What AI can do and what it can't.
It occurred to me that I haven’t talked about AI for a while. I guess I ought to do that since it’s become my “thing,” so to speak. Go spelunking in my previous entries to find those thoughts.
I won’t link to it since you can find it quickly enough on your own, but Adam Conover did something recently in which he interviewed a couple of AI skeptics who came on to reaffirm his anti-AI bias. They claimed that AI won’t ever become sentient and will never be able to create coherent art with just a prompt.
Don’t be so quick to judge, guys. People were mocking generative AI not so long ago because it couldn’t consistently produce pictures of photorealistic hands. They used this as “proof” that AI was useless and fundamentally flawed. And guess what? Hands aren’t an issue anymore. And neither are many things generative AI had a hard time doing.
Large Language Models (LLMs) aren’t conscious but are a step toward sentience. As they grow more powerful and sophisticated, they are hurtling toward Artificial General Intelligence (AGI), which is a huge progression.
Anytime someone says AI can’t do something, I always append yet. AI has repeatedly proven that what it appears incapable of today, it will do tomorrow. That won’t stop because Adam Conover wants it to. And soon, AI won’t even need us to learn and grow. Think about that.