I think that the mistake is thinking that "smart" is a meaningful word. I'd encourage you to learn about the technology you're critiquing and not listen to memetic bullshit from articles like the one we're discussing. Consider:
AI/cybernetics/robotics (same field, different perspectives) is always only useful for specific tasks, never for general replacement of humans
Black-box treatments of machine learning are only done at the most introductory level and there are several ways to examine how e.g. a Transformers-based language model's weights contribute to its outputs
We have many useful theories about how to learn functions in general, with machine learning as a special case
This has happened before and it will happen again. I'm sure you've seen the phrase "AI winter" floating around.
I encourage you to read the comment I left down lower in response to someone else. I go into more details about the point I'm trying to make. As I mentioned above I'm not an expert on the topic. But I am open to criticism. It's kinda the reason I make comments like this in the first place. I'm not trying to rile people up but get a discussion going.
I think that the mistake is thinking that "smart" is a meaningful word. I'd encourage you to learn about the technology you're critiquing and not listen to memetic bullshit from articles like the one we're discussing. Consider:
This has happened before and it will happen again. I'm sure you've seen the phrase "AI winter" floating around.
I encourage you to read the comment I left down lower in response to someone else. I go into more details about the point I'm trying to make. As I mentioned above I'm not an expert on the topic. But I am open to criticism. It's kinda the reason I make comments like this in the first place. I'm not trying to rile people up but get a discussion going.
I appreciate your feedback. I'm learning as I go