Professor Andrew W. Lo expects AI models to quickly advance and be able to serve as fiduciaries in coming years. As of now, he cautioned, they aren't ready for prime time."We believe that it is possible to train an LLM, just like we train humans, to provide fiduciary duty," Lo said. "But they don't have it right now, and the guardrails aren't there to protect individuals."

In this episode of Vox's "Today, Explained" podcast, professor Deborah Lucas said: "It never makes economic sense to bail out a small, chronically unprofitable enterprise. There's a lot of talk that it would've helped the employees of Spirit Airlines, but a typical bailout really doesn't just help the employees. The main beneficiaries are the debt holders of the company. A lot of bailouts are subsidies to the rich, if you will. They don't necessarily reach the people you're hoping to when you think about bailing out a company in order to save jobs."

"If you keep negotiating, and keep getting just a little more every time, the effect grows and grows," lecturer John Richardson said. It can be uncomfortable for many people to ask for more. "Your body thinks it's life or death. But it isn't. Two things help you calm down. Experience, and preparation. And those are things you can start working on today," he said.

AI may also be causing workers to spend more time on the job because they're still learning how to integrate it into their workflows, said principal research scientist Neil Thompson. "Usually there's a transition period where you have to modify the processes the organization has," he said. "Initially, you become less efficient."

"The U.S. does not manufacture most of the advanced chips. It requires other inputs for infrastructure investment and it requires global data. Hence, the AI industry is dependent upon global trade," said Institute Professor Daron Acemoglu. "I don't think we have any chance of doing these things ourselves."

Companies betting against entry-level Gen Z talent by automating their roles may be making a costly long-term mistake. That's the warning from MIT research scientist Andrew McAfee. "How else are people going to learn to do the job except via on-the-job learning and training apprenticeship?" he said.

According to a study by principal research scientist Neil Thompson and co-authors, leading AI models in mid-2024 successfully completed 50 percent of white-collar tasks that would have taken a human three to four hours to complete; just over a year later, they completed 65 percent. The authors estimate AI systems will be able to complete 80 to 95 percent of text-based tasks by 2029. "This pace of improvement isn't quite as fast as what we've seen with AI and coding," research scientist and co-author Matthias Mertens said. "But it's still really, really fast."

Keeping California's last operating nuclear station, Diablo Canyon, running from 2030 to 2045 could offer net savings of capital and operating costs totaling more than $7.6 billion, or more than $500 million per year of continued operations, according to a new analysis by senior lecturer John Parsons, a research affiliate with the MIT Center for Energy and Environmental Policy Research.

Senior lecturer Robert Pozen wrote: "The lesson from the telecom debacle is that financial engineering can obscure, for years, the difference between real customer demand and demand driven by incentives. When AI companies begin to finance their own product distribution, guaranteeing returns to investors and subsidizing sales, it's a signal for investors to dig deeper."

Erin Dawicki (Sloan Fellows: MBA '24) founder of LymeAlert, the first at-home rapid tick testing kit, said: "Tick populations are expanding, making them more prevalent in areas where they weren't as common before. You can pick up ticks anywhere there is grass and trees, which can include places such as Central Park in New York and Boston Common."

A recent field experiment by associate professor Jackson Lu and co-authors randomly assigned 250 employees at a technology consulting firm in China to either use ChatGPT to assist with their work or to work without it. The employees with ChatGPT access were judged as significantly more creative by both their supervisors and outside evaluators. But the gains showed up exclusively among employees with strong metacognitive strategies.