imh

41 ms·

331.
by imh 10y ago
More specifically, I hear the top top entry level salaries are going to fresh PhDs from just a small handful of labs.
332.
by imh 10y ago
The Lifecycle of Software Objects is one of my all time favorites. Exhalation is really great too.
333.
by imh 10y ago
It's really a shame. Some fields are undergoing reproducibility crises at the moment, and I'd expect there are innumerable questions that you could survey those communities about, only to return with a strong but irreproducible co
334.
by imh 10y ago
Yeah there's a sense, but a bunch of the negative stories about Uber/Kalanick have originated in our community too. Would Susan Fowler's story have gone viral had it not been for the outrage within Silicon Valley? Including &
335.
by imh 10y ago
>Two questions face the company. One is whether Uber will continue prospering under Mr Kalanick’s leadership. Silicon Valley and its denizens may celebrate his type, but ... Well fuck you too, Economist. That seems like an unnecessary an
336.
by imh 10y ago
Why neural nets though? It seems like total overkill compared to simpler statistics.
337.
by imh 10y ago
Why GANs for this?
338.
by imh 10y ago
Bind is a great name. When you have a function that takes a monadic value of type `m a` (e.g. List[A_] in scala I think) it becomes convenient to program with lambda functions. For list comprehension, you can program like this (python notat
339.
by imh 10y ago
Maybe I phrased it poorly. Being ok or acceptable isn't exactly what I meant re: leveling the playing field. Maybe it needs to be expected or something. I wouldn't expect something so cultural to balance out in under a generation
340.
by imh 10y ago
The point isn't that discrimination being rational makes it ok. The point is that if discrimination is irrational, the best way to fix it is different from the best solution if it were rational. It's an important factor in decidin
341.
by imh 10y ago
>It is true, of course, that physics chooses certain mathematical concepts for the formulation of the laws of nature, and surely only a fraction of all mathematical concepts is used in physics. It is true also that the concepts which wer
342.
by imh 10y ago
If the denominator doesn't matter, then there have been 0 accidents of any kind in the last femtosecond due to humans. Clearly that means autonomous vehicles are a menace, right? Or maybe it's rate of accidents not the absolute nu
343.
by imh 10y ago
All the facial recognition stuff is extremely creepy. Who could think the internet version of privacy is a model for real life?
344.
by imh 10y ago
Doesn't that bring the question back to the 5th amendment on the other 198 counts?
345.
by imh 10y ago
I've never understood the foregone conclusion doctrine. If it's a foregone conclusion, the search should be unnecessary, not a special privilege. IIRC, he also stated that he has forgotten his password (and after 18 months, it has
346.
by imh 10y ago
Also interesting: https://github.com/iassael/learning-to-communicate and the paper: https://arxiv.org/abs/1605.06676 "Learning to Communicate with Deep Multi-Agent Reinforcement Learning&quot
347.
by imh 10y ago
It would only work for a certain type of game. Something story heavy couldn't work for example.
348.
by imh 10y ago
Lots of discussion about whether tips are good or bad. Friendly reminder that people don't tend to tip all races equally [0]. This means that having driver compensation partially in the form of tips will mean black drivers get paid les
349.
by imh 10y ago
The minimum wage stuff is not true in California, at least.
350.
by imh 10y ago
I wrote out the numbers I worked out for a tesla, but I had also done it from basic physics (plus referenced efficiencies) for a gas car, and it's not horribly different. I didn't include it because I have no idea what so ever
351.
by imh 10y ago
I expected that 1kW would be a drop in the bucket for a car, since moving that huge hunk of metal from A to B seems like it should totally dominate. But I worked it from first principles, and sanity checked against Tesla anecdotes [0]. At u
352.
by imh 10y ago
I recently found this one: https://www.reddit.com/r/hmmm/ It's very good.
353.
by imh 10y ago
I wish they clarified whether this claim that humans have a 5.1% error rate is in "listen to this sentence once and transcribe it" or "study this recording however you like and transcribe it." edit: They talk about this
354.
by imh 10y ago
Followup question: Are there good applied books that cover examples in broader fields than just crypto and coding theory?
355.
by imh 10y ago
That's kinda true, but (regularization aside) for standard loss functions it's minimized at the point it's well calibrated, right? Given the scores in the image (97% animal, 90% tiger, etc) they seem to be binary classifiers
356.
by imh 10y ago
The demo picture they chose is interesting. It's obviously a tiger, and is identified as such with only 90% probability. I appreciate the difficulty of the problem and how big of a success it is to achieve even that level of confidence
357.
by imh 10y ago
I made a little lambda calculus with super limited possible ASTs last summer to see if I could get a RL algo to learn to program in it. It failed horribly, but it was fun and I learned a lot.
358.
by imh 10y ago
it's not just "current time" you can ask "(12+431)/42" or "how tall is a poodle" or "when does san francisco city hall close today" or "directions to nearest gas station." These ar
359.
by imh 10y ago
Can we get a title change? About half of the people seem to be commenting on the HN title instead of the article content.
360.
by imh 10y ago
>... consigning herself to being nothing more than a "housewife" for her best adult years. >Finally, what kind of woman has no career goals? Usually one with little or no education. Educated men usually aren't intereste

More ›