Jeroen Baertjbaert@mastodon.social
Oct 14, 2026, 12:55 AMIt's astonishing how many people reply to the bit in my keynote about AI hallucinations with "we don't have that problem, see, because all our input data is 100% correct". This is something that - even after 5 years of LLM discourse - people (often management level) still cannot grasp.