in nuclear power. By applying computational methods, the work could make nuclear energy a more practical option globally. Read full story
Deep-learning model predicts how fruit flies form, cell by cell
MIT researchers developed a deep neural network that can predict cellular movements during the early development of fruit flies. By modeling cell-by-cell dynamics in embryonic growth, this AI-driven approach could be extended to study human organ development and identify early disease mechanisms. The technique showcases AI’s potential in computational biology. Read full story
Enabling small language models to solve complex reasoning tasks
MIT researchers introduced a “self-steering” system called DisCIPL, which allows multiple smaller language models (LLMs) to work together on tasks with constraints (like planning itineraries within budgets). The model orchestrates these LLM agents so they specialize and collaborate, enabling complex reasoning tasks without relying on one huge model. This advances use of lightweight AI models for real-world problems. Read full story
New MIT program to train military leaders for the AI age
MIT launched a certificate program to equip U.S. Naval officers with AI skills needed for modern challenges. This initiative reflects the military’s growing emphasis on AI, where graduates learn to apply data science and machine learning to defense problems. The goal is to prepare leaders in national security to leverage AI in strategy and operations. Read full story
New method improves the reliability of statistical estimations
MIT scientists developed a technique to assess the trustworthiness of statistical estimates. The method helps researchers in fields like economics and public health determine whether results are reliable. By using this approach, people can gauge when experimental findings are solid or likely due to random variation, improving the robustness of scientific conclusions. Read full story
New materials could boost the energy efficiency of microelectronics
Researchers described a novel approach to improve microchip efficiency by stacking multiple active components made of new materials on the chip’s back end. This prototype reduces wasted energy during computation in things like memory and logic units. The innovation could lead to greener, faster electronics by shrinking the energy cost of processing. Read full story
MIT affiliates named 2025 Schmidt Sciences AI2050 Fellows
Ten MIT affiliates, including professors and alumni, were named AI2050 Fellows by Schmidt Futures. Each will receive support to pursue “moonshot” AI research addressing difficult societal and scientific challenges. The fellows will investigate areas like AI alignment, natural language processing, and physics-informed machine learning over the coming year. Read full story
MIT researchers “speak objects into existence” using AI and robotics
MIT scientists combined 3D generative AI with robotic assembly so that objects can be created on demand from speech. Users describe an object verbally, the AI generates a 3D model, and robots then fabricate it. This “speech-to-reality” system demonstrates a new way to merge natural language understanding with manufacturing. Read full story
Robots that spare warehouse workers the heavy lifting
Pickle Robot Company, founded by MIT alumni, unveiled autonomous robots that can load and unload trucks in warehouses and logistics centers. These green robots use suction to lift and move boxes, automating a task that is physically strenuous for workers. The innovation could reduce injuries and increase efficiency in shipping operations. Read full story
US stocks rise after turbulent 2025, AI worries subside
Wall Street closed the year on a high note, as US stock indexes reached record levels ([www.forbes.com](https://www.forbes.com/sites/aliciapark/2025/12/25/ai-minted-more-than-50-new-billionaires-in-2025/#:~:text=AI%20Minted%20More%20Than%2050,New%20Billionaires%20In%202025)). Despite volatility earlier in 2025 due to trade tariffs and interest rate concerns, investor sentiment improved over the year. Movers included technology and AI-related stocks, which rebounded late in the year after regulatory outlooks became clearer. Read full story
Alphabet to buy Intersect to power AI operations
Google’s parent company, Alphabet Inc., announced plans to acquire Intersect, a data-center energy specialist, for $4.75 billion ([www.forbes.com](https://www.forbes.com/sites/aliciapark/2025/12/25/ai-minted-more-than-50-new-billionaires-in-2025/#:~:text=AI%20Minted%20More%20Than%2050,New%20Billionaires%20In%202025)). This move is aimed at securing more renewable energy to meet the rapidly growing electricity demands of artificial intelligence computing at Google data centers. Intersect will help Alphabet expand its clean power capacity for AI services. Read full story
Starlink in the crosshairs: Russia’s concerns over Musk’s satellites
Analysis reports that Russia may target SpaceX’s Starlink satellite network as a strategic move. Starlink, owned by Elon Musk, provides satellite internet and has played roles in Ukraine’s defense. Experts warn that Russia could attempt jamming or attacking Starlink satellites, which are key to communications and potentially to AI-powered navigation. The story explores “how Russia could attack Elon Musk’s conquering of space.” Read full story
AI deepfakes pose increasingly serious cyberbullying threat
Schools are grappling with AI-generated “deepfake” images being used for cyberbullying. One case in Louisiana involved students sharing AI-made nude images of a classmate. The situation escalated into physical altercations and discipline issues. Experts warn that as AI image generation improves, educators must develop new policies to address this emerging form of harassment. Read full story
Girl expelled after fighting classmates who spread AI nude images
An AP News investigation reports on a middle school student who was suspended after a physical altercation, despite being the target of classmates who shared AI-generated fake nude images of her on Snapchat. The incident highlights policy gaps: current rules against sexting do not neatly apply to AI content, and schools are unsure how to discipline such behavior. The story illustrates the legal and ethical challenges of AI in schools. Read full story
Trump Media to merge with nuclear fusion company that wants to power AI
AP News reports that Trump Media & Technology Group plans to merge with a startup developing compact nuclear fusion reactors. The fusion company’s goal is to provide cheap, abundant energy — potentially to power data centers for AI computing. The merger follows investor Devin Nunes bringing Trump Media to Wall Street, and could position the group in the emerging market for fusion-powered AI infrastructure. Read full story
Wall Street ticks higher after Fed update; AI stocks stabilize
After a new inflation report, U.S. stock markets finished higher on December 18, 2025. Tech companies, including AI chipmaker Micron, led gains and helped reverse a recent slide in AI-related stocks. Investors appeared relieved by any sign that the Fed might pause interest rate hikes, even as economic uncertainty remains. Read full story
Wall Street plunges as AI stocks slump, marking worst day in weeks
Major U.S. indexes fell sharply after dropping technology and AI-related stocks hit bottom. Concerns about rising interest rates and slower tech demand dragged markets to their worst performance in nearly a month. The slide was partly blamed on investor jitters over potentially overheated AI valuations. Read full story
Texas universities review courses on race, gender using AI
In compliance with state directives, some Texas colleges are employing AI tools to scan course curricula for content on race and gender. The software flags potential violations of new state laws limiting academic material on these topics. The initiative has sparked debate over academic freedom, as administrators and professors debate whether AI can fairly judge complex course content. Read full story
Militant groups are experimenting with AI, and the risks are expected to grow
A report reveals that militant organizations like ISIS are beginning to use AI tools—such as generating propaganda videos with deepfake technology. Security experts warn this new use of AI makes disinformation and radicalization efforts more potent. The story outlines how freely available AI tools can be weaponized by extremist groups, raising global security concerns. Read full story
Merriam-Webster’s 2025 word of the year is ‘slop’
The Merriam-Webster dictionary named “slop” (meaning sloppy or imprecise work) as its 2025 Word of the Year. The term’s popularity surged with people criticizing inaccuracy and “sloppiness” in various contexts, including media and politics. Its selection reflects cultural concerns over precision and misinformation. Read full story
Tumbling tech stocks drag Wall Street to worst day in 3 weeks
Global stock markets slumped on concerns that the recent rally in tech and AI stocks was overdone. The S&P 500 and Nasdaq dipped as investors worried that a potential bubble in artificial intelligence-related companies was forming. This pullback erased some recent gains and reminded analysts of the volatility still in the market. Read full story
Trump signs executive order to block state AI regulations
President Trump signed an executive order preempting proposed state-level regulations of artificial intelligence. The move prohibits states from enacting their own AI rules for three years, aiming to create a uniform federal approach. The order is part of broader GOP efforts to remove what they see as regulatory obstacles to technology development. Read full story
David Thewlis: AI threatens all human creativity
In an AP Television News interview, actor David Thewlis (famous for playing Professor Lupin in Harry Potter) warns that AI challenges human creativity. He argues that as AI systems generate art, music and writing, society must consider how to preserve genuine creativity and artistry. Thewlis emphasizes that human-driven creativity is “threatened” by machines that can imitate it. Watch video
El Salvador teams up with Elon Musk’s xAI to bring AI to 5,000 public schools
El Salvador and Elon Musk’s AI company xAI announced a partnership to introduce AI educational tools to 5,000 public schools. The plan includes providing students and teachers access to AI tutoring and learning platforms. This initiative is one of the first nationwide educational uses of generative AI, aiming to improve learning outcomes through technology. Read full story
Time names “Architects of AI” as Person of the Year
Time magazine announced that its Person of the Year for 2025 is “The Architects of AI,” referring to key individuals driving artificial intelligence innovation. The winners include tech CEOs, startup founders, and researchers who have significantly shaped AI this year. The AP summary notes that the decision reflects AI’s pervasive impact on global society and economy. Read full story
AP’s 2025 Breakthrough Entertainers on AI and Hollywood
Forbes explores how emerging entertainers (actors, directors, storytellers) are addressing AI in Hollywood. The article highlights five rising stars and producers (the 2025 Breakthrough Entertainers) and their perspectives on artificial intelligence’s role in the film industry. Topics include using AI for special effects, creating movies, and handling AI ethics on set. Watch video
Technological Disruption: Strategic Inflection Points from 2026–2036
This Forbes analysis predicts that from 2026–2036, the convergence of AI, quantum computing, connectivity, and cyber-physical integration will transform society. Author Chuck Brooks discusses how these technological “inflection points” will reshape industries, geopolitics, and daily life. The article outlines expected milestones in areas like AI ethics, space exploration, and climate tech over the next decade. Read full story
Silicon and the Business Context of AI
Forbes contributor John Werner highlights hardware advances in AI, such as Cerebras’s giant chip, and notes how these developments reflect a shift beyond Moore’s Law. The article explains how innovations in silicon and chip design in 2025 have made AI more energy-efficient and accessible, enabling businesses to deploy advanced AI without the previous limitations. Read full story
Make 2026 The Year You Actually Learn AI
Forbes technology contributor Ron Schmelzer argues that 2026 should be the year individuals invest time learning AI. He explains that AI skills will yield long-term benefits across careers. The piece outlines practical steps to begin learning AI and machine learning fundamentals, from online courses to experimenting with AI tools. Read full story
How 2025 Recalibrated AI’s Models Race
Gerui Wang has analyzed how priorities in the AI model race shifted in 2025. Rather than just focusing on scale, model developers moved toward efficiency, reliability, and integration. The article discusses benchmarks where smaller, more efficient models made gains and how this trend is reshaping business, science, and everyday work with AI. Read full story
The Holiday Dissonance: Why AI’s Real Gift Isn’t Efficiency—It’s Sanity
Forbes contributor Sahar Hashmi writes that during the holiday season, AI’s greatest benefit is reducing stress. Using AI tools for tasks like gift-shopping or planning gives people “permission to protect their sanity.” The article encourages readers to let AI handle routine tasks this season, freeing up time to enjoy family and rest. Read full story
How To Rethink Your Job As AI Agents Reshape Work
AI agents are beginning to automate routine tasks in many professions. Forbes author Bernard Marr advises workers on adapting: focus on uniquely human skills and higher-value work. He suggests reprioritizing roles around creativity, strategy, and interpersonal work that AI cannot easily replicate. The article provides a guide to future-proofing careers in the AI era. Read full story
The $40B Healthcare AI Failure—and the EMR Divide Sabotaging Progress
Forbes reports that despite nearly $40 billion invested in AI for healthcare, few projects have succeeded. Much of the failure is blamed on the “EMR divide”—the lack of standardized electronic medical records. Without interoperable data, advanced AI tools struggle to gain traction. The article highlights that only ~4% of healthcare organizations see real AI impact, often due to fragmented records and workflow silos. Read full story
Google Notebook Rolls Out Lecture Mode With British Flair
Google’s AI study app NotebookLM introduced a “lecture mode” that turns long-form text into interactive audio summaries. The mode provides automatically generated overviews read by a British-accented voice. This upgrade transforms passive document review into an audio-visual learning experience. The feature showcases how large language models can create engaging, podcast-like study materials from user content. Read full story
AI Minted More Than 50 New Billionaires In 2025
Forbes data analysis shows a surge of new billionaires whose wealth comes from AI-driven industries. Record investment in AI this year led to dozens of startup founders and tech executives joining the billionaire ranks ([www.forbes.com](https://www.forbes.com/sites/aliciapark/2025/12/25/ai-minted-more-than-50-new-billionaires-in-2025/#:~:text=AI%20Minted%20More%20Than%2050,New%20Billionaires%20In%202025)). The article profiles several such individuals and examines how AI venture funding and public market gains fueled their fortunes. Read full story
Is The AI Job Apocalypse Real Or Overhyped?
Fears abound that AI will eliminate jobs, but Forbes contributor Bernard Marr contends the situation is more nuanced. While AI will automate tasks, it also creates new roles. In many organizations, the AI “apocalypse” has been overhyped; most companies are augmenting human work rather than replacing it entirely. The article argues that workers should adapt by gaining AI-related skills to stay relevant. Read full story
The Robot Olympics Will Have Zero Sports. Here’s Why
The first “Robot Olympics” event will not include traditional sports. Instead, it features competitions in robotics tasks like warehouse sorting and drone rescue missions. In this analysis, John Koetsier explains how real-world utility, not athleticism, is the focus—robots compete in challenges that demonstrate practical skills in manufacturing, search-and-rescue, and logistics. This showcases how robots are becoming useful in everyday tasks, not sport. Read full story
Seeing The Lay Of The Land With AI
Open-source tools like the Clay platform use AI to automate detailed Earth mapping. Forbes contributor John Werner discusses large Earth-observation models (LEOMs) that can rapidly process satellite imagery. By training on massive geospatial datasets, these models enable quick terraforming analyses for research and planning. This democratizes planetary mapping, making geographic insights available for agriculture, disaster response, and science. Read full story
What Small Businesses Should Know Ahead Of 2026
As 2026 approaches, Forbes contributor John Hall outlines key trends for small businesses. Topics include ongoing AI adoption, cybersecurity, and supply chain changes. He advises small business owners to prepare for increased AI integration in customer service and operations, and to stay agile amid tech-driven market shifts. The article offers tactical steps (like leveraging AI marketing tools) for SMBs to thrive in the coming year. Read full story
Pixel 10’s GPU Fix: Gaming Speeds Up 18%, But There’s A Catch
Google’s new phone, the Pixel 10, received a behind-the-scenes graphics driver update. Benchmarks show the Pixel 10 now runs the game Genshin Impact up to 18% faster ([www.forbes.com](https://www.forbes.com/sites/iainmartin/2025/12/22/open-source-project-with-little-revenue-in-talks-to-raise-at-least-160-million/#:~:text=Million%20www,tech%20that%20can%20make%20AI)) (and a 26% boost in one other game). The catch is that the stability is not perfect—some games crash and frame rates can fluctuate. This story reveals how a software fix can dramatically improve mobile gaming performance, albeit with limitations. Read full story
Stop Wasting Money On AI Agents: 5 Rules for Choosing Use Cases
Consultant Bernard Marr advises companies to be strategic when implementing AI agents. He outlines five rules, such as focusing on repetitive tasks and ensuring data readiness, to avoid common pitfalls. The article emphasizes that not every problem needs an AI agent and that organizations should align AI projects with clear business value to avoid wasteful spending. Read full story
Can Your Future Self Make You Happier?
This article explores how AI can boost personal well-being by focusing on “future selves.” Tools that connect people’s goals to their future selves (like reimagining future achievements) can increase motivation and happiness. Author John Werner suggests that AI-enabled journaling and reflection apps can strengthen a person’s sense of purpose by highlighting long-term progress on personal stories and goals. Read full story
ChatGPT’s App Store Moment Has Arrived
OpenAI introduced an “App Store” for ChatGPT, allowing developers to create and sell apps (plugins) built on ChatGPT. Forbes reports this marks the shift of ChatGPT from a standalone chatbot to a platform with an ecosystem of applications. Users can now add features like web browsing or image generation through third-party plugins, greatly extending ChatGPT’s functionality and distribution. Read full story
AI Analyzes My 2025 ChatGPT Usage: Top 10 Use Cases & One Insight
A Forbes tech staffer used an AI to analyze his personal ChatGPT usage over 2025. The AI identified his top 10 use cases (such as drafting emails, coding help, learning topics) and concluded that ChatGPT became a “reasoning engine” rather than just a search tool. It showed how integrated the assistant was in his workflow, offering quantitative insight into personal productivity. Read full story