Highlights
- Royal Observatory warns AI reliance may weaken questioning and evaluation skills.
- Early astronomers’ “unnecessary” work led to breakthroughs 150 years later.
- Studies show cognitive outsourcing to AI negatively impacts memory and learning skills.
The Royal Observatory Greenwich, known for 350 years of astronomical research, told BBC that relying too heavily on artificial intelligence could make humans less intelligent.
Paddy Rodgers, director of the Royal Museums Greenwich group, pointed to the observatory's history as proof that human curiosity and independent thinking drive real progress.
His comments come as AI systems become more common in daily life, from Google search results to social media platforms like TikTok and X.
1. People stop asking questions
Rodgers noted that getting instant answers from AI means people lose "the habits of questioning and evaluation that underpin knowledge, expertise and innovation." When you stop asking your own questions, you miss the deeper understanding that comes from working through problems yourself.
Students who rely on AI for homework may struggle to develop problem-solving skills needed for exams and real-world situations. As a result pupils cannot explain how they reached an answer because they simply copied what AI provided.
2. Machines miss unexpected discoveries
Early astronomers at the Royal Observatory collected huge amounts of data about space. They did things "a machine would not do" because it seemed unnecessary at the time. But 150 years later, that same data helped scientists understand navigation on Earth.
AI focuses on quick answers and skips the exploration that leads to unexpected breakthroughs. AI systems follow patterns and may ignore anomalies that could lead to important findings.
3. Brains get lazy
Dr Anuschka Schmitt from the London School of Economics says AI makes it too easy for people to stop thinking hard about problems. When people let AI do their thinking, their memory gets worse and they learn less. Scientists call this "cognitive outsourcing" and studies prove it harms how our brains work. Depending on AI for answers may weaken our ability to recall information.
4. Cannot verify facts easily
With Wikipedia, you could check where information came from and decide if it was reliable. AI answers often skip this step. "You're getting more and more distanced from relatable or checkable information," Rodgers noted. You get answers but cannot trace where they originated or verify if they are correct.
AI chatbots have been caught providing false information presented as fact, sometimes inventing sources that do not exist. Without clear references, users cannot differentiate between accurate data and AI hallucinations.
5. AI can help if used carefully
The observatory says AI is not all bad. Sir Demis Hassabis won the 2024 Nobel Prize for Chemistry using AI to study proteins. LinkedIn founder Reid Hoffman suggested asking AI to challenge your ideas instead of just agreeing with you. An Oxford lecturer told BBC that AI helps students when used responsibly, but simply handing over all thinking to machines shows its limits.
