Prof. Jacob Andreas explored the concept of language guided program synthesis at CSAIL’s Imagination in Action event, reports research affiliate John Werner for Forbes. “Language is a tool,” said Andreas during his talk. “Not just for training models, but actually interpreting them and sometimes improving them directly, again, in domains, not just involving languages (or) inputs, but also these kinds of visual domains as well.”
Prof. Daniela Rus, director of CSAIL, emphasizes the central role universities play in fostering innovation and the importance of ensuring universities have the computing resources necessary to help tackle major global challenges. Rus writes, “academia needs a large-scale research cloud that allows researchers to efficiently share resources” to address hot-button issues like generative AI. “It would provide an integrated platform for large-scale data management, encourage collaborative studies across research organizations, and offer access to cutting-edge technologies, while ensuring cost efficiency,” Rus explains.
Writing for Forbes, research affiliate John Werner spotlights Prof. Stefanie Mueller’s presentation at the CSAIL Imagination in Action event on her work developing a new type of paint that allows users to change the color and pattern of different objects. “The long-term vision here, really, is to give those physical objects the same capabilities as we have in digital,” said Mueller. “I hope in the future we will all get some free stuff, and we would just have an [app] where we can download different textures we can apply, and change our outfits.”
During her talk at CSAIL’s Imagination in Action event, Prof. Daniela Rus, director of CSAIL, explored the promise of using liquid neural networks “to solve some of AI’s notorious complexity problems,” writes research affiliate John Werner for Forbes. “Liquid networks are a new model for machine learning,” said Rus. “They're compact, interpretable and causal. And they have shown great promise in generalization under heavy distribution shifts.”
A study by researchers from MIT and Stanford University explored how customer service agents use generative AI chatbots in their response to tech support questions, reports Meghan McCarty for Marketplace. “The most experienced workers see no to even small negative effects of the AI being turned on for them, while the less-experienced and least-productive workers see on average 30 to 35% improvements in productivity,” explains graduate student Lindsey Raymond.
Prof. Regina Barzilay speaks with Nicole Estephan of WCVB-TV’s Chronicle about her work developing new AI systems that could be used to help diagnose breast and lung cancer before the cancers are detectable to the human eye.
In conversation with Matthew Huston at Science, Prof. John Horton discusses the possibility of using chatbots in research instead of humans. As he explains, a change like that would be similar to the transition from in-person to online surveys, “"People were like, ‘How can you run experiments online? Who are these people?’ And now it’s like, ‘Oh, yeah, of course you do that.’”
Researchers from MIT have found that using generative AI chatbots can improve the speed and quality of simple writing tasks, but often lack factual accuracy, reports Richard Nieva for Forbes. “When we first started playing with ChatGPT, it was clear that it was a new breakthrough unlike anything we've seen before,” says graduate student Shakked Noy. “And it was pretty clear that it was going to have some kind of labor market impact.”
Prof. Daron Acemoglu and graduate student Todd Lensman have created “the first economic model of how to regulate transformative technologies,” like artificial intelligence, reports Tim Fernholz for Quartz. “Their tentative conclusion is that slower deployments is likely better, and that a machine learning tax combined with sector-specific restrictions on the use of the technology could provide the best possible outcomes,” writes Fernholz.
Writing for Forbes, Prof. Daniela Rus, director of CSAIL, makes the case that liquid neural networks “offer an elegant and efficient computational framework for training and inference in machine learning. With their compactness, adaptability, and streamlined computation, these networks have the potential to reshape the landscape of artificial intelligence and drive further breakthroughs in the field.”
Wall Street Journal reporter Emily Bobrow spotlights Laurel Braitman PhD '13 for her work teaching writing and communication skills to healthcare workers. “We need people who are trained in science and medicine to be able to tell stories about what matters in public health in a way that makes people listen,” says Braitman. “But to do that, they have to be in touch with what they really feel.”
Researchers at MIT have developed PIGINet (Plans, Images, Goal and Initial facts), a neural network designed to bring task and motion planning to home robotics, reports Brian Heater for Tech Crunch. “The system is largely focused on kitchen-based activities at present. It draws on simulated home environments to build plans that require interactions with various different elements of the environment, like counters, cabinets, the fridge, sinks, etc,” says Heater.
Prof. Max Tegmark speaks with Guardian reporter Steve Rose about the potential of artificial intelligence. “The positive, optimistic scenario is that we responsibly develop superintelligence in a way that allows us to control it and benefit from it,” says Tegmark. “If we can build and control superintelligence, we can quickly go from being limited by our own stupidity to being limited by the laws of physics. It could be the greatest empowerment moment in human history.”
MIT Schwarzman College of Computing Dean Daniel Huttenlocher discusses how artificial intelligence has impacted print media at the Aspen Ideas Festival, reports John Frank for Axios. “Most of us grew up in a world where the word print was something that was authoritative,” says Huttenlocher, of how people will need to be on the lookout for misinformation.
MIT Schwarzman College of Computing Dean Daniel Huttenlocher speaks at the Aspen Ideas Festival on how to regulate AI while maximizing its positive impact, reports NBC. “I think when we think about regulation [of artificial intelligence] we need to think about this in the ways we’ve traditionally thought about things – risk, reward, tradeoffs – and that tends to be domain specific,” says Huttenlocher. “It’s hard to have sort of an abstract notion of this new technology and what the risk [and] reward is across all domains.”