pesenti

47 ms·

421.
by pesenti 13y ago
Actually NLP has made great progress in the last few years and there are commercial products now based on it: Watson, Siri, Google Now are good examples.
422.
Why Are We Still Waiting for Natural Language Processing?(chronicle.com)
19 points by pesenti 13y ago|18 comments
423.
by pesenti 13y ago
In case anyone wants to give it a try, here is the paper: http://www.kurims.kyoto-u.ac.jp/~motizuki/Invitation%20to%20... and it looks like he is about to give a lecture in Tokyo: http://www.kurims.kyoto-u.ac.jp/~motizuki/news-english.ht
424.
by pesenti 13y ago
Perelman did lecture on his proof: http://en.wikipedia.org/wiki/Grigori_Perelman#Verification
425.
by pesenti 13y ago
Is it only in life sciences? Remember the cold fusion and the faster than light particle fiascos?
426.
by pesenti 13y ago
The problem is not that research is not reproducible. Researchers are humans and full of biases. The problem is too much emphasis is put on novelty vs. reproducibility. It took a long time for software engineers to recognize the value of te
427.
by pesenti 13y ago
One reason the western world is appearing more indeterminate and becoming less optimistic is because of the diminishing returns of progress. That's the main problem in Europe. A hundred years ago, most people had difficult lives and technic
428.
by pesenti 14y ago
This is no different from fortune telling ( http://en.wikipedia.org/wiki/Forer_effect ). Make a bunch of predictions, some will be accurate, some will be a off, and some will be really bad. But people will remember the accurate ones and ada
429.
by pesenti 14y ago
The problem with Reinhardt-Rogoff has nothing to do with Excel. It was a programing error that could have happened anywhere. The problem is that it took them years to reluctantly share their spreadsheet. Open data research and computation i
430.
by pesenti 14y ago
I am an entrepreneur. And I now work in a big company. It has been 10 months since they acquired my company. As far as I am concerned this article is completely inaccurate. I don't give a damn about compensation. IBM gave me enough money fo
431.
by pesenti 14y ago
Seeing how many people agree with the post I am left wondering who the commenters on HN are. With a large code base and actual customers, comprehensive unit testing is by far the most cost effective way to create and maintain software. This
432.
by pesenti 14y ago
I dont really see why. You can always play Bing against Google and trying to compete in both markets when you haven't made a dent in the first one doesn't seem reasonable. When we switched to their ads our revenue increased 10x which made u
433.
by pesenti 14y ago
No it would not be reasonably good compared to Google. It wouldn't even match what AltaVista used to be (given the relevance algorithms used in these products).
434.
by pesenti 14y ago
Couldn't you just use Google or Bing to serve ads?
435.
by pesenti 14y ago
Hmm... I spent 12 years building my business and you tell me I wasn't passionate. We actually did beat Google but in the enterprise space. The best way to take on Google is to render web search obsolete (the same way Microsoft is becoming i
436.
by pesenti 14y ago
You just weren't born yet.
437.
by pesenti 14y ago
No that was after we sold it.
438.
by pesenti 14y ago
This feels like history repeating. We tried the same thing 10 years ago (remember Clusty?). We got a lot of press with the same kind of silly titles - "Dump Google use Clusty", "Should you ditch Google? Clusty", "Will Clusty be the next Goo
439.
by pesenti 14y ago
You are right. I used "traders" a bit too vaguely. Kanheman's research applied more to stock pickers and wealth advisors. But I think the argument applies to all: none of these people's skills justify the 8 or 9 figure bonuses that Wall Str
440.
by pesenti 14y ago
Why stop at $100M? A company worth $250B could justify $50B salaries following this argument. And the same exact argument can justify outrageous salaries for traders who gamble huge amount of money that's not theirs. Yes, large losses/profi