xamuel

11 ms·

91.
by xamuel 6y ago
Not sure what you mean exactly. Note, the set P of intuitive ordinal notations is not computably enumerable. Proof: if X were a computer program which enumerated all the intuitive ordinal notations and nothing else, then X itself would (by
92.
by xamuel 6y ago
As an example to illustrate how this initially-simple-looking project spans the whole range of all possible difficulty levels: suppose you had a source-code for a superhuman flawless obedient honest AGI, call it HAL. From that source-code,
93.
by xamuel 6y ago
Here's an open-ended programming project which, in a certain formal sense, spans the entire range of all difficulty levels: write an "intuitive ordinal notation" for as large of an ordinal number as you can. What is an "
94.
by xamuel 6y ago
I do sometimes wonder to what extent Socrates'/Plato's/Aristotle's legacy benefited from Aristotle's connection to Alexander the Great. Had Alexander not been so famous, would those three have vanished into obs
95.
by xamuel 6y ago
It's easy to turn a blind eye to the benefits we've gotten from philosophy, it sounds like a bunch of truisms and we think surely we would've just figured that easy stuff out ourselves without the philosophers. But suppose yo
96.
by xamuel 6y ago
Not still alive? He released an original new album six months ago , which reached #2 in the U.S.
97.
by xamuel 6y ago
You can already see this in academia, where there are sometimes more people trying to write papers than there are people actually reading said papers
98.
by xamuel 6y ago
I always find it odd that people take such a hands-off approach when it comes to things like Vitamin D usage, while at the same time cheerleading for governments to crack down hard on things like gyms, birthday parties, church services, ele
99.
by xamuel 6y ago
I wish it were possible for similar prompts to appear before all sorts of policy-makers and bureaucrats. "It appears you are about to institute a policy which will require 400 million patients to sign an additional waiver every time th
100.
by xamuel 6y ago
Here's something fun to think about. Instead of evil Skynet, suppose the entity resurrecting long-dead brains is a benevolent entity. As such, it would avoid resurrecting people against their will. How would it determine which people w
101.
by xamuel 6y ago
Suppose I claimed it's impossible to understand Shakespeare with all the consonants are deleted, leaving only vowels. Most people would accept that claim without pedantically asking me to define "understand". Same story here.
102.
by xamuel 6y ago
It's not about whether the agent will or will not find an optimal solution (no agent can possibly find good solutions in every environment). It's about whether the agent could even understand the true environment based on your
103.
by xamuel 6y ago
The argument in the paper has nothing to do with cardinality. Review my comment a few steps above about Big-O. The reasoning there shows the real numbers, despite having cardinality of the continuum, are inadequate for measuring even the fo
104.
by xamuel 6y ago
I make no attempt to disprove an entire field, rather to question an implicit assumption, namely that real number rewards are flexible enough to capture all relevant environments an AGI should be able to comprehend. Quite the opposite, I in
105.
by xamuel 6y ago
Whether man is machine can't be trivially answered using a few handwavy applications of Godel's theorem nor of observations about the structure of the reals. My paper does not make any attempt to weigh on that question, neither do
106.
by xamuel 6y ago
I'm surprised to see that that's actually a thing people write about. I'm afraid I can't answer your question right now, as I have no idea why anyone would do that (as opposed to, say, vector-valued rewards---i.e., why i
107.
by xamuel 6y ago
>Humans don't receive infinite rewards because any "reward system" in our brain is implemented by receiving a finite amount of electrochemical pleasure signals over a finite amount of time. This is like saying computers ca
108.
by xamuel 6y ago
Question: when you say "I can't see why you couldn't use real numbers to represent utility", does your reasoning for that have anything to do with Dedekind cuts, Cauchy sequences, or complete ordered fields? Because that
109.
by xamuel 6y ago
Easy to come up with examples using exotic money-related constructions. Suppose there's something called a "superdollar". If you have a superdollar, you can use it to create an arbitrary number of dollars for yourself, any ti
110.
by xamuel 6y ago
Hi, thanks for looking at my paper. I do not assume that an AGI must be capable of performing optimally on all tasks in general--indeed, that's quite impossible. When measuring the performance of RL agents, one must come up with some w
111.
by xamuel 6y ago
One could certainly contemplate versions of RL with non-linear orderings. I guess the reason people care about linear ordering is because you want the agent to at least understand "this outcome is better than that outcome". How wo
112.
by xamuel 6y ago
Hi, thanks for looking at my paper. If you're interested in the relation between Lucas-Penrose stuff and enumeratability of ordinal codes, you might like I.J. Good (1969), "Godel's theorem is a red herring" (2 pages). Ca
113.
by xamuel 6y ago
Broken in the sense that it's not flexible enough to apply to all conceivable environments a genuine AGI could navigate, without misleading that AGI. But I should stress that real number objective functions are probably fine for many s
114.
by xamuel 6y ago
>Humans don't receive infinite rewards Maybe you're right, but I don't think it's as obvious as you imply. Consider this question. Can there exist hypothetical rewards x1,x2,x3,..., each one of which is significantly
115.
by xamuel 6y ago
I think you understood the basic gist of the paper quite well, that's a good way of describing it, and thanks for the link. >it speculates that a higher-order number system (such as the hyperreal or surreal numbers) would be able to
116.
by xamuel 6y ago
The argument isn't so much about the type of agent (which I think is what Neural Turing Machines etc. are about), it's about the type of environment. In traditional Reinforcement Learning, environments give real-number-valued rewa
117.
by xamuel 6y ago
Yes you can, but said representations will necessarily be misleading. It's illuminating to consider Big-O notations: why don't we "simplify", since real numbers are so much easier, why don't we declare, e.g., that O
118.
by xamuel 6y ago
Author here. One particularly topical observation (topical because HN has recently featured discussions about ways to merge statistical and symbolic approaches to AI), from Section 4.2.3: certain cutting-edge number systems such as Conway&#
119.
Why traditional reinforcement learning will probably not yield AGI [pdf](philpapers.org)
82 points by xamuel 6y ago|115 comments
120.
by xamuel 6y ago
The article is about simulation results, but I have a mathematically proven result for an idealized abstract species notion, published in an article in Acta Biotheoretica. Definition: An "infinitary species" is a set X of organism

More ›