Comment by JoshCole | Hacker News Reader

Comment by JoshCole | Hacker News Reader

Parent Full thread

> What I mean by that, is that AI is often non-explainable, can be non-deterministic and only as good as it’s training data.

By your logic, humans can't discover things, but AI can. You think you made the opposite claim - you didn't.

1. Consider that you can freeze a random seed, making an algorithm deterministic, but you can't do the same for a human.

2. Secondly, I dare you to explain to me exactly, down to the very activation of cells and synapses in your mind, how you came to understand these words as having meaning. You can't do it, but an AI that did this would do this implicitly. So it will do a much better job of explaining how this happens than you will.

So you haven't ruled out AI being able to discover things. You've claimed that humans can't, but AI can. It directly contradicts your thesis.

3. We can point to specific examples of novel discoveries produced by modern ML that were made without the use of good training data provided by humans - indeed, without any training data provided by humans. Moreover, these techniques weren't limited by their training data being bad - they were built around the idea that the training data they would generate was bad.

> 2. Secondly, I dare you to explain to me exactly, down to the very activation of cells and synapses in your mind, how you came to understand these words as having meaning. You can't do it, but an AI that did this would do this implicitly. So it will do a much better job of explaining how this happens than you will.

I think you're working with a very different definition of "explainable" than is commonly used in the machine learning literature.

They are trying to create a distinction, not between the explainability of machine learning approaches, but between human approaches and machine learning approaches. Whatever paper you think I should be using for my definition, just go to there citations. You'll trace it back eventually to things like the Imitation Game paper by Alan Turing. You'll trace it back to the debate about whether machines can think, discover, learn, know. The papers you reference are built atop a mountain of papers that came before them which largely reject the claim that machines aren't capable of discovery. You'll also find my term usage early on in most books about machine learning. Most authors like introducing machine learning by pointing out why we need machine learning. They point out that humans not being able to explain solutions to very hard problems means they can't build programs that solve those problems. If they could explain their solution, they could code it up and they would have a program. Since they can't, they reach to learning tools to figure it out.

It really shouldn't surprise you that we have to go back to the early parts of the field; his argument restated is "machines can't make discoveries, only humans can" and this confused thinking is something that gets addressed in the early writings much moreso than the latter writings.

Frankly, we've known algorithms as far back as Aristotle that were capable of discovery. Peter Norvig in AIMA cites Aristotle when talking about some search problems and path discovery is still discovery. Some of these algorithms literally include frontier in the algorithmic description; clearly a framing that implies thinking of them in terms of discovery. Honestly, what is really happening in my estimation is some people do this thing where they pretend AI is categorically reduced to the subset of things which they can criticize. It is a really weird tendency, but extremely common. You can see it in the persons post actually. Look at the contortions they go through to only focus on supervised learning - to only focus on a narrow subset of supervised learning restricted to deep learning - to only focus on a narrow problem domain of supervised learning - classification - and to only focus on a specific example of that problem domain (we can give good explanations for smaller models) - the ones that humans can't easily explain.

Explainability isn’t an issue in humans, because we prove our findings. An AI cannot typically do so, unless trained to do so. Formal systems such as mathematics can be proven via AI. But an AI output isn’t proof it’s a statistical probability. You still have to go and take measurements / samples (what have you) to prove your case. Further, you must make a hypothesis, then run experiments - things an AI cannot do.

> Further, you must make a hypothesis, then run experiments - things an AI cannot do.

Uh, you do realize that AI and classifiers are two different things right? Like, there are literally, right now - like, already - AI which make a hypothesis and then run an experiment. Just because it is popular to phrase things as supervised learning problems and just because we often want to have classifiers as components of systems we build doesn't mean that is the only technique in AI that exists.

Honestly, your entire point reeks of this really unjustified hypocrisy that makes you think you don't realize this. You seem to pick the cherries for the human, but grab only the poisonous mushrooms for the AI. For example, you seem to close your eyes when you hear about the replication crisis in the social science acting as if humans only emit proofs when they don't, but you're more than willing to focus on how some subsets of AI don't emit proofs despite there being subsets of humans which don't emit proofs.

Reply on news.ycombinator.com