Question: What are the biggest misconceptions when using AI?

Carl Peter Robinson answered on 17 Jan 2024: last edited 5 Feb 2024 8:08 am

Nice question 🙂 We’re in the AI Hype phase of things so it’s a great question to keep people grounded. A few I would list are as follows:

1. That AI will solve all our problems. For example, there’s no way AI is solving climate change on its own. That’s going to take a massive global effort of cooperation and compromise by human beings and their habits and their own self-interests. AI cannot enforce that but it can provide data analyses and run lots of experiments as to how we humans can help ourselves.

2. That if Artificial General Intelligence (AGI) comes around, it will destroy the human race. I don’t think AGI would do that or that we humans would let it get that far down such a path. Our greatest threat is ourselves, not the tools we create.

3. That ChatGPT is an example of AGI, or on the way to being one. Perhaps this is controverisal but I just don’t see it that way. It’s a massive, next-level knowledge-base that is very good at knowing which word to predict next given the previous words it has output. I know it’s been successfully applied in other use cases, but I still can’t get past the mechanical nature of the system. For me, an AGI would feel more organic, and maybe not require so much computing resource, energy, and money to run. Or is that a pipedream on my part, perhaps?