Marco Tedesco has worked with artificial intelligence since the late 1990s, but he does not sound like a salesman. A research professor at Columbia Climate School's Lamont-Doherty Earth Observatory, he has watched AI's slow march into every corner of science and now wants to slow the conversation down before habits harden. About a year ago, he approached Lamont leadership about creating an AI working group-a shared space where researchers could examine the technology's potential without surrendering their judgment, ethics, or curiosity.
"The problem is not whether AI is going to be here," Tedesco said. "AI is already here. The problem is how we are going to use it, and how we are going to continue building it."
AI's promise and peril
For Tedesco, AI is a tool, not a force of nature. He sees it shaped by the economic, political, and environmental systems that produce it. "To me, AI can be a manifestation of evil," he said, "but it is not evil itself." The working group he set in motion is not anti-AI. It seeks to prevent researchers from drifting into passive reliance on systems they do not understand.
The AI Learning Path for Research Scientists is one example of the kind of structural guidance researchers need to use these tools critically. Tedesco's concern is that without that critical distance, scientists may trade their own analytical rigor for a fast, convenient answer.
The promise is real. Tedesco points to satellite data from MODIS or Landsat: an AI can draft Google Earth Engine code on the spot, letting a researcher produce a preliminary time series far faster than before. It can assist with synthesis, pattern recognition, and a first draft of analysis. But the scientific question-deciding what matters, what to test, and how to interpret it-still belongs to the human behind the screen.
An interdisciplinary check on technology
Pavithra Priyadarshini Selvakumar, a postdoctoral research scientist at Columbia Climate School, said Tedesco's initiative is distinctive because it pushes past disciplinary lines. Through the working group, introductory sessions, and a symposium, researchers are examining AI through ethics, neuroscience, mental health, gender, community impact, and data science. "It shows that these issues are not separate, but deeply connected," she said. "He is actively creating spaces where people can come together, learn, question and think differently."
Selvakumar argues that questions of equity and representation cannot be bolted on after the fact. AI systems, she said, are shaped by the same structural privileges and inequalities that exist around us. When women, gender-diverse people, and frontline communities are missing from datasets or design processes, AI tools can deepen that invisibility. In climate work, that can mean certain risks go unrecognized, certain knowledge discounted, and certain needs deprioritized. At its worst, she said, it becomes a form of digital colonialism.
Tedesco's working group is part of a wider push for responsible AI for Science & Research, where scientists can ask hard questions before integrating these systems into their work.
Training the next generation of climate researchers
For David Sathuluri, founder and co-director of Lamont's Laboratory of AI, Climate and Society, the working group arrives at an urgent moment. "I am glad that Marco has taken this important initiative to build an interdisciplinary space at the intersection of AI, climate and society at Columbia," he said. "This work is especially urgent right now, as AI is shaping our lives from every direction, and we need real checks and balances to ensure it is used ethically."
Training younger researchers sits at the center of that effort. Students are entering professional life when AI tools can accelerate coding, writing, and analysis-and also encourage dependency, blur authorship, and skip past the critical thinking habits researchers are still trying to build. Tedesco's advice is not to avoid AI but to use it with guardrails. "The greatest danger," he said, "is that you become a tool for AI rather than AI becoming a tool for you."
Why this matters for science and research professionals
The contradiction is inescapable. A climate scientist might use AI to study the environmental footprint of data centers while relying on the same infrastructure that produces that footprint. The choice of models, training data, and governance structures is often nonexistent. For researchers who want lower-impact, ethically governed tools, the practical options are few. That lack of choice, Tedesco argues, risks concentrating power in a small number of companies and reinforcing the very inequities Earth science seeks to understand.
For scientists, the takeaway is not to abandon AI. It is to approach it with the same rigor applied to any research method. Before habits harden, the questions-about data provenance, transparency, hidden costs, and who benefits-need to be part of every lab, every proposal, and every classroom.
Your membership also unlocks:
