karpathy

32 ms·

241.
by karpathy 14y ago
Personally, I've consistently found that the best predictor of whether or not I was going to enjoy a movie was the NUMBER of ratings, not the rating itself. This also works for restaurants and other things on sites like Yelp. It almost seem
242.
by karpathy 14y ago
So... it is a forum but posts in each thread are ordered based on upvotes instead of time, and displayed in Pinterest format? That's what I gather from 30 minutes of an exhaustive all-links traversal, at least.
243.
by karpathy 14y ago
I ported this to the newly announced Khan Academy CS sandbox :) http://www.khanacademy.org/cs/lorenz-attractor/966338554
244.
by karpathy 14y ago
Don't miss the link to Bret Victor's lecture on responsive programming: http://vimeo.com/36579366 . I found it to be very enjoyable and it further shows what is possible with this paradigm. [minutes 2-23 are most interesting and relevant]
245.
by karpathy 14y ago
"share specific topics only with circles that you know to be passionate about them" Sounds wrong, doesn't it? The circles sharing model is backwards-- why should I have to make guesses about what other people are interested about? I've writ
246.
by karpathy 14y ago
What's going to prevent people using app.net from talking about things you don't care about? There is nothing that forces you to only talk about tech and related, as far as I can tell. The reason this doesn't happen on HN is because the top
247.
by karpathy 14y ago
Here are some more I've collected: Awesome 3D live in-browser visualization of Curiosity's journey: http://eyes.nasa.gov/ [doesn't work on ubuntu :( ] There are Curiosity landing parties going on around the world, see map: http://bit.ly/
248.
by karpathy 14y ago
I took 2 courses on Quantum Mechanics during my undergrad and I'm taking the Coursera course now as well for fun. My biggest quibble with how QM is presented is that it is, paradoxically, never tied into physics. Quantum Physics is really a
249.
by karpathy 14y ago
This sounded more like some movie maker pitching a scene for a sci-fi movie. Except they are actually serious.
250.
by karpathy 14y ago
I'm not surprised that this is a research topic but compared to all the other applications, I still think this will take much longer to get running reliably. They are working on centered, pre-registered faces and stationary cameras. Once yo
251.
by karpathy 14y ago
Nice video, but it is a mixed bag of interesting and implausible. In particular, it's unfortunate that the video had to be mostly wasted on the implausible date scene, with an app reading emotions (such as impatience) and suggesting things
252.
by karpathy 14y ago
Since you're just starting, can you maybe go beyond what Twitter does in some obvious ways? We've learned a lot about social media in the last few years and you're still agile enough to incorporate all the lessons and make things significan
253.
by karpathy 14y ago
Agreed, i spent a while looking for this. Looking at Figure 1B in the paper though suggests that it is the latter. There are no actual molecules bouncing around.
254.
by karpathy 14y ago
I don't believe that the intent is to replace higher education for the tiny fraction of us who are lucky enough to be able to afford it thanks to money, time or location, but to make it much more widely available. Sebastian's quote from the
255.
by karpathy 14y ago
Any idea about what prevents them from combining the data?
256.
by karpathy 14y ago
At the risk of sounding silly a few years from now, I'll draw connection to when personal computers were first coming out. It was hard to predict all the uses we have today, and spreadsheets were thought to be one of the killer apps. Simila
257.
by karpathy 14y ago
I noticed that too. It was a sensational, unexpected, brilliant demo that made me feel wonderful things inside, mixed with reflections about the entire human race, and then it was followed by an underwhelming, slightly silly and slow presen
258.
by karpathy 14y ago
If anyone is interested to read more on this topic, there is another recent, closely related and perhaps slightly more accessible paper ( "High-Level Invariant Features with Scalable Clustering Algorithms" http://bit.ly/KDuN04 ) from Stan
259.
by karpathy 14y ago
Actually, I don't think I agree. It's quite fascinating but the friend I mentioned is not very good at telling the easily automatable tasks apart from those that aren't. On a later occasion the friend came to me again with a different probl
260.
by karpathy 14y ago
There are also many people in biology research who do these kinds of repetitive tasks on daily basis for hours without realizing that much of it could easily be automated. While we're sharing horror stories, a shining example from my experi
261.
by karpathy 14y ago
I first figured out something was fishy when they mentioned that it allows their employees to come to work in shorts.
262.
by karpathy 14y ago
It's a fair point. Doing well in my undergrad classes felt like a game, and I've developed many rules of playing the game to score many points. I think it's fair to say that good grades are some kind of a measure of combination of ambition,
263.
by karpathy 14y ago
The point about dictionary is an important one. When I teach I try to keep redefining the words I'm using over and over again, even if I've defined them already. I feel this is one of the more underestimated complications. Especially in my
264.
by karpathy 14y ago
It's funny because education is actually is backwards here. You're supposed to want to be taught by an expert in the field.
265.
by karpathy 14y ago
That quote was the highlight of the class :) He's not only playing the audience-- In an earlier class (class 3 or so), Peter Thiel justified his views on Stanford vs. Berkeley based on their track record with creating successful companies.
266.
by karpathy 14y ago
Disagree. Users are most vulnerable when they are bored, staring at their feed for last 5 minutes and with nothing else to do or look at, not when they are actively engaged with a specific intent. That's when the ad gets in the way.
267.
by karpathy 14y ago
Few of my observations from this announcement: In Google's vision of the future, your computer is a thin client. There is no My Documents folder, as everything personal lives on gdrive. The Chrome app store becomes the new operating system.
268.
by karpathy 15y ago
I would recommend using AutoHotkey instead of AutoIt. I wrote some bots for fun for a few similar games but skipped Bejeweled in favor of (a simpler) alternative called Diamond Dash. The bot could hold its own but it couldn't beat some of m
269.
by karpathy 15y ago
I'm a PhD student at Stanford working with Andrew Ng, who is known for his work on Deep Learning. I've worked on these networks for the last few years. I think it's great that people get excited about these advances, but it is also easy to
270.
by karpathy 15y ago
Interesting points, the matrix edge cases seem a bit weird. I program in Matlab all the time, but I basically can't imagine leaving it for Julia no matter how good the language is. The main reason being that the single most valuable part of

More ›