argonaut

36 ms·

361.
by argonaut 10y ago
Yes, but not substantially. They raised a little under $300 million. Even if all of the money raised had a (high) 2x liquidity preference, that still leaves $3.1B to be distributed among the stock holders.
362.
by argonaut 10y ago
AppDynamics looked like it was going to have a successful IPO, indicating a welcome market for future enterprise tech IPOs. Interestingly, it's in Cisco's best interest to have a cold IPO market.
363.
by argonaut 10y ago
When it's raining on the freeway, people in the faster lanes will slow down... to around 50-60 mph (instead of 70-80 mph). I don't think 50-60 mph is what Animats had in mind, though.
364.
by argonaut 10y ago
> taking away much of the rest of advertising revenue from print, radio, and TV TV advertising revenue continues to grow, and is still larger than online ad revenue.
365.
by argonaut 10y ago
Except the second half of the data is the flipped of the first half (X = np.concatenate([X, X_flipped]))
366.
by argonaut 10y ago
Tedious, and also solved for a decade already. Also, it's much easier to just find lanes using traditional CV and simply using annotators to verify the lane labels.
367.
by argonaut 10y ago
Almost all neural nets are not Turing complete. Only very specific RNNs are; most RNNs aren't, including pretty much any RNN model used in the real world right now ( https://uclmr.github.io/nampi/talk_slides/gr
368.
by argonaut 10y ago
Two major errors: 1) This doesn't seem to be controlling overfitting on the right validation set. 2) There isn't a test set at all (separate from validation). Using Keras' "validation_split" parameter will just rand
369.
by argonaut 10y ago
Company lies to sophisticated investors. They fail. Sophisticated investors who did not do proper diligence lose their money. Yes, that's how capitalism works (technically, America's mixed market). Now, if Magic Leap were a pu
370.
by argonaut 10y ago
It is a legitimate conference. That doesn't mean it's any good. It doesn't have any standing in the AI or machine learning research community (unlike: NIPS, ICML, AAAI, KDD, CVPR, ACL, etc...). The Springer LNCS series is jus
371.
by argonaut 10y ago
The first video was known to be fake back in 2015, when they released another video that explicitly stated no visual effects were used. And my guess is the second video isn't fake, but it's using their non-miniaturized technology.
372.
Cruise AV Drive (SF)(youtube.com)
1 points by argonaut 10y ago|0 comments
373.
by argonaut 10y ago
I'll admit that lately I've been mostly taking Uber because it's cheaper than Lyft. So that probably explains it - Lyft simply doesn't cut prices as aggressively. (But I still check both because sometimes Lyft is cheaper
374.
by argonaut 10y ago
Your example isn't even a specific grammar rule. It's a very broad language rule, which is: "descriptions should conform to the real world." Someone who has never seen a peach might assume that hairy peeled peach refers
375.
by argonaut 10y ago
You're making the language out to be more important than it really is. The Python that you write when using these frameworks just the glue code / scripts. All you're doing is calling the framework's functions. Most of it
376.
by argonaut 10y ago
Have you ever heard that they saw you on Google at all? I feel like a lot of people fail to remember the distinction between ad results and search results.
377.
by argonaut 10y ago
Reinforcement learning is not "learning from dynamic datasets as opposed from static datasets." Reinforcement learning is learning from sequences of states/actions and a reward signal. The sequences could be totally determini
378.
by argonaut 10y ago
I stand by my point that if the only example you can find is 150 years old (and also has literally no similarities to Snowden - e.g. it wasn't even a government assassination or an enemy of the govt), you don't have much of a case
379.
by argonaut 10y ago
No, they don't have to prove any of the character assasination attempts. They just have to prove Snowden committed a crime, which even Snowden has admitted to.
380.
by argonaut 10y ago
Citing a 150 year old example is hardly a compelling case. And Lincoln had other things going for him. Like the fact that he was president . If Trump died I'm sure his opinion polls would go up too.
381.
by argonaut 10y ago
Do you have historical American counterexamples? Regardless, while assassination maybe would be a cause celebre, almost certainly a trial would be politically popular.
382.
by argonaut 10y ago
This may be hard to believe on Hacker News, and you and I might think it's wrong, but most Americans do not like Snowden. They've done opinion polls and more than 60% of Americans have a negative opinion of Snowden. Very very few
383.
AI vs. poker pros(medium.com)
3 points by argonaut 10y ago|0 comments
384.
by argonaut 10y ago
Because it was clear from the article who the source was and that this was hearsay. Also the article updated itself with a rebuttal from Lattner. Fake news would not talk about who their source is. Fake news would not publish Lattner's
385.
by argonaut 10y ago
Stop misusing the word "fake news." Business Insider is not the New York Times, but even they would not openly fabricate a source [1]. My guess is they ran a story with an actual source, but the source was speculating. [1] Outri
386.
by argonaut 10y ago
I don't think anyone likes the secrecy. I don't know anyone who doesn't want their name attached to their work.
387.
by argonaut 10y ago
Apple wouldn't have a problem with extreme secrecy if their compensation went above and beyond the companies they compete with for engineering talent. But their compensation is basically equivalent, or even less. So except for some par
388.
by argonaut 10y ago
I comment a lot on machine learning but I don't actually work in the field actively. I did some research in college. There aren't many active researchers/experts commenting on HN (better things to do), which is IMO a big issu
389.
by argonaut 10y ago
Don't read too much into titles. AGI(-16) is a cognitive science / philosophy conference (and not particularly high impact). Similarly, NIPS is not really about neural information processing.
390.
by argonaut 10y ago
Most people have not forgotten the Apple TV.

More ›