argonaut

44 ms·

421.
by argonaut 10y ago
This doesn't make sense. You're already paying for the content, but sometimes the DRM makes it unwatchable for your platform so you torrent. Oddly enough it seems to me like Netflix et al. have no incentive to cater to you, becau
422.
by argonaut 10y ago
Does the A10 have 12 cores? No, it has 2 (technically has 4 but only 2 can be active at any time). Also it's 4 year old Xeon. If you're going to compare it to a 4 year old Xeon, compare it to the 2 core versions, which will have m
423.
by argonaut 10y ago
1) "Scales badly with the number of images" seems like a research problem to me. 2) In theory almost everything is "just" a compute problem (just enumerate or search over all possibilities - of course this may not finish
424.
by argonaut 10y ago
Those are not the labels for the problem I was talking about (visual SLAM). Those are labels for a different problem.
425.
by argonaut 10y ago
It is heavily bonus centric. In theory it is unlimited, based on how much you make the firm, which is why you could theoretically make millions any given year (rare). My friends in quant finance made more than 200k guaranteed their first ye
426.
by argonaut 10y ago
We were comparing the amount of time it takes to write a textbook versus the amount of time it takes to write all the papers the textbook cites (or relies on, I guess). So I don't think "hardness" and the fact that papers als
427.
by argonaut 10y ago
There is nothing original in textbooks? They just need to cite the relevant papers? I find it perfectly intuitive that it is harder to come up with an original argument/idea than it is to summarize one you've already heard. Papers
428.
by argonaut 10y ago
TIL cashiers are considered salespeople.
429.
by argonaut 10y ago
Tesla does not have the world's experts in visual SLAM. Those people are in academia, or at Google/Microsoft. The 100k car/driver thing is a red herring. Steering angle / brake / acceleration info is not the kind of
430.
by argonaut 10y ago
Hedge funds and quant finance pays way way way more than Google/Facebook/etc. To whatever extent they have trouble attracting researchers it's because you will never be able to publish anything or talk about your work in a f
431.
by argonaut 10y ago
Tesla is doing highway driving and some turns without LIDAR. The point of LIDAR is to help you parse a complicated urban scene (and so far what Tesla has are demo videos of urban driving). Of course you don't need LIDAR for the highw
432.
by argonaut 10y ago
This is a research problem, not a compute problem. Directly trying to replicate what LIDAR gives you (SLAM, a 3d map of your current environment) using just cameras is still unsolved today. And it is an active area of research.
433.
by argonaut 10y ago
Nope, writing papers takes longer. The collective amount of time it took to write the papers (I'm not counting the research time) that a textbook is based on, is much greater than the time it takes to write the textbook. Also, there is
434.
by argonaut 10y ago
Because investment bankers are just salespeople? Nobody is suggesting AI will replace salespeople in the near future. I could see there being fewer grunt analysts in the future, though.
435.
by argonaut 10y ago
To be fair, it seems like you linked to a textbook, rather than a paper. Textbooks have different aims than papers (pedagogy vs. succinctly conveying necessary info to an expert in the field)
436.
by argonaut 10y ago
The fair comparison is comparing the number of "free variables" in the system. This is how machine learning models are compared. You don't compare "number of free parameters in linear regression" with "number o
437.
by argonaut 10y ago
This seems overcomplicated. Theoretical computer science papers just write the algorithm in words, using widely accepted mathematical notation wherever they can (e.g. sigma for sums, set notation, and so on).
438.
by argonaut 10y ago
It scales better than complaining on Hacker News.
439.
by argonaut 10y ago
A parameter != a neuron. A single neuron probably has thousands, if not hundreds of thousands, of "parameters" (of course the vast majority are probably unrelated to learning - but given we barely understand neurological learning
440.
by argonaut 10y ago
Email them politely asking for the code.
441.
by argonaut 10y ago
No, it's not an overstatement. If half the time the prediction turns out not to be true, then there is no relationship - you may as well flip a coin.
442.
by argonaut 10y ago
That's irrelevant. It's a positive filter, not a negative one (I didn't claim the absence of everyone using it indicates lack of success).
443.
by argonaut 10y ago
Yes. You would have no idea of this reading Hacker News threads about Snapchat 4 years ago, though.
444.
by argonaut 10y ago
Yeah, also it's worth pointing out that he finished a PhD in math . That is already genius level mathematics as far as most people are concerned; certainly relative to your average undergraduate STEM student.
445.
by argonaut 10y ago
True, but it's not unheard of for math PhD's to graduate with no refereed publications (certainly they'll have a thesis), so it really was the lack of his advisor's recommendation that killed his career.
446.
by argonaut 10y ago
If you cannot describe something in a way that would allow for a precise formulation, then I don't know what you're trying to say. This is the difference between something that is implementable / testable, and something that&
447.
by argonaut 10y ago
I think it was obvious. I remember being in college during the acquisition and thinking it was obviously a smart deal. And Instagram was pursued hotly, so many tech insiders thought it was a good deal too. People on Hacker News are only n
448.
by argonaut 10y ago
You have to be a little careful here. Neural networks are not "very general computation techniques." A dot product and a rectified linear function (or some other function of choice) are not "general computation techniques&quo
449.
by argonaut 10y ago
No it's not. Most is quite a strong assertion.
450.
by argonaut 10y ago
I don't think there are any serious AGI efforts using OpenAI Universe.

More ›