{"id":425989,"date":"2026-01-20T18:14:08","date_gmt":"2026-01-20T18:14:08","guid":{"rendered":"https:\/\/www.newsbeep.com\/au\/425989\/"},"modified":"2026-01-20T18:14:08","modified_gmt":"2026-01-20T18:14:08","slug":"ai-cannot-automate-science-a-philosopher-explains-the-uniquely-human-aspects-of-doing-research","status":"publish","type":"post","link":"https:\/\/www.newsbeep.com\/au\/425989\/","title":{"rendered":"AI cannot automate science \u2013 a philosopher explains the uniquely human aspects of doing research"},"content":{"rendered":"<p>Consistent with the general trend of <a href=\"https:\/\/www.nytimes.com\/2026\/01\/17\/business\/tech-firms-ai-retailers.html?smid=nytcore-ios-share\" rel=\"nofollow noopener\" target=\"_blank\">incorporating artificial intelligence into nearly every field<\/a>, researchers and politicians are increasingly using AI models trained on scientific data to infer answers to scientific questions. But can AI ultimately replace scientists?<\/p>\n<p>The Trump administration signed an executive order on Nov. 24, 2025, that announced <a href=\"https:\/\/www.whitehouse.gov\/presidential-actions\/2025\/11\/launching-the-genesis-mission\/\" rel=\"nofollow noopener\" target=\"_blank\">the Genesis Mission<\/a>, an initiative to build and train a series of <a href=\"https:\/\/theconversation.com\/what-is-an-ai-agent-a-computer-scientist-explains-the-next-wave-of-artificial-intelligence-tools-242586\" rel=\"nofollow noopener\" target=\"_blank\">AI agents<\/a> on federal scientific datasets \u201cto test new hypotheses, automate research workflows, and accelerate scientific breakthroughs.\u201d<\/p>\n<p>So far, the accomplishments of these so-called <a href=\"https:\/\/doi.org\/10.1145\/3769733.3769747\" rel=\"nofollow noopener\" target=\"_blank\">AI scientists have been mixed<\/a>. On the one hand, AI systems can process vast datasets and detect subtle correlations that humans are unable to detect. On the other hand, their lack of commonsense reasoning can result in unrealistic or irrelevant experimental recommendations.<\/p>\n<p>While AI can assist in tasks that are part of the scientific process, it is still far away from automating science \u2013 and may never be able to. <a href=\"https:\/\/scholar.google.com\/citations?user=_UUMxr0AAAAJ\" rel=\"nofollow noopener\" target=\"_blank\">As a philosopher<\/a> who studies both the history and the conceptual foundations of science, I see several problems with the idea that AI systems can \u201cdo science\u201d without or even better than humans.<\/p>\n<p>AI models can only learn from human scientists<\/p>\n<p>AI models do not learn directly from the real world: They have to be <a href=\"https:\/\/news.mit.edu\/2024\/generative-ai-lacks-coherent-world-understanding-1105\" rel=\"nofollow noopener\" target=\"_blank\">\u201ctold\u201d what the world is like<\/a> by their human designers. Without human scientists overseeing the construction of the digital \u201cworld\u201d in which the model operates \u2013 that is, the datasets used for training and testing its algorithms \u2013 the breakthroughs that AI facilitates wouldn\u2019t be possible.<\/p>\n<p>Consider the AI model AlphaFold. Its developers were awarded the <a href=\"https:\/\/theconversation.com\/machine-learning-cracked-the-protein-folding-problem-and-won-the-2024-nobel-prize-in-chemistry-240937\" rel=\"nofollow noopener\" target=\"_blank\">2024 Nobel Prize in chemistry<\/a> for the model\u2019s ability to infer the structure of proteins in human cells. Because so many biological functions depend on proteins, the ability to quickly generate protein structures to test via simulations has the potential to accelerate drug design, trace how diseases develop and advance other areas of biomedical research.<\/p>\n<p>As practical as it may be, however, an AI system like AlphaFold does not provide new knowledge about proteins, diseases or more effective drugs on its own. It simply makes it possible to analyze existing information more efficiently. <\/p>\n<\/p>\n<p>            AlphaFold draws upon vast databases of existing protein structures.<\/p>\n<p>As philosopher Emily Sullivan put it, to be successful as scientific tools, AI models must <a href=\"https:\/\/doi.org\/10.1093\/bjps\/axz035\" rel=\"nofollow noopener\" target=\"_blank\">retain a strong empirical link<\/a> to already established knowledge. That is, the predictions a model makes must be grounded in what researchers already know about the natural world. The strength of this link depends on how much knowledge is already available about a certain subject and on how well the model\u2019s programmers translate highly technical scientific concepts and logical principles into code.<\/p>\n<p>AlphaFold would not have been successful if it weren\u2019t for the <a href=\"https:\/\/www.rcsb.org\/pages\/about-us\/index\" rel=\"nofollow noopener\" target=\"_blank\">existing body of human-generated knowledge about protein structures<\/a> that developers used to train the model. And without human scientists to provide a foundation of theoretical and methodological knowledge, nothing AlphaFold creates would amount to scientific progress.<\/p>\n<p>Science is a uniquely human enterprise<\/p>\n<p>But the role of human scientists in the process of scientific discovery and experimentation goes beyond ensuring that AI models are properly designed and anchored to existing scientific knowledge. In a sense, science as a creative achievement derives its legitimacy from human abilities, values and ways of living. These, in turn, are grounded in the unique ways in which humans think, feel and act.<\/p>\n<p>Scientific discoveries are more than just theories supported by evidence: They are the <a href=\"https:\/\/archive.org\/details\/thomas-s.-kuhn-the-structure-of-scientific-revolutions\" rel=\"nofollow noopener\" target=\"_blank\">product of generations of scientists<\/a> with a variety of interests and perspectives, working together through a common commitment to their craft and intellectual honesty. Scientific discoveries are never the products of a single visionary genius. <\/p>\n<p>            <a href=\"https:\/\/images.theconversation.com\/files\/712946\/original\/file-20260116-66-w3m4gh.jpg?ixlib=rb-4.1.0&amp;q=45&amp;auto=format&amp;w=1000&amp;fit=clip\" rel=\"nofollow noopener\" target=\"_blank\"><img decoding=\"async\" alt=\"Group of people wearing white lab coats and protective eye equipment working in a lab\" src=\"https:\/\/www.newsbeep.com\/au\/wp-content\/uploads\/2026\/01\/file-20260116-66-w3m4gh.jpg\" class=\"native-lazy\" loading=\"lazy\"  \/><\/a><\/p>\n<p>              Breakthroughs are possible through collaboration across generations of scientists.<br \/>\n              <a class=\"source\" href=\"https:\/\/www.gettyimages.com\/detail\/photo\/experiment-science-and-group-of-people-in-royalty-free-image\/2163671991\" rel=\"nofollow noopener\" target=\"_blank\">Jacob Wackerhausen\/iStock via Getty Images Plus<\/a><\/p>\n<p>For example, when researchers first proposed the <a href=\"https:\/\/www.nature.com\/scitable\/topicpage\/discovery-of-dna-structure-and-function-watson-397\/\" rel=\"nofollow noopener\" target=\"_blank\">double-helix structure of DNA<\/a>, there were no empirical tests able to verify this hypothesis \u2013 it was based on the reasoning skills of highly trained experts. It took nearly a century of technological advancements and several generations of scientists to go from what looked like pure speculation in the late 1800s to a discovery honored by a 1953 Nobel Prize.<\/p>\n<p>Science, in other words, is a <a href=\"https:\/\/plato.stanford.edu\/entries\/scientific-knowledge-social\/\" rel=\"nofollow noopener\" target=\"_blank\">distinctly social enterprise<\/a>, in which ideas get discussed, interpretations are offered, and disagreements are not always overcome. As other philosophers of science have remarked, scientists are <a href=\"https:\/\/doi.org\/10.2307\/j.ctt32bbxc\" rel=\"nofollow noopener\" target=\"_blank\">more similar to a tribe<\/a> than \u201c<a href=\"https:\/\/press.uchicago.edu\/ucp\/books\/book\/chicago\/T\/bo37447570.html\" rel=\"nofollow noopener\" target=\"_blank\">passive recipients\u201d of scientific information<\/a>. Researchers do not accumulate scientific knowledge by recording \u201cfacts\u201d \u2013 they create scientific knowledge through skilled practice, debate and agreed-upon standards informed by social and political values. <\/p>\n<p>AI is not a \u2018scientist\u2019<\/p>\n<p>I believe the computing power of AI systems can be used to accelerate scientific progress, but only if done with care.<\/p>\n<p>With the active participation of the scientific community, ambitious projects like the Genesis Mission could prove beneficial for scientists. Well-designed and rigorously trained AI tools would make the more mechanical parts of scientific inquiry smoother and maybe even faster. These tools would compile information about what has been done in the past so that it can more easily inform how to design future experiments, collect measurements and formulate theories.<\/p>\n<p>But if the guiding vision for deploying AI models in science is to replace human scientists or to fully automate the scientific process, I believe the project would only turn science into a caricature of itself. The very existence of science as a source of authoritative knowledge about the natural world fundamentally depends on human life: shared goals, experiences and aspirations.<\/p>\n","protected":false},"excerpt":{"rendered":"Consistent with the general trend of incorporating artificial intelligence into nearly every field, researchers and politicians are increasingly&hellip;\n","protected":false},"author":2,"featured_media":425990,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[20],"tags":[256,254,255,64,63,105],"class_list":["post-425989","post","type-post","status-publish","format-standard","has-post-thumbnail","category-artificial-intelligence","tag-ai","tag-artificial-intelligence","tag-artificialintelligence","tag-au","tag-australia","tag-technology"],"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/posts\/425989","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/comments?post=425989"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/posts\/425989\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/media\/425990"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/media?parent=425989"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/categories?post=425989"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/tags?post=425989"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}