acdc4life

7 ms·

1.
by acdc4life 6y ago
> evolution selected for a language skill Yes it did, and we do know a lot about general principles behind, from different disciplines. This is still an early science. NLP research still hasn’t considered many important aspects of langua
2.
by acdc4life 6y ago
Not Google search, but many companies are exploiting cheap labor over seas for different industrial use cases because our algos suck. Not just in manufacturing, there are tech companies outsourcing labeled data for tasks like object detecti
3.
by acdc4life 6y ago
I think this is false. My general opinion of computer scientists and engineers (especially in silicon valley) lack the scientific training that one gets in other disciplines like in physics or other hard sciences. Psychology, linguistics, c
4.
by acdc4life 6y ago
> Comparing your pre-trained brain You’re making sweeping assertions that require domain experts in several different disciplines. The scientific evidence points to the contrary. It is demonstrated that humans have a general ability to l
5.
by acdc4life 6y ago
Well, that’s not entirely true. Alpha go self played 29 million games of Go, which is in feasible for humans. Assuming a game of Go is 5mins, it would take a human 286 years to achieve the same, assuming this person doesn’t sleep, eat or do
6.
by acdc4life 6y ago
I see BERT as non sensical. You need to be scientific and have a mathematical theory on how humans learn language, which is a multi disciplinary task requiring physicists, mathematicians, neuroscientists, cognitive psychologists and linguis
7.
by acdc4life 6y ago
My brain does nlp better than any system out there. I’m also able to ride bicycles and do motor control better than Boston Dynamics. I can also construct and prove math, do physics and code all of this in Matlab and C. My brain can handle a
8.
by acdc4life 6y ago
Slippery slope, with all evidence pointing to the contrary. Apple will never have a monopoly, linux will always be an alternative.
9.
by acdc4life 6y ago
> There should be a name for this kind of ridiculous hubris Pseudoscience? To me this paper is no different from astrology.
10.
by acdc4life 7y ago
The wiki post mentioned that people are right 45% of the time. How does dl and ml stack up? Also not mentioned, how long does it take for humans to start recognizing new races, vs ml and dl?
11.
by acdc4life 7y ago
Why doesn’t the model trained on one race generalize to different races? That sounds inferior to human vision.
12.
by acdc4life 7y ago
I prefer Googles snippet to visiting a shitty site. This is one of the few instances that I fully support Google. If websites want to stay relevant, they need to be more inventive.
13.
by acdc4life 7y ago
"No matter what the function, there is guaranteed to be a neural network so that for every possible input, x , the value f(x) or some close approximation) is output from the network" Okay, so what? You require more and more neuron
14.
by acdc4life 7y ago
There is no "ai", it's the same bull crap we had in the 80s. We haven't had true innovation in this space for a very long time, the current hype was caused by GPU, cloud, and abundance of data for training.
15.
by acdc4life 7y ago
Huh, a quick Google search gave me: https://www.ee.columbia.edu/~dpwe/papers/Wang03-shazam.pdf This was a paper from Shazam from 2003. This is essentially what I proposed, there is no training. Shazam works pretty
16.
by acdc4life 7y ago
> A neural net can capture all those smaller eigenvectors in the signal that are routinely thrown away during traditional feature engineering What on earth are you talking about? >Deep nets are here to stay. Maybe in silicon valley fo
17.
by acdc4life 7y ago
The biggest issue I have is: "To do this we developed an entirely new system using convolutional neural networks to turn a few seconds of audio into a unique “fingerprint.” Why did you pick a neural network? What mathematical propertie
18.
by acdc4life 7y ago
>you do realize that NN/AI is totally state of the art for many tasks? Being state of the art doesn't imply that these things will solve these problems. In ML terms, how do you know that NN/AI isn't a local maxima tha
19.
by acdc4life 7y ago
>Some of the most brilliant people in the world have been working on image recognition, voice recognition etc. and AI is crushing all of their work. This is very true. I take my stronger statements back, MAINSTREAM mathematicians attempt
20.
by acdc4life 7y ago
Machine learning is glorified curve fitting. DL isn't even mathematically sound, back propagation has no proof of convergence. Quantitative math is about extracting natural laws, and mapping them to mathematical structures. You could u
21.
by acdc4life 7y ago
Quantitative math, or applied math isn't based on fitting data to an arbitrary mathematical structure. It's looking at real life, and deriving the mathematical laws that govern what you see. You could have a neural net predict pla
22.
by acdc4life 7y ago
Quantitative math might be the wrong word, maybe applied math? In quantitative finance, you make quantitative models about the world, and build math that realize those assumptions and understanding. A simple example: options are a great fin
23.
by acdc4life 7y ago
Deep learning and machine learning don’t work. Quantitative math will always prevail, as it always has. Unfortunately, mathematical research isn’t there yet. We don’t have models for vision, audition and linguistics. Neuroscience and psycho
24.
by acdc4life 7y ago
“A separate class action lawsuit was filed on behalf of 64,000 tech workers for an estimated $3 billion in lost wages.“ So they went through all this effort to save 3 billion/64000 = $46,875 per employee. Is that how it should be inter
25.
by acdc4life 7y ago
None of these tech companies have the ability to do that. What you are suggesting would mean the death of every social media company.
26.
by acdc4life 7y ago
No one has that kind of AI. It doesn't matter if you are a billion dollar company, the scientific innovation is nowhere close in the present day and age. Facebook is relying on deeplearning and machine learning, which is the current pa
27.
by acdc4life 7y ago
"[They] allow the live streaming of suicides, rapes, and murders, continue to host and publish the mosque attack video, allow advertisers to target ‘Jew haters’ and other hateful market segments, and refuse to accept any responsibility
28.
by acdc4life 7y ago
How long until stadia bites the dust?
29.
by acdc4life 8y ago
This not 100% true. I am not advocating first principles. I am advocating mathematical modeling. Mathematically modeling vision seems pretty much impossible. I do think there are many theoretical parts of topology that might be useful like
30.
by acdc4life 8y ago
If you are interested in how mathematical modeling works I recommend any course that covers the contents of this book: https://www.amazon.com/Mathematical-Financial-Derivatives-Sp... Again, experience beats academia. These

More ›