zackchase

6 ms·

1.
by zackchase 6y ago
That's not quite right. Salary bands tend to be guidelines, not constraints. And in ML there are plenty of people getting paid far out of band.
2.
The Superheroes of Deep Learning(approximatelycorrect.com)
7 points by zackchase 6y ago|0 comments
3.
Habits of Highly Effective Data Scientists(approximatelycorrect.com)
4 points by zackchase 6y ago|0 comments
4.
by zackchase 8y ago
I assure you, people have pointed it out...
5.
OpenAI Trains Language Model, Mass Hysteria Ensues(approximatelycorrect.com)
152 points by zackchase 8y ago|114 comments
6.
by zackchase 8y ago
Thanks for sharing! Happy to discuss here if anyone wants to engage in some (pre)debate. Also happy to butt out :).
7.
by zackchase 8y ago
These are excellent questions. I often ask Google researchers, "How Deep is the Mind?" They don't seem to know. You see, it all comes down to democratizing the blockchain, and the only way to do that is with AI in each and ev
8.
by zackchase 8y ago
Thanks for the interest in JnJ AGI Research! As inter-global head of research I'm happy to answer any questions the HN community may have!
9.
ICML 2018 Registrations Sell Out Before Submission Deadline(approximatelycorrect.com)
4 points by zackchase 9y ago|0 comments
10.
by zackchase 9y ago
Flag-planting is relevant because the practice of flag-planting (and other abuses of the pre-print) is what has prompted some of this over-reaction.
11.
by zackchase 9y ago
Author here: Grisha Perelman's proof of the Poincaré conjecture has never been (by him, to my knowledge) submitted to or published in any journal. He decided, as is his right, that he could not care less about the professional communit
12.
Do I really have to cite an arXiv paper?(approximatelycorrect.com)
9 points by zackchase 9y ago|0 comments
13.
DeepMind Solves AGI, Summons Demon(approximatelycorrect.com)
3 points by zackchase 9y ago|0 comments
14.
by zackchase 9y ago
It's problematic that the "experts in AGI" are generally self-designated, and have mostly produced little that is concrete, either in the form of theory or engineering. We should be concerned about long-term outcomes of techn
15.
by zackchase 9y ago
While I don't take exception with the word machine learning (it's reasonably well defined), I agree with most of the points here. Some are grossly exaggerated (e.g. needing petabytes of data to do machine learning). In my research
16.
by zackchase 9y ago
As someone with an economics background (BA), and medical research career (concurrent with ML), I can say (subjectively), that this feels different.
17.
by zackchase 9y ago
I generally agree with this. I used machine learning almost exclusively in the article. Yet "AI misinformation" seems more appropriate in the title since the misinformation, generally, is in reference to "artificial intellige
18.
by zackchase 9y ago
Some strident Singularitarians joined the comment thread.
19.
by zackchase 9y ago
Some strident Singularitarians have joined the comment thread.
20.
Dance Dance Convolution(arxiv.org)
2 points by zackchase 10y ago|0 comments
21.
by zackchase 10y ago
The performance improvement over
22.
by zackchase 10y ago
The performance of multilayer perceptrons and linear models is significant. The paper additionally allows the MLP and linear models to "cheat" by giving them heavily hand-engineered features incorporating extensive prior knowledge
23.
by zackchase 10y ago
The two statements here may be factually true. 1) You had a hard time. 2) Your female classmates, all else being equal, did better. But there's something unsettling about the implication. Namely, by putting these two assertions togethe
24.
by zackchase 10y ago
These are strong arguments. Unfortunately, while sometime interpretability addresses a fundamental (not frivolous concern), it's also true that often interpretability is sometimes contrived to serve some irrational bias or political mo
25.
by zackchase 10y ago
I have good reason to believe the paper's author agrees with this argument.
26.
The Neural Network That Remembers(spectrum.ieee.org)
3 points by zackchase 11y ago|0 comments
27.
by zackchase 11y ago
Hi smackfu. So one wonderful virtue of this technology is that it can handle messes quite gracefully. Part of the fun is to see that it can learn to stay on topic - to decide it's talking about a raspberry beer and continue to talk abo
28.
by zackchase 11y ago
Hi larsga. Thanks for your interest. To begin, yes the concatenated input RNN is much better. A Markov chain could not do this feat (or make comprehensible text at the character level generally). Occasionally the reviews contradict themselv
29.
by zackchase 11y ago
Hi Ian, First off, thanks for your interest in our work! The tech is a homegrown recurrent neural network (deepx) which is available via pip. https://pypi.python.org/pypi/deepx/ We use theano for compilation to GP
30.
by zackchase 11y ago
Hi, thanks for posting our work! The system generates reviews at the character level based on both category and star rating.

More ›