News
Entertainment
Science & Technology
Life
Culture & Art
Hobbies
News
Entertainment
Science & Technology
Culture & Art
Hobbies
- home
›- MIT News - Artificial intelligence
›- Software
MIT News - Artificial intelligence - Software
8 | Follower
Advancing urban tree monitoring with AI-powered digital twins | MIT News | Massachusetts Institute of Technology
The new "Tree-D Fusion" system integrates generative AI and genus-conditioned algorithms to create precise simulation-ready models of 600,000 existing urban trees across North America.
MIT researchers develop an efficient way to train more reliable AI agents | MIT News | Massachusetts Institute of Technology
MIT researchers developed an efficient approach for training more reliable reinforcement learning models, focusing on complex tasks that involve variability. This could enable the leverage of reinforcement learning across a wide range of applications.
Can robots learn from machine dreams? | MIT News | Massachusetts Institute of Technology
LucidSim is an AI-powered simulator that trained a robot dog to perform parkour using generated images without any real-world data. This approach, from MIT CSAIL researchers, scales up training data, helping robots transfer skills to the real world without additional fine-tuning.
Four from MIT named 2025 Rhodes Scholars | MIT News | Massachusetts Institute of Technology
Four from MIT — Yiming Chen ’24, Wilhem Hector, Anushka Nair, and David Oluigbo — have been selected as 2025 Rhodes Scholars and will begin fully funded postgraduate studies at Oxford University in the U.K. next fall.
A causal theory for studying the cause-and-effect relationships of genes | MIT News | Massachusetts Institute of Technology
MIT researchers developed theoretical foundations for methods that could identify the best way to aggregate genes into modules and efficiently learn the underlying cause-and-effect relationships among them. This approach holds promise for investigating the mechanisms of diseases and identifying new drug targets.
A portable light system that can digitize everyday objects | MIT News | Massachusetts Institute of Technology
The portable light system and design tool “PortaChrome” uses UV and RGB LEDs to activate photochromic dye, reprogramming everyday objects like shirts. The MIT CSAIL researchers' software can help users turn items into multicolor displays of fashion designs and health data.
Nanoscale transistors could enable more efficient electronics | MIT News | Massachusetts Institute of Technology
Nanoscale 3D transistors made from ultrathin semiconductor materials can operate more efficiently than silicon-based devices, leveraging quantum mechanical properties to potentially enable ultra-low-power AI applications.
Empowering systemic racism research at MIT and beyond | MIT News | Massachusetts Institute of Technology
Researchers in the cross-disciplinary MIT Initiative on Combatting Systemic Racism are building an open data repository to advance research on racial inequity in domains like policing, housing, and health care.
Artist and designer Es Devlin awarded Eugene McDermott Award in the Arts at MIT | MIT News | Massachusetts Institute of Technology
Es Devlin is the recipient of the 2025 Eugene McDermott Award in the Arts at MIT. The $100K prize includes an artist residency at MIT in spring 2025, during which Devlin will present her work in a public lecture May 1, 2025.
Despite its impressive output, generative AI doesn’t have a coherent understanding of the world | MIT News | Massachusetts Institute of Technology
Large language models can achieve incredible performance on some tasks without having internalized a coherent model of the world or the rules that govern it, MIT researchers find. This means these models are likely to fail unexpectedly if they are deployed in situations where the environment or task slightly changes.
MIT Schwarzman College of Computing launches postdoctoral program to advance AI across disciplines | MIT News | Massachusetts Institute of Technology
MIT Schwarzman College of Computing launched the Tayebati Postdoctoral Fellowship Program to support leading postdocs to bring cutting-edge AI to bear on research in scientific discovery or music.
Making it easier to verify an AI model’s responses | MIT News | Massachusetts Institute of Technology
A new system helps human fact-checkers validate the responses generated by a large language model. By speeding validation time by 20 percent, the system could improve manual verification and help users spot errors in AI models deployed in real-world situations.
Combining next-token prediction and video diffusion in computer vision and robotics | MIT News | Massachusetts Institute of Technology
Diffusion Forcing, a method led by researchers at MIT CSAIL, can train a neural network to sort corrupted data while anticipating next steps. It can make flexible plans for robots, generate high-quality video, and help AI agents navigate digital environments.
Equipping doctors with AI co-pilots | MIT News | Massachusetts Institute of Technology
Ambience Healthcare, founded by MIT alumni, automates documentation for clinicians before, during, and after patient visits. The founders say clinicians using Ambience save time on documentation, report lower levels of burnout, and develop higher-quality relationships with their patients.
Modeling relationships to solve complex problems efficiently | MIT News | Massachusetts Institute of Technology
MIT Associate Professor Julian Shun develops high-performance parallel graph algorithms and frameworks that improve the efficiency of online recommendation systems, transportation routing platforms, and financial fraud detectors.
Q&A: A new initiative to help strengthen democracy | MIT News | Massachusetts Institute of Technology
Professor David Singer, head of the MIT Department of Political Science, outlines a new Strengthening Democracy Initiative to advance rigorous evidence-based analysis of democratic resiliency, with a research focus on election administration, public opinion, and political participation.
AI simulation gives people a glimpse of their potential future self | MIT News | Massachusetts Institute of Technology
“Future You” is a generative AI tool that enables users to have a simulated conversation with a potential version of their future selves. The chatbot is meant to reduce users’ anxiety, improve positive emotions, and guide them toward making better everyday choices.
AI pareidolia: Can machines spot faces in inanimate objects? | MIT News | Massachusetts Institute of Technology
MIT CSAIL researchers have introduced a human-labeled dataset of pareidolic faces found in everyday objects, revealing key differences between human and AI face detection and discovering a mathematical "Goldilocks zone" where pareidolia is most likely to occur.
New security protocol shields data from attackers during cloud-based computation | MIT News | Massachusetts Institute of Technology
MIT researchers developed a technique guaranteeing that data remain secure during multiparty, cloud-based computation. This method, which leverages the quantum properties of light, could enable organizations like hospitals or financial companies to use deep learning to securely analyze confidential patient or customer data.
3 Questions: Should we label AI systems like we do prescription drugs? | MIT News | Massachusetts Institute of Technology
To mitigate potential harms of AI when deployed in high-stakes health care applications, MIT and BU researchers propose responsible-use labels for these systems, similar to the FDA-mandated labels placed on prescription medications.
Fifteen Lincoln Laboratory technologies receive 2024 R&D 100 Awards | MIT News | Massachusetts Institute of Technology
MIT Lincoln Laboratory received 15 R&D 100 Awards for 2024. The "Oscars of Innovation" recognized the technologies for their ability to map the ocean floor and the brain, prevent heat stroke and cognitive injury, expand AI processing and quantum system capabilities, and introduce new fabrication approaches.
New AI JetPack accelerates the entrepreneurial process | MIT News | Massachusetts Institute of Technology
The MIT Entrepreneurship JetPack is a generative artificial intelligence tool that helps students navigate the 24-step Disciplined Entrepreneurship process developed by Trust Center’s managing director Bill Aulet.
AI model can reveal the structures of crystalline materials | MIT News | Massachusetts Institute of Technology
MIT chemists developed a generative AI model to help determine structures of powdered crystal materials. This could help researchers characterize materials for use in batteries, magnets, and many other applications.
Study: AI could lead to inconsistent outcomes in home surveillance | MIT News | Massachusetts Institute of Technology
MIT researchers find large language models apply different norms when deciding whether similar home surveillance videos show a crime and whether the police should be called. The models often recommend calling police even when they state no crime occurred in the video.
Enhancing LLM collaboration for smarter, more efficient solutions | MIT News | Massachusetts Institute of Technology
“Co-LLM,” an algorithm developed at MIT CSAIL, uses a base large language model to begin crafting a response to a prompt, and calls upon an expert large language model at specific tokens to improve accuracy. A switch variable enables this token-level collaboration.
A fast and flexible approach to help doctors annotate medical scans | MIT News | Massachusetts Institute of Technology
The interactive segmentation tool “ScribblePrompt” allows users to scribble, click, and use bounding boxes to delineate anatomical structures in biomedical images. This manual annotation algorithm was trained on synthetic data that simulated how humans interact with medical scans.
Study: Transparency is often lacking in datasets used to train large language models | MIT News | Massachusetts Institute of Technology
The Data Provenance Explorer can help machine-learning practitioners make more informed choices about the data they train their models on, which could improve the accuracy of models deployed in the real world.
A framework for solving parabolic partial differential equations | MIT News | Massachusetts Institute of Technology
A new algorithm solves complicated partial differential equations on triangle meshes by breaking them down into simpler problems, potentially guiding complex modeling and analysis in computer graphics and geometry processing.