ChatGPT may be an excellent tool in case your strongly-worded email to your landlord about that ceiling leak needs a second pair of eyes. It also excels at coming up with a rough first draft for non-mission-critical writing, allowing you to carefully pick it apart and refine it.<\/p>\n
But like all of its competitors, ChatGPT is plagued by plenty of well-documented shortcomings as well, from rampant hallucinations to a sycophantic tone that can easily lull users into gravely mistaken beliefs<\/a>.<\/p>\n In other words, it\u2019s not exactly a tool anybody should rely on to get important work done \u2014 and that\u2019s a lesson University of Cologne professor of plant sciences Marcel Bucher learned the hard way.<\/p>\n In a column for Nature<\/a>, Bucher admitted he\u2019d \u201clost\u201d two years\u2019 worth of \u201ccarefully structured academic work\u201d \u2014 including grant applications, publication revisions, lectures, and exams \u2014 after turning off ChatGPT\u2019s \u201cdata consent\u201d option.<\/p>\n He disabled the feature because he \u201cwanted to see whether I would still have access to all of the model\u2019s functions if I did not provide OpenAI with my data.\u201d <\/p>\n But to his dismay, the chats disappeared without a trace in an instant.<\/p>\n \u201cNo warning appeared,\u201d Bucher wrote. \u201cThere was no undo option. Just a blank page.\u201d<\/p>\n The column was met with an outpouring of schadenfreude on social media, with users questioning how Bucher had gone two years without making any local backups<\/a>. Others were enraged, calling on the university to fire him<\/a> for relying so heavily on AI for academic work.<\/p>\n Some, however, did take pity.<\/p>\n \u201cWell, kudos to Marcel Bucher for sharing a story about a deeply flawed workflow and a stupid mistake,\u201d Heidelberg University teaching coordinator Roland Gromes wrote in a post on Bluesky<\/a>. \u201cA lot of academics believe they can see the pitfalls but all of us can be naive and run into this kind of problems!\u201d<\/p>\n Bucher is the first to admit that ChatGPT can \u201cproduce seemingly confident but sometimes incorrect statements,\u201d arguing that he never \u201cequated its reliability with factual accuracy.\u201d Nonetheless, he \u201crelied on the continuity and apparent stability of the workspace,\u201d using ChatGPT Plus as his \u201cassistant every day.\u201d<\/p>\n The use of generative AI in the scientific world has proven highly controversial.<\/p>\n Scientific journals are being flooded with poorly sourced AI slop<\/a>, turning the process of peer review into a horror show, as The Atlantic reported this week<\/a>. Entire fraudulent scientific journals are popping up to capitalize on others who are trying to get their AI slop published. The result? AI slop being peer-reviewed by AI models, further entrenching the polluting of scientific literature.<\/p>\n For their part, scientists are constantly being informed of how their work is being cited in various new papers \u2014 only to find that the referenced material was entirely hallucinated<\/a>.<\/p>\n To be clear, there\u2019s zero evidence that Bucher was in any way trying to sell off AI slop to his students or get dubious, AI-generated research published. <\/p>\n Nonetheless, his unfortunate experience with the platform should serve as a warning sign to others.<\/p>\n In his column, Bucher accused OpenAI of selling subscriptions to its ChatGPT Plus despite not assuring \u201cbasic protective measures\u201d to stop years of his work from vanishing in an instant.<\/p>\n In a statement to Nature, OpenAI clarified that chats \u201ccannot be recovered\u201d after being deleted, and challenged Bucher\u2019s claim that there was \u201cno warning,\u201d saying that \u201cwe do provide a confirmation prompt before a user permanently deletes a chat.\u201d<\/p>\n The company also helpfully recommended that \u201cusers maintain personal backups for professional work.\u201d<\/p>\n