Home mit_idePosts

MIT IDE

@mit_ide

The MIT Initiative on the Digital Economy Shaping a brighter digital future💡 Latest news, podcast episodes & events👇
Followers
2,424
Following
155
Account Insight
Score
29%
Index
Health Rate
%
Users Ratio
16:1
Weeks posts
AI agents are quickly moving from assistants to decision-makers in online markets. A recent paper from researchers at the @mit_ide examines what happens when AI agents begin shopping, negotiating, and making purchases on behalf of consumers — and how the design choices being made today could shape future markets. The research, from Benjamin Manning, Peyman Shahidi, and collaborators in IDE Research Lead John Horton’s AI, Marketplaces and Labor Economics research group, explores a key tension emerging in agentic commerce: Will AI agents work for consumers, or for the platforms that deploy them? As Manning notes: “This is a choice a lot of companies need to make soon.” From insurance and housing to financial services and e-commerce, these systems could fundamentally reshape how people navigate markets, compare options, and make decisions online. Read more about the future of AI agents for commerce and what’s at stake as these systems evolve via link in bio 🔗
0 0
1 day ago
We’re excited to announce the winners of the 2026 GenAI Lab, an @mitsloanactionlearning course sponsored by the @mit_ide These exceptional students have demonstrated outstanding innovation and creativity in the field of #GenerativeAI. 🎊 Winners: Susy Liu, Benjamin Plotnik, and Jordan Sandford from team Hunter Water for their project “Development Application Assessment Assistant” Throughout the course, these students tackled real-world challenges, pushing the boundaries of what’s possible with Generative AI. Their projects showcased not only technical prowess but also a deep understanding of the societal impacts of AI technologies. Thank you to GenAI Lab faculty John Horton, Tim Valicenti, and Michiel Bakker, the TA’s, students, mentors, and companies who made this course possible.
0 0
2 days ago
Most people think AI influence starts with the answer. Renée Richardson Gosline argues it starts much earlier — at the prompt. @mit_ide Research Group Lead and @mitsloan Professor @reneegosline explains how AI can shape decision-making through both trust and skepticism. Even people who actively resist AI recommendations still respond to them in measurable ways. A thoughtful look at how AI is starting to shape human judgment in ways most people probably don’t notice day to day.
59 1
3 days ago
Will AI automate work? Probably. Will most jobs become fully automated? @mit researchers say it’s far more complicated than that. New research from the @mit_ide explores why the future of work may not be about replacing humans entirely, but finding the “just right” balance between people and AI. In many cases, the research suggests partial automation may be the real story: AI handling repetitive, scalable tasks while humans focus on judgment, creativity, and decision-making. As MIT IDE Research Scientist Martin Fleming put it: “Partial automation is pervasive.” Read the full blog for insights from Neil Thompson, Martin Fleming, David Autor, and collaborators on how AI may reshape work, expertise, and automation via link in bio 🔗
0 0
4 days ago
AI isn’t just assisting anymore, it’s acting. Sinan Aral (@professorsinan ), @mit_ide Director and Professor of IT and Marketing at @mitsloan and Eric So, MIT IDE Research Group Lead and Professor of Global Economics and Management at MIT Sloan, break down what happens when AI agents don’t wait for permission—and what that means for all of us. Key takeaways: • Agents can make decisions + take action independently • Humans risk stepping out of the loop • The real skill now: knowing what to delegate (and what not to) Who’s in control when AI starts acting on its own? . . . . . . . . . . #ai #aiagents #chatgpt #chatbot #mitsloan
35 0
9 days ago
At BIG.AI@MIT , we spent two days exploring the business, economic and societal impacts of AI with the world's leading thinkers. The results were unanimous: this is the most important technological revolution in human history and steering it requires us to understand it, not (only) from a technical perspective, but more importantly from an economic, sociological, cognitive, and generally socio-technical perspective. Changes from AI will be social and economic. Understanding these aspects of the technology are much harder than understanding it technically.
27 2
13 days ago
Join us on Thursday, April 30 at 11am, for the final @mit_ide Lunch Seminar of the semester with Vartan Shadarevian for a talk on “General Strategic Intelligence: AI Agents for the New Economy.” What happens when AI agents stop acting like tools and start acting like strategic players? Vartan introduces GENSTRAT, a new benchmark designed to test how AI agents behave in complex, unfamiliar economic environments. His research explores where these agents succeed, where they break down, and what that means as they begin to coordinate, compete, and make decisions with increasing autonomy. The implications go far beyond technology, touching market dynamics, regulation, and the future structure of the economy. 🔗 Register via link in bio (virtual or in-person options available)
0 0
23 days ago
What if targeted ads aren’t actually helping you? At a recent @federaltradecommission workshop, Alessandro Acquisti, Professor at @mitsloan and Research Lead at the @mit_ide was asked to testify on the economics of behavioral advertising. His research challenges one of the industry’s core assumptions: that targeted ads are a win-win. After reviewing decades of evidence, the findings suggest: • Little proof that consumers, publishers, or merchants meaningfully benefit • Targeted ads can steer users toward higher prices and lower-quality options • The biggest gains may go to data intermediaries As the FTC builds the case for future privacy enforcement, and states continue advancing their own data privacy laws, these questions are becoming harder to ignore. 🔗Read more on what this could mean for the future of online advertising and data privacy via link in bio
0 0
1 month ago
You might not just work with AI. You might report to it. At BIG.AI@MIT , @ranaelkaliouby shared a vision of organizations becoming hybrids of humans and AI—where management itself could look very different. In a fireside chat with @mit_ide Director @professorsinan , she raised the real challenge: not the technology, but what comes next. How do you build trust, accountability, and clarity in teams where AI plays a central role? Would you work for an AI manager? 👉 Check out the full fireside chat via link in our bio 💻 Visit ide.mit.edu to explore more insights from BIG.AI@MIT
0 0
1 month ago
AI is moving fast. But getting it right is the real challenge. At The Business Implications of Generative AI @ MIT (BIG AI@MIT ) event, leaders across industry and academia dug into what it actually takes to make AI work inside organizations. A few themes that stood out: • AI adoption is a management problem, not a technology problem • Many companies are still in the early “J-curve” dip • Human judgment and oversight are becoming more valuable, not less • Responsible AI isn’t optional—it’s foundational • The future isn’t human vs AI, it’s how well they work together If you’re thinking about how to move from experimentation to real impact, this is worth a read. 🔗 Read the full recap and access event videos via link in bio
0 0
1 month ago
This Thursday, join us for the @mit_ide Lunch Seminar with Emma Wiles, professor @buquestrom and IDE Digital Fellow, for her talk, “AI Agents as Employees.” Emma’s research explores what happens when organizations move beyond using AI as a tool and begin formally embedding AI agents into their org charts, with defined roles and responsibilities. What does it actually mean to treat AI as part of the organization? Find out this Thursday! Register now via link in bio 🔗
0 0
1 month ago
What if better AI outputs don’t actually lead to better outcomes? In a new piece from @mit_ide Michael Schrage, Research Fellow at the IDE, challenges a core assumption in today’s AI conversation: that more advanced models automatically deliver better results. The reality is more nuanced. As AI systems improve, outcomes depend just as much on how people use them—how teams interpret outputs, integrate them into workflows, and adapt decision-making in response. In many cases, the biggest gains come not from the model itself, but from rethinking how work gets done around it. For organizations, that shifts the focus: ➡️ Better outputs ≠ better outcomes ➡️ Human judgment and execution are still the differentiator AI doesn’t replace the need for strategy—it raises the bar for it. Read the full Q&A via link in bio 🔗
0 0
1 month ago