Machine learning

A new proposal by graduate student Shayne Longpre and other AI researchers suggests “a new scheme supported by AI companies that gives outsiders permission to probe their models and a way to disclose flaws publicly,” reports Will Knight for Wired. “The authors suggest three main measures to improve the third-party disclosure process: adopting standardized AI flaw reports to streamline the reporting process; for big AI firms to provide infrastructure to third-party researchers disclosing flaws; and for developing a system that allows flaws to be shared between different providers,” writes Knight. 

Writing for Fast Company, graduate student Sheng-Hung Lee and Devin Liddell of Teague highlight four types of AI technologies that could aid senior citizens in their homes. “To better understand how seniors want AIs and robots to help in their homes, we asked them,” they write.  “We recruited seniors from the MIT AgeLab’s research cohort—each around 70 years old and in the early stages of retirement—and then engaged in wide-ranging conversations about their aspirations and fears about these technologies.”

Research Scientist Eva Ponce, director of online education for the MIT Center for Transportation & Logistics, speaks with Financial Times reporter Rafe Uddin about how companies are shifting toward automation and the impact on employees. “Companies are investing more in upskilling associates… ensuring they’re ready for a new style of work,” says Pone. “More complex tasks will still need to be done by people… These technologies are disruptive. The warehouse of the future is a combination of robotics, sensors and computer vision.” 

Prof. Sara Beery speaks with TechCrunch reporter Kyle Wiggers about the use of AI tools in the advancement of science. “Most science isn’t possible to do entirely virtually — there is frequently a significant component of the scientific process that is physical, like collecting new data and conducting experiments in the lab,” explains Beery. 

Principal Research Scientist Andrew McAfee co-founded Workhelix, a “tech-enabled service startup that works with enterprises to better understand and monitor AI automation at their companies,” reports Rebecca Szkutak for Tech Crunch. “Workhelix breaks down a company’s employee positions into specific job functions and tasks and scores each task for its suitability for AI adoption,” explains Szkutak. “This helps companies build roadmaps for how and where to adopt AI and gives enterprises a way to monitor if the AI they adopted is working.” 

Mohamed Elrefaie speaks with Automotive World reporter Will Girling about his work developing an open-source dataset of 8,000 car designs, including their aerodynamic characteristics, which could be used to develop novel car designs in a more efficient manner. “If an automaker wants to reduce drag and improve performance, it can guide the GenAI model to produce those specific designs,” Elrefaie explains. “The standard development cycle for a design using legacy tools can take anywhere from three to five years, as it requires collaboration between many specialized departments. With AI, you could validate up to 600 designs in just one or two minutes.”

Writing for Forbes, Paula Schneider, CEO of Susan G. Komen, highlights Prof. Regina Barzilay’s research using AI to detect breast cancer. “Using her own mammograms in her research at MIT, Dr. Barzilay demonstrated how AI could have detected her breast cancer much earlier, potentially improving her prognosis,” writes Schneider. “Studies show that incorporating AI into mammogram analysis boosts cancer detection rates by 20%, without increasing false positives. This is a significant leap forward, as early detection is key to a better chance at positive outcomes and survival.” 

Eva Ponce, director of the MIT Center for Transportation & Logistics speaks with Financial Times reporter Rafe Uddin about how companies are integrating automation. “Labor shortages are a persistent theme and this is another driver for this investment,” says Ponce.

Writing for Project Syndicate, Prof. Daron Acemoglu addresses the potential benefits and risks posed by AI advancements. “AI, properly developed and used, can indeed make us better – not just by providing ‘a bicycle for the mind,’ but by truly expanding our ability to think and act with greater understanding, independent of coercion or manipulation,” explains Acemoglu. “Yet owing to its profound potential, AI also represents one of the gravest threats that humanity has ever faced. The risk is not only (or even mainly) that superintelligent machines will someday rule over us; it is that AI will undermine our ability to learn, experiment, share knowledge, and derive meaning from our activities.”

Varun Mohan '17, SM '17 and Douglas Chen '17 co-founded Codeium, an AI-powered coding startup designed to help users write code in a faster and more efficient manner, reports Marina Temkin for TechCrunch. “Codeium tries to distinguish itself from competitors by targeting companies rather than individual developers,” writes Temkin. 

A new study by researchers at MIT has found that “while highly skilled workers reported a 40% surge in performance when artificial intelligence was used within the boundary of its capabilities, overreliance on AI resulted in a performance drop of 19%,” reports Chris Callagher for USA Today. 

Prof. Daron Acemoglu speaks with New York Times reporter Jeff Sommer about the anticipated impact of future AI on various industries. “There is a lot of hype in the industry,” says Acemoglu. While some AI companies have “impressive achievements,” Acemoglu adds that many financial and economic calculations were being based on mere “projections into the future that are sometimes exaggerated.”

Lincoln Lab Senior Scientist Vijay Gadepally speaks with Washington Post reporter Nicolas Rivero about ways to make AI more sustainable. “Whatever we do, energy usage is likely going to go up,” says Gadepally. “That train has left the station.”

Prof. Daron Acemoglu speaks with Bloomberg reporter Jeran Wittenstein about the implications of new AI advancements on areas such as productivity, the labor market and economic growth. “I hope I’m wrong, I hope we get some productivity growth. That would be really cool,” says Acemoglu. “But I don’t see it yet.”

Prof. Max Tegmark speaks with CNBC “Beyond The Valley” podcast hosts Arjun Kharpal and Tom Chitty about concerns surrounding the future of AI systems. “I think, on an optimistic note here, we can have almost everything that we’re excited about with AI,” says Tegmark, “if we simply insist on having some basic safety standards before people can sell powerful AI systems.”