Spoilers: this short article isn’t going to answer the premise of the article. Further spoiler: so many people have discussed this that know more than I do. But an introduction to an idea, or even repetition of common talking points, is fertile ground for thinking of something more deeply.
So, what is artificial intelligence, as compared to “regular” intelligence? Where do we draw the line, exactly? Artificial, as a word, means that it was created. We made a computer that can “think” in some capacity. It derives information from other information.
But how artificial is human intelligence? Yes, we are born with a brain. It’s “grown.” Nothing artificial about that. But how many concepts does the average human have if not taught them? Education, as a concept, exists because we do not know things and should, ideally, know them. Is that not artificial? Are you expecting the average person to learn calculus unprompted and without access to previous documentation? Should the word intelligence even be used when “knowledge” might be more accurate?
Perhaps it is the learning method that defines the difference between humans and computers. The difference in how information is gathered. Humans don’t learn stuff by dropping raw numerical data into a massive container alone—connections and relevance are a part of this. Memory and experience inform input as much as the output.
Furthermore, it is the inference that simply isn’t represented in artificial thought? That’s a potent line of thinking. Might get us somewhere. I don’t—currently—expect a computer to be introduced to a concept, given incomplete data, and then go out of its way to fill in data gaps. That feels distinctly human. The desire to learn is an organic thing.
I also don’t expect an artificial intelligence to ask seemingly unrelated questions to infer a greater conceptual framework and make connections that are not linearly related. It likely can’t solve a brainteaser that relies on meta-textual understanding to undermine the initial premise. I imagine it would have a lot of trouble with an unestablished pun.
And, like I said, I can’t answer these questions. But I think we should—you and I—keep asking them. In art, in science, in the day-today: these questions are important. Because once AGI exists, we should really already have an answer.
