Why Trump’s Genesis Mission to hand science over to AI might be a catastrophic miscalculation

DCM Editorial Summary: This story has been independently rewritten and summarised for DCM readers to highlight key developments relevant to the region. Original reporting by The Conversation, click this post to read the original article.

If you’re following recent developments around Donald Trump’s “Genesis Mission,” you’ll notice it’s a bold push to transform scientific research using artificial intelligence. The idea is to let AI systems design, run, and learn from experiments, all while generating new research questions. The goal? To dramatically boost the speed and productivity of federally funded science, tackling big problems in fields like biotech, robotics, and clean energy. This aligns with what’s happening globally, including in the UK, where governments see AI breakthroughs like DeepMind’s AlphaFold as a model for future scientific progress.

But before you get too excited, you should know that automating science isn’t as simple as it sounds. Leading philosophers argue that real scientific discovery isn’t just about pattern recognition. It’s about spotting anomalies, forming new theories, and using deep human insight to choose which ideas are worth exploring. Today’s AI systems can mimic some parts of this process, but they lack the understanding, creativity, and judgement of human researchers. Without these qualities, AI might miss meaningful breakthroughs that lie outside predictable patterns—or worse, steer science toward easier, short-term wins rather than bold exploration.

You also need to consider what happens when AI makes decisions that affect which experiments are run or which lines of research get abandoned. AI systems aren’t transparent; you often can’t tell why they made a certain decision or whether it came from biased data. If these systems are left unchecked, you risk serious missteps—something real-world events, like the Dutch childcare scandal driven by an algorithm, have already shown. That same opacity could creep into science, affecting research credibility and direction without enough human oversight.

Most importantly, you’ve got to think about how scientific knowledge becomes accepted. Data alone isn’t enough—scientists persuade their peers through arguments, evidence, and critique. If AI starts writing research papers and forming conclusions with little human input, who’s responsible for defending those claims? Will journals and peers judge machine-generated work with the same skepticism they use with human-led studies? Taking past AI successes like AlphaFold out of context can be risky, especially when those tools became valuable precisely because they were guided by human scientists.

So if you care about how science moves forward, it’s crucial to see AI not as a replacement but as a tool—something that supports, not substitutes, your human-driven research communities. Keeping the imagination, debate, and even the messy disagreements alive is what makes science thrive. The Genesis Mission may hold potential, but it must be carefully grounded in the messy, human-centered culture that science truly depends on.