tgflynn

68 ms·

691.
by tgflynn 7y ago
That is true historically but I doubt that it is true necessarily. I mean I suspect that starting from what we now know it should be possible to reconstruct calculus (at least for all practical purposes) without reference to infinities or
692.
by tgflynn 7y ago
> Infinities have the opposite effect then you might think, they make things simpler. It's much easier to reason about an infinite list of numbers then to reason about 64 bit numbers. I would argue that a few extra lines in a proof
693.
by tgflynn 7y ago
Yes, my current inclination would be to define truth (in a practical sense) as that which is computable. Everything else belongs to the realm of fiction or fantasy. Note that I don't mean to imply that this is the entirety of my philo
694.
by tgflynn 7y ago
Computers handle calculus quite nicely using numerical algorithms, which don't involve any infinite sets. > And you can't easily only partially include infinity. Sure you can. You just need to use a dx that's small enough
695.
by tgflynn 7y ago
Unexplained downvotes are antithetical to rational discourse. Either downvotes should be abolished or they should require an explanatory comment.
696.
by tgflynn 9y ago
Say you let n be a number equal to all of the memory bits that currently exist on Earth then showed that the size of the smallest boolean circuit solving 3SAT for variable counts near n scaled as 2^n. That would essentially be enough to con
697.
by tgflynn 9y ago
I don't see how it could be unprovable. As far as I know the only way a theorem can be unprovable is if it makes statements involving infinity. You can reduce P vs NP to a question involving only finite sets by choosing some finite n,
698.
by tgflynn 9y ago
I don't know the specifics of your situation so my suggestion may or may not be applicable. That said I would suggest you spend your own time trying to solve some difficult problem that your team faces or designing or prototyping a sol
699.
by tgflynn 9y ago
Link is broken (extra stuff after correct link).
700.
by tgflynn 9y ago
"The LHC use magnets to accelerate protons" - That statement would really confuse someone with a basic knowledge of physics because static magnetic fields can't do work and hence can't accelerate particles. The accelerat
701.
by tgflynn 10y ago
I don't quite fit your profile, I actually have quite a lot of prior experience with machine learning, especially neural networks. However my availability is limited for various reasons that I don't want to post publicly. If you
702.
by tgflynn 10y ago
It looks to me like in Remark 4 he's assuming that each instruction in the program always runs on the same memory location. So that step works only if the program executes linearly, without loops. If that's true it's probab
703.
by tgflynn 10y ago
Yes some natural phenomenon also seems possible. What I'm really questioning is their ETI hypothesis. How do they explain that 100's of different civilizations would all choose the same modulation period ?
704.
by tgflynn 10y ago
From the abstract: Signals having the same period were found in only 234 stars overwhelmingly in the F2 to K1 spectral range. Why would they expect signals from 100's of different stars to have the same period ? Doesn't this su
705.
by tgflynn 10y ago
Because "consciousness" and "subjective experience" are essentially two different words for the same thing. Some people try to define "consciousness" as something other than "subjective experience" bu
706.
by tgflynn 10y ago
> What's problematic about the "say no to consciousness" case? It denies the most fundamental fact of human experience, the existence of that experience itself.
707.
by tgflynn 10y ago
What form does the "not directly intelligible" transcript take ? If it's possible to obtain that maybe there could be a way to generate an intelligible transcript automatically.
708.
by tgflynn 10y ago
It's not a primality test. A primality test determines whether or not a single given integer is prime. The circuit I described checks that the product of two arbitrary input integers is a specific known integer. By solving the decisio
709.
by tgflynn 10y ago
I think any algorithm that could do that would count as an artificial general intelligence (AGI) and I think the current consensus is that we currently have no idea how to create one. What relation an AGI has to NP hardness is unclear. I t
710.
by tgflynn 10y ago
It's worst-case exponential, sure, but it's hard to find that worst-case. It's not that hard. Just generate a boolean circuit for the multiplication of 2 64 bit prime numbers and convert the circuit to a 3-SAT formula. I
711.
by tgflynn 10y ago
Solving any coding challenge is NP-hard problem. To prove that statement you'd have to carefully define your terms. If "Develop a polynomial time algorithm to determine whether any given 3SAT problem is satisfiable" is a va
712.
by tgflynn 10y ago
One of the most clueless things Google does is use general search queries to suggest articles in Google News. Just because I search for something, say a business, on the web doesn't in any way mean that I'm interested in seeing fu
713.
by tgflynn 10y ago
My intent was actually not to take any firm position on the nature of consciousness but rather to present one possible model, that I think many would find plausible, as a counterexample to the progressive replacement argument (ie. the argum
714.
by tgflynn 10y ago
I've never found that argument convincing. It seems to suppose that consciousness is an either/or phenomenon. Either you're conscious or you're not. But what if consciousness is more of a continuum and the intensity of
715.
by tgflynn 10y ago
I think that the brain does something beyond or in addition to information processing because I don't see how you can get qualia or conscious experiences out of information processing alone. However to claim that information processin
716.
by tgflynn 10y ago
Even if all the fonts are rigorously identical the viewing conditions won't be. There will be plates seen at various angles, bent plates, dented plates, dirty plates etc. I think that if you're going to use a net you should try t
717.
by tgflynn 10y ago
OK this isn't a very scientific test but just eyeballing some outputs for a lightly trained net (classification error about 10% whereas a fully trained net can get well below 2%) what I'm seeing is that with the test set inputs ev
718.
by tgflynn 10y ago
On your first point in my experience if you use a net with logistic output nodes and you feed it an input that looks nothing like any of the (labeled) training examples I would expect to see low values for all outputs. I would add the cavea
719.
by tgflynn 10y ago
It sounds like you want to stack another probability on top of a probability. The net gives you an estimate of P(C|x) and then you want a probabilistic measure of how accurate an estimate that is. This starts to get a little difficult for
720.
by tgflynn 10y ago
I think you're describing accuracy, as in prediction accuracy. Confidence is simply how much confidence you have that the predicted class is correct. This is often important because in a real system you often want to know whether to ac

More ›