4 ms·

Investigative skills. AI can't do anything a detective does. Formulating a theory, testing it, refining it etc. Also quickly grabbing arbitrarily shaped object

by andy99 3y ago

Investigative skills. AI can't do anything a detective does. Formulating a theory, testing it, refining it etc. Also quickly grabbing arbitrarily shaped objects from a pile.

sweetheart 3y ago
I'm not disagreeing, but this answer is ironic considering how Ilya Sutskever once used detective work as a way to describe the relationship between LLMs and "understanding": "So, I’d like to take a small detour and to give an analogy that will hopefully clarify why more accurate prediction of the next word leads to more understanding, real understanding. Let’s consider an example. Say you read a detective novel. It’s like complicated plot, a storyline, different characters, lots of events, mysteries like clues, it’s unclear. Then, let’s say that at the last page of the book, the detective has gathered all the clues, gathered all the people and saying, “okay, I’m going to reveal the identity of whoever committed the crime and that person’s name is”. Predict that word. Predict that word, exactly. My goodness. Right? Yeah, right. Now, there are many different words. But predicting those words better and better and better, the understanding of the text keeps on increasing. GPT-4 predicts the next word better."
- andy99 3y ago
  He's talking about the model as a "decider" which it's good at, not a detective or investigator.
  - ilaksh 3y ago
    For the model to output the highest likelihood for the correct token, it must literally solve the mystery by inference with the given clues/narrative.
    - andy99 3y ago
      That's completely different from investigation which involves iteratively hypothesizing and gathering evidence, and at each step coming up with ideas from an open field of possibilities. Having everything laid out for you as a classification problem and then just pronouncing is what ML models do very well and is an unrelated problem formulation. "AI" is great at solving well defined problem. Deciding there is a problem, formulating it, deciding and collecting what you need to solve or test it, iterating, not so much.
recursive 3y ago
Your recipe for investigative skills also sounds a lot like science. Even today, the various chat bots can formulate theories and test protocols. I don't see anything that puts any of that out of reach.
ben_w 3y ago
I'm sure I've seen articles about AI doing the former and videos demonstrating robots doing the latter.

by andy99 3y ago

Investigative skills. AI can't do anything a detective does. Formulating a theory, testing it, refining it etc. Also quickly grabbing arbitrarily shaped objects from a pile.

sweetheart 3y ago

I'm not disagreeing, but this answer is ironic considering how Ilya Sutskever once used detective work as a way to describe the relationship between LLMs and "understanding": "So, I’d like to take a small detour and to give an analogy that will hopefully clarify why more accurate prediction of the next word leads to more understanding, real understanding. Let’s consider an example. Say you read a detective novel. It’s like complicated plot, a storyline, different characters, lots of events, mysteries like clues, it’s unclear. Then, let’s say that at the last page of the book, the detective has gathered all the clues, gathered all the people and saying, “okay, I’m going to reveal the identity of whoever committed the crime and that person’s name is”. Predict that word. Predict that word, exactly. My goodness. Right? Yeah, right. Now, there are many different words. But predicting those words better and better and better, the understanding of the text keeps on increasing. GPT-4 predicts the next word better."

andy99 3y ago
He's talking about the model as a "decider" which it's good at, not a detective or investigator.
- ilaksh 3y ago
  For the model to output the highest likelihood for the correct token, it must literally solve the mystery by inference with the given clues/narrative.
  - andy99 3y ago
    That's completely different from investigation which involves iteratively hypothesizing and gathering evidence, and at each step coming up with ideas from an open field of possibilities. Having everything laid out for you as a classification problem and then just pronouncing is what ML models do very well and is an unrelated problem formulation. "AI" is great at solving well defined problem. Deciding there is a problem, formulating it, deciding and collecting what you need to solve or test it, iterating, not so much.

recursive 3y ago

Your recipe for investigative skills also sounds a lot like science. Even today, the various chat bots can formulate theories and test protocols. I don't see anything that puts any of that out of reach.

ben_w 3y ago

I'm sure I've seen articles about AI doing the former and videos demonstrating robots doing the latter.