2 weeks ago
Wikipedia Founder Jimmy Wales Says AI Can't Be Trusted
A very smart person named Jimmy Wales started Wikipedia, the big online encyclopedia.
He says computers that write with AI can write stories and answers, but we should not trust them.
These AI tools work by guessing which word should come next, like a game of word prediction.
They pick the word that seems most likely to follow the words before it.
Sometimes the answers they write sound really good and make sense.
But Jimmy Wales says that does not mean the AI knows what is true.
He calls the mistakes these tools make 'hallucinations.'
Hallucinations happen when the AI says something confidently that is not real or true.
So even though AI can write well, it does not actually understand facts.
That is why we need to be careful and check what AI tells us.
Jimmy Wales, the founder of Wikipedia, says AI can write but cannot be trusted.
He identifies 'hallucinations' as one of the core problems with AI today.
Large language models work by predicting the next word based on what came before.
AI responses can sound coherent and reasonable, which makes them impressive.
Wales argues that coherent output does not mean AI understands reality or facts.
- Who
- Jimmy Wales, the founder of Wikipedia
- What
- Explained that AI can write but cannot be trusted because of hallucinations
- Where
- Not specified
- When
- Not specified
- Why
- Because large language models predict likely next words rather than understanding reality or facts
Key facts
- Speaker
- Jimmy Wales
- Role
- Founder of Wikipedia
- Topic
- Trustworthiness of AI writing
- Core problem
- AI hallucinations
- How LLMs work
- Predicting the next word based on what came before
- Key claim
- Coherent responses do not mean understanding of reality or facts
Quotes
Jimmy Wales
Founder of Wikipedia and CEO of the Wikimedia Foundation
“One of the core problems with AI today is what we call hallucinations. Large language models work by predicting the next word based on what came before, choosing what seems most likely. The results can be impressive. We’ve all seen AI produce responses that sound coherent and reasonable. But that doesn’t mean it actually understands reality or facts.”
timesnownews.com







