Why Do Scientists Use AI?
From Computational Efficiency to Climate Innovation

The most likely way of reversing the trajectory of climate change is through people using AI to develop previously unthought-of methods and technologies.

Daniel Brouse1, Sidd Mukherjee2

1Climatologist, Economist
2Physicist

September 2026

Why Do Scientists Use AI?

Why Scientists Use AI

The short answer is:

We’ve conducted case studies on the use of AI in a wide variety of environments, including everything from the dynamics of nonlinear chaotic systems to songwriting and musical production.

All produced similar results—benefits to the environment, the economy, and society.

For instance, the climate education and graphics used for this post are based on this scientific paper as well as the publicly accessible version.

If I had created them manually, the costs in time, labor, energy, and resources would have been substantial. Using AI, the direct dollar cost to me was effectively zero, the graphics were generated in about a minute, and the estimated natural-resource cost was roughly $0.067 per image.

By comparison, producing similar graphics through traditional methods could easily have cost hundreds of dollars when accounting for labor, software, revisions, and associated natural-resource use. Mathematically, the AI-generated version required approximately 0.005% of those resources—just five thousandths of one percent of the traditional cost. These savings are equivalent to a reduction in environmental impact.

The benefits—in both environmental savings to the planet through reduced physical-resource use and in the ability to produce and distribute education—far outweigh the costs.

Most importantly:

The only effective way to stop the acceleration of climate change is through the reduction of fossil-fuel combustion.

However, the most likely way of reversing the trajectory of climate change is through people using AI to develop previously unthought-of methods and technologies that can fundamentally change how we produce, consume, and manage resources—and potentially remove greenhouse gases from the atmosphere.

Anti-Science and Technology Rhetoric

Unfortunately, the lack of science and technology literacy is at the root of many of society’s problems. Anti-AI opinions can sometimes resemble the anti-science rhetoric used by climate-change deniers, particularly when conclusions are formed before the underlying evidence or technology has been examined.

Research in cognitive and political psychology suggests that rejection of scientific evidence is more complicated than simply a lack of intelligence or education. People can engage in motivated reasoning—evaluating evidence in ways that protect existing political, ideological, or cultural identities. When scientific findings conflict with strongly held beliefs, individuals may scrutinize contradictory evidence more critically, accept supporting evidence with less scrutiny, or reinterpret information so that it remains compatible with their existing worldview. This phenomenon has been documented across a range of politically contested scientific issues, including climate change. (Kahan, 2017; Borghi et al., 2026.)

Importantly, this does not mean that people who reject scientific evidence necessarily lack intelligence or analytical ability. Some research has found that greater reasoning ability can actually strengthen politically motivated reasoning when individuals use that ability to construct arguments supporting conclusions they are already motivated to accept.

Conspiracy beliefs provide another example. A review of the literature found that conspiracy belief is generally associated with greater reliance on intuition and less reflective reasoning. Other preregistered research has found that endorsement of particularly implausible conspiracy theories can be associated with reduced information sampling and less reflective reasoning. (Binnendyk & Pennycook, 2022; Hattersley et al., 2022.)

The same general principles apply to public opinions about technology. Someone who has never used AI as a scientific research assistant is evaluating the technology from a fundamentally different position than a scientist who has spent years developing, testing, and refining specialized AI systems for research. An opinion formed without direct experience may therefore reflect assumptions about what AI is rather than knowledge of what AI can actually do.

Scientists increasingly use AI as a specialized research tool. AI is artificial SUPPLEMENTAL intelligence. It is not necessarily a replacement for the scientist. Instead, it can function as an extension of the researcher’s computational, analytical, and creative capabilities—particularly when the system has been deliberately trained or configured for a specific scientific task.

Science communication research also suggests that curiosity and willingness to engage with information can matter. Kahan found that science curiosity was associated with more open-minded engagement with information that conflicted with a person’s political predispositions. In other words, the ability to consider information that challenges one’s existing beliefs may be at least as important as simply possessing scientific knowledge. (Kahan, 2017.)

In my case, I have spent years developing multiple specially trained “lab assistants” that have learned from scores of our own work. When Sidd and I started, our work had to be done on Ohio State computers and supercomputers that Sidd helped create. Things like “a floating decimal” were, and still are, a significant problem. Problems that needed solving would take months or years to process. The amount of energy spent on computation was enormous compared with today.

Things have changed for the better.

This year, we were able to establish third-derivative behavior across multiple climate indicators—extremely important developments in climate science that confirm that the acceleration rate of climate change is itself accelerating. This work was completed within four months. Historically, I would have been unlikely to complete this work in my lifetime if it were not for AI.

The broader lesson is that scientific progress depends not only on the ability to acquire information but also on the willingness to investigate unfamiliar methods and revise established assumptions. Research on science denial emphasizes that rejection of expert conclusions cannot be explained simply by ignorance or irrationality. Trust, identity, perceived competence, political conflict, and the social cues people use to determine whom to believe can all influence whether scientific evidence is accepted. (Lewandowsky et al., 2022; Sinatra et al., 2019.)

In the meantime, the United States has gone extremist in anti-science and anti-technology ideology, much to the detriment of science and scientists in the U.S. In fact, we are now working on bringing up our own AI computer using Chinese technology. It will be running on DeepSeek. The single computer will do the work of what used to take at least a dozen servers to accomplish, at a fraction of the price. It will not be run in a data center. It will not be costing society anything. In fact, it will be greatly benefiting society and helping to solve the climate crisis.

The distinction is therefore important: criticizing a technology based on evidence, testing, or demonstrated limitations is part of science. Rejecting a technology without understanding how it works or without examining its actual applications is something very different.

So, the next time you want to voice your opinion about AI, do your homework first. Start by getting at least a master’s-level education in climate science. Then spend years training your AI assistants. After completion, I would be happy to discuss AI with you.

References

Binnendyk, J., & Pennycook, G. (2022). Intuition, reason, and conspiracy beliefs. Current Opinion in Psychology, 47, 101387.

Borghi, O., Tappin, B. M., Smets, K., & Tsakiris, M. (2026). Mind over bias: How is cognitive control related to politically motivated reasoning? Cognition, 268, 106373.

Hattersley, M., Pummerer, L., & Van Prooijen, J.-W. (2022). Of tinfoil hats and thinking caps: Reasoning is more strongly related to implausible than plausible conspiracy beliefs. Cognition, 226, 104956.

Kahan, D. M. (2017). Science curiosity and political information processing. Political Psychology, 38(S1), 179–199.

Lewandowsky, S., Smillie, L., Garcia, D., Hertwig, R., Weatherall, J., Egidy, S., Robertson, R. E., O’Connor, C., Kozyreva, A., Lorenz-Spreen, P., Blaschke, Y., & Leiser, M. (2022). When science becomes embroiled in conflict: Recognizing the public’s need for debate while combating conspiracies and misinformation. Proceedings of the National Academy of Sciences, 119(34).

Sinatra, G. M., Kienhues, D., & Hofer, B. K. (2019). Addressing challenges to public understanding of science: Epistemic cognition, motivated reasoning, and conceptual change. Educational Psychologist, 54(1), 1–10.

1. Advanced Clinical Screening and Diagnostics

2. Vaccine Development and Engineering

3. Oncology and Tailored Cancer Therapeutics

4. Materials Science, Physics, and Energy

5. Climate Science and Earth-System Modeling

6. Astronomy and Space Science

7. The Common Thread

These examples span medicine, biology, materials science, physics, astronomy, energy, and climate science. The common thread is not that AI replaces scientists. It is that AI gives scientists a new computational instrument for exploring problems that were previously too large, too complex, or too time-consuming to investigate.

In many cases, AI does not eliminate the need for conventional scientific methods. Instead, it reduces the amount of computational or experimental work required to identify promising possibilities. Scientists still have to formulate questions, evaluate evidence, design experiments, validate results, and determine whether an AI-generated prediction is actually correct.

That is why we describe AI as artificial SUPPLEMENTAL intelligence: a tool that extends human scientific capabilities rather than replacing the scientist.

AI Science Bleeding Edge Technology 2
AI Science Bleeding Edge Technology

Easy-to-Read Resources

Climate Change Simplified

→ The Hitchhiker’s Guide to Climate Change

For the basics: Climate Change Simplified

Climate Change: What Can I Do?

Anthropogenic Global Warming 101

Intro to Advanced Anthropogenic Global Warming