How do you solve a problem like out-of-control AI?
Last week Google revealed it is going all in on generative AI. At its annual I/O conference, the company announced it plans to embed AI tools into virtually all of...
Automating the math for decision-making under uncertainty
One reason deep learning exploded over the last decade was the availability of programming languages that could automate the math — college-level calculus — that is needed to train each...
Solving a machine-learning mystery
Large language models like OpenAI’s GPT-3 are massive neural networks that can generate human-like text, from poetry to programming code. Trained using troves of internet data, these machine-learning models take...
3 Questions: Leo Anthony Celi on ChatGPT and medicine
Launched in November 2022, ChatGPT is a chatbot that can not only engage in human-like conversation, but also provide accurate answers to questions in a wide range of knowledge domains....
Helping companies deploy AI models more responsibly
Companies today are incorporating artificial intelligence into every corner of their business. The trend is expected to continue until machine-learning models are incorporated into most of the products and services...
Efficient technique improves machine-learning models’ reliability
Powerful machine-learning models are being used to help people tackle tough problems such as identifying disease in medical images or detecting road obstacles for autonomous vehicles. But machine-learning models can...
MIT-Takeda Program heads into fourth year with crop of 10 new projects
In 2020, the School of Engineering and Takeda Pharmaceutical Company launched the MIT-Takeda Program, which aims to leverage the experience of both entities to solve problems at the intersection of...
Integrating humans with AI in structural design
Modern fabrication tools such as 3D printers can make structural materials in shapes that would have been difficult or impossible using conventional tools. Meanwhile, new generative design systems can take...
Elon Musk tells Tesla staff he must approve all hiring: report
Tesla Chief Executive Elon Musk has said that the company can make no new hires unless he personally approves them, including contractors, the Information reported on Monday, citing an email to...
Robot armies duke it out in Battlecode’s epic on-screen battles
In a packed room in MIT’s Stata Center, hundreds of digital robots collide across a giant screen projected at the front of the room. A crowd of students in the...
Larger language models do in-context learning differently
Posted by Jerry Wei, Student Researcher, and Denny Zhou, Principal Scientist, Google Research There have recently been tremendous advances in language models, partly because they can perform tasks with strong...
Twitter CEO Linda Yaccarino shared ‘chemistry’ with Elon Musk
Linda Yaccarino’s onstage appearance with Elon Musk at a recent marketing conference in Miami Beach amounted to a job interview that eventually netted her the role of Twitter’s new chief...
Large language models are biased. Can logic help save them?
Turns out, even language models “think” they’re biased. When prompted in ChatGPT, the response was as follows: “Yes, language models can have biases, because the training data reflects the biases...
TikTok parent allowed Chinese Communist Party ‘supreme access’ to US data: suit
The Chinese Communist Party “maintained supreme access” to all data held by TikTok parent ByteDance, including information stored in the US, a former top executive at the company alleged in...
New insights into training dynamics of deep classifiers
A new study from researchers at MIT and Brown University characterizes several properties that emerge during the training of deep classifiers, a type of artificial neural network commonly used for...
