Neural networks generally don't extrapolate, they interpolate
You could test that if someone were willing to define what they mean by "generally" I suppose. I think it's fairly safe to say that they work best when they're interpolating, like any model, but you can certainly ask them to extrapolate as well.
You are welcome.
I thought not. Your "main point" is based on two logical fallacies. You might be familiar with the saying "two wrongs don't make a right." Your "reply" was a third.
It was based on solving a maths equation.
True.
There's a big and very obvious difference between "scientific research" and "mathematics".
Ehhhhh
Nobody was out there putting clocks on satellites
Technically true, but they were definitely doing experiments. The inconsistencies in Maxwell's electrodynamics and previous physics were the hot topic of late 19th century physics. To the point where various people thought resolving them would put the finishing touches on physics. Even the popular account includes the Michealson-Morely experiment.
Einstein himself says in "On the Electrodynamics of Moving Bodies" (i.e. the special relativity paper):
It is known that Maxwell’s electrodynamics—as usually understood at the
present time—when applied to moving bodies, leads to asymmetries which do
not appear to be inherent in the phenomena. Take, for example, the reciprocal electrodynamic action of a magnet and a conductor. The observable phenomenon here depends only on the relative motion of the conductor and the
magnet, whereas the customary view draws a sharp distinction between the two
cases in which either the one or the other of these bodies is in motion....
Examples of this sort, together with the unsuccessful attempts to discover
any motion of the earth relatively to the “light medium,” suggest that the
phenomena of electrodynamics as well as of mechanics possess no properties
corresponding to the idea of absolute rest. They suggest rather that, as has
already been shown to the first order of small quantities, the same laws of
electrodynamics and optics will be valid for all frames of reference for which the
equations of mechanics hold good.
There were a whole bunch of relevant experiments. Lorentz reviews many of them in "On the influence of the earth's motion on luminiferous phenomena”, published in 1886.
Anyway, the author's point is not that AI can't think because it can't find the consequences of equations. Regular old numerical simulations and logic engines are pretty good at that, no AI required. His point is that AI can't think because it cannot generate ideas out of thin air, presumably the "pure reason" of ancient greek philosophy, and he uses Einstein as an example.
And as a supporting argument he used a fallacy. That's my point.
Crappy people at work. We really need to make greed a punishable offense in some areas.
You mean unlike the US pharma price-gouging, where people pay 20x as much as they do for basically the same product with the same safety in other places? Let's hope so. Americans may find out that most things can actually be treated without sending you into medical bankruptcy.
Indeed. Profit-driven is not enough.
The problem I think is that now the student has become the master, and the west is finding that out.
Like hat has never happened before and nobody could anticipate that....
Japanese cameras and then electronics for one close example. Or look up where "Made in Germany" came from.
Sue Taco Bell!
In other news, the catholic church is suing OpenAI because they had the idea to simply make suicide illegal a thousand years ago and have been using it ever since.
Not sure if it's a trade secret or a copyright case, the news often don't mention the fine details.
That's a good point. Here on
Or not. Kind of like a drug-dealer that prints warnings on their product...
Western pharmaceutical companies have gotten fat and lazy and do not take the risk of doing R&D anymore, with rare exceptions. If China kicks their collective behinds, all the better.
"You know, we've won awards for this crap." -- David Letterman
