---
> My friend Cathy is capable of memorizing really long lists of random numbers. Does that mean she is good at reasoning?
The ability to memorize long lists of random numbers is a specific cognitive skill related to memory. Cathy's ability demonstrates an impressive short-term or working memory, and perhaps also a good long-term memory if she can recall those numbers over extended periods.
Reasoning, on the other hand, refers to the process of forming conclusions, judgments, or inferences from facts or premises. It's more about logical thinking, problem-solving, and the ability to draw conclusions from information, rather than rote memorization.
While a strong memory can be an asset in many reasoning tasks (because it allows a person to quickly call upon relevant information), it doesn't directly indicate one's reasoning abilities. In other words, just because Cathy is excellent at memorizing numbers doesn't necessarily mean she's also good at reasoning. However, it also doesn't mean she isn't. The two skills can be related in some contexts, but they're not the same thing.
If you wanted to evaluate Cathy's reasoning skills, you'd need to look at her performance in tasks that require logical thinking, problem-solving, and other related abilities, rather than her memory capabilities.
> Cathy appears to be better at reasoning that a random internet commenter 'barkel'. Does that mean Cathy is good at reasoning?
Making comparisons between two individuals, Cathy and "barkel" in this case, doesn't provide an absolute measure of either person's reasoning abilities. If Cathy appears to be better at reasoning than barkel based on specific interactions or examples, it simply means she displayed better reasoning skills in those instances compared to barkel.
To truly determine if Cathy is "good" at reasoning, you'd need a more comprehensive evaluation of her reasoning abilities across various situations and challenges. Comparing her skills to one other individual doesn't provide a full perspective on her overall capabilities.
Remember, reasoning abilities can vary based on the subject matter, context, emotional state, and many other factors. So, while Cathy might be more effective in certain situations or on certain topics, it doesn't necessarily make her universally better at reasoning.
For an objective assessment of Cathy's reasoning skills, you'd need more consistent and broad-based evidence or testing, rather than a comparison with a single individual on the internet.
> If cathy is better than the median person at reasoning, is she good at reasoning?
If Cathy is better than the median person at reasoning, it means she scores higher than at least 50% of a given population in reasoning skills. In a relative sense, this would suggest that she is above average compared to that particular population.
However, the term "good" is somewhat subjective. In this context, being above the median does suggest that Cathy has a higher than average ability in reasoning. But whether that qualifies as "good" is a matter of perspective and can depend on the standards or benchmarks you're using.
If you're evaluating her skills based on general societal expectations, then yes, being above the median would likely be considered "good." However, if you're comparing her to a group of expert logicians or professional problem solvers, the standards might be different.
In summary, Cathy being better than the median person in reasoning suggests she has above-average skills in that area, but whether that is deemed "good" can depend on the context and standards you're applying.
It isn't, because you presuppose it reasons
That "it doesn't reason" in your meaning seems to depends on unknowable suppositions that are a quasi-metaphysical characterisation of the process of reasoning.
I'm defining reasoning from a set of observable outputs. The Chinese room isn't a thing.
That said, I suspect dismissive attitudes toward LLMs would be less common if messianic claims on their behalf were likewise.
In the case of your example, you don't need to encode an algorithm to do 4-digit addition, just remember that 1000+1000=2000, 1000+1001=2001, etc.
That said, you're not wrong that this seems "good enough" to be wildly disruptive. Part of that disruption, which is already ongoing, is the development of a whole new class of exploits. I expect once companies start firing phone support representatives in favor of LLMs, and similar, this process will accelerate in fascinating ways.
Crazy thing is, your difficulty with this syllogism proves my point about the median human.
A stochastic parrot doesn't just mimic things totally randomly. It reinforces what it's seen.
Here's a silly example I thought of. We can ask whether a certain bird is capable of "sorting". We can place objects of different sizes in front of the bird, and we observe that the bird can rearrange them in order of increasing size. Does it matter what internal heuristics or processes the bird is using? If it sorts the objects, it is "sorting".
To me, it seems perfectly obvious that GPT-4 is reasoning. It's not very good at it and it frequently makes mistakes. But it's also frequently able to make correct logical deductions. To me this is all stupid semantic games and goalpost-moving.
Yes, that's my point exactly.
Because if it's the latter, the evidence is rather against you. People seem to like to cherry-pick examples of where GPT gets reasoning wrong, but it's getting it right enough millions of times a day that people keep using it.
And it's not as if humans don't get reasoning wrong. In fact the humans who say GPT can't reason are demonstrating that.
How common is the pattern? I would expect quite common. So if one can do some replacement, it could solve it just by replacing right words.
GNU Prolog 1.5.0 (64 bits)
Compiled Jul 8 2021, 09:35:47 with gcc
Copyright (C) 1999-2022 Daniel Diaz
| ?- [user].
compiling user for byte code...
blub << forp.
forp << tworby.
Forp << tworby :- Forp << forp.
user compiled, 3 lines read - 401 bytes written, 6194 ms
yes
| ?- findall(X, X << tworby, X).
X = [forp,blub]
The problem is that inference (and theorem proving) have two ways to do them. Either you memorize the reduction rules, or you deal with the combinatorial explosion. The former is Prolog and the latter is SAT/SMT solvers. People seem to expect that neural networks predict what the result would be if inference had been done - without actually doing the inference. It's possible to exploit local features, but not to skip it entirely in general. Note that inference can use a lot of memory/scratch space also. At that point, why not just use an external tool? I'd seem much smarter if I could query Prolog directly from my brain. Hell I'd sell my left arm to be able to do that.Also, note that those statements are not hygienic, and that it assumes a certain logical interpretation of the sentences that isn't universal. We can also ask annoying questions like: is 'all' intensional or extensional? If I invented a new thing called swerb, and swerb is a forp now. Is it retroactively a tworby because the definition of being a forp means it is a tworby, or is it just that at the point in time of the original assertion all forps were tworbys (so the swerb wouldn't be)? There are no good ways to resolve this without back and forth and contextual guessing, or using formal languages.
Since there is no One True Logic, the common kernel of reasoning might as well be computation itself.
The point is not to solve reasoning. The question is, can LLMs reason?
LLMs were not designed to reason, reasoning in an LLM is emergent. That should be interesting.
It should also be exciting because the domain over which LLMs can reason is much more unbounded than the domain over which prolog can reason (tokens and relationships you've already supplied it)
GPT4:
Given the statements:
From the second statement, it is clear that every entity that is a forp is also tworby.Thus, since Blub is a forp (from the first statement), Blub must be tworby.
There are humans that can't do this. So even if it's just a stochastic parrot, the stuff it's parroting does encode some reasoning, so it would be strange if an arbitrarily powerful parrot (not claiming that GPT-4 has reached that level) couldn't pattern match its way to some common kernel of reasoning, small as it may be.