What is AI?

Definition

By Artificial Intelligence (AI), I specifically mean systems powered by Large Language Models (LLMs).

How we should understand AI

There are many views on this. Some say it is just a statistical model and fundamentally just a computer program, though a very complicated one. Others say it has intelligence comparable to a human, and perhaps even has a consciousness.

I think both these extremes get it wrong. On the one hand, you cannot reduce today’s LLM-powered AI systems to just a statistical model and a regular computer program and believe that helps you understand it. On the other hand, I see no reason to believe that today’s AI systems have human-like intelligence, let alone consciousness.

I believe they are different from everything we have seen before, we have to consider and understand them as a category of its own and avoid hasty analogies with other things (traditional computer programs, humans, animals). We need to consider LLMs as complex systems with emergent behavior when they grow large enough (and many LLMs in use today are large enough for this).

How we should approach AI

I think we need to consider AI systems as tools, just like computers and traditional computer programs. Very powerful and versatile tools, but fundamentally just tools for us to use. We need to resist the temptation of anthropomorphism and avoid treating them as a human (or an animal). An AI system is your tool, it is not your friend, your partner or your pet.