Post 28: Let’s define Artificial General Intelligence (AGI)
Quick post today. I really just want to define AGI a little better than we have in the past for the purposes of the Substack going forward as I think it is about to become more relevant. The definition I received from ChatGPT 4o is:
Artificial General Intelligence (AGI) is an AI system capable of understanding, learning, and applying knowledge across a wide range of tasks at a human or superhuman level, with the ability to generalize beyond its training data.
After writing “Post N: How an LLM is trained,” AGI seems a lot more relevant to me in terms of where the Substack will be going at least for a while, (or at least until I get bored of the whole consciousness discussion and move on to other topics. . .), so I think this brief (re)introduction is in order.
Aside from just giving this definition, it is important to note that people are working towards AGI right now, (this will be covered more in tomorrow’s post and doubtlessly many posts going forward), and regardless of when or if it happens, (some people say it is coming this year, others 5, 10, 15 years in the future. . . or never), some knowledge of it will certainly make the AI consciousness discussion easier, and well, more practical? So, see below for a slightly expanded definition and some hints at things to be covered soon:
