At the end of the day, it is a tool and it depends on how you use it. It may destroy the research ability of the average person, but for the power user it is an intelligence accelerator IMO.
At the end of the day, it is a tool and it depends on how you use it. It may destroy the research ability of the average person, but for the power user it is an intelligence accelerator IMO.
In this “post-truth” era, I think this is deeply concerning and has the potential to far outweigh the benefits in the long run. People are already not good at critically evaluating information, and this is already leading to major real world impact.
I say this as someone who has personally found LLMs to be a learning multiplier. But then I see how many people treat these tools like some kind of oracle and start to worry.
If the societies of 2005 had the technologies of 2025, I expect OpenAI/Anthropic etc would have a much more challenging time convincing people that "convincingly incorrect" systems should send Nvidia to a $1tn+ valuation.
You live in a planet with billions of other humans. Maybe you are using LLMs carefully and always verifying outputs but most people definitely are not and it is naive to believe that is only their problem. It will soon be your problem too, because what those people do will eventually come back to bite you.
An unrelated quote from John Green feels appropriate:
> Let me explain why I like to pay taxes for schools even though I personally don’t have a kid in school. It’s because I don’t like living in a country with a bunch of stupid people.
One day you’ll be deeply affected by a code bug or clerical decision caused by someone who blindly accepted the words of whatever LLM they were using. An LLM which can itself be created with specific bias, like denying the existence of a country, rejecting scientific consensus, or simply trying to sell you a product.