vonnik

51 ms·

511.
Fashion Moves Philosophy(aeon.co)
1 points by vonnik 10y ago|0 comments
512.
How State-Sponsored Blackmail Works in Russia(theatlantic.com)
41 points by vonnik 10y ago|3 comments
513.
Emergency Dialect(reallifemag.com)
1 points by vonnik 10y ago|0 comments
514.
The Secret Source of Putin(vanityfair.com)
3 points by vonnik 10y ago|0 comments
515.
Why We Fight Russia(tabletmag.com)
15 points by vonnik 10y ago|22 comments
516.
Picketty on Productivity: France versus Germany(piketty.blog.lemonde.fr)
2 points by vonnik 10y ago|0 comments
517.
Saint of Saint Bernards(theparisreview.org)
1 points by vonnik 10y ago|0 comments
518.
The Ten-Year Anniversary of Apple TV(stratechery.com)
1 points by vonnik 10y ago|0 comments
519.
by vonnik 10y ago
It would be more interesting to see aggregate metrics by yearly cohort, like rev, valuation, etc.
520.
I Want Wings(roughtype.com)
2 points by vonnik 10y ago|0 comments
521.
by vonnik 10y ago
We should wait for Medium ... to make a responsive web site? These guys. We'd to fix something.
522.
by vonnik 10y ago
Deeplearning4j had done this for a long time. Tensorflow has been slow to serve Windows.
523.
by vonnik 10y ago
> The underlying premise of what you write, seems to be that money is the only yardstick that matters. That's not what I said at all, and it does not reflect what I think, but it does smack of the presumptuous and dogmatic pigeon-ho
524.
by vonnik 10y ago
The absurdity conceals a reality you don't understand yet, which is that VCs often have to make investments not according to their deep understanding of a specific market, since there are too many markets to study, but by using a few s
525.
by vonnik 10y ago
DHH's clock is stopped because he keeps saying the same thing again and again: don't raise VC. When a startup fails that raises VC, he points out how right he was. When a startup succeeds that raises VC, he says nothing. In fact,
526.
by vonnik 10y ago
It's a pleasure to see DHH pat himself on the back each time his stopped clock overlaps with the hour. He shows the palpable schadenfreude of a man, sitting on the sidelines, who won't shut up about the players in the game. But I
527.
by vonnik 10y ago
I don't think Medium understood the online media economy when it began, and I don't think they understand it now. With the advent of the Internet, content creators like newspapers and television networks saw the barriers to compet
528.
by vonnik 10y ago
It does. In a way you face a parsing problem similar to Asian languages that don't use spaces to separate words.
529.
by vonnik 10y ago
The biggest challenge you're going to face is that the state of the art in NLP is not sufficient for chatbots to conduct a prolonged conversation that is either useful or interesting. This will lead to disappointment on the part of you
530.
by vonnik 10y ago
He makes a good point about continuous functions. But actually neural nets are quite good at handling discrete elements like words, likes, atoms in a molecule, etc. Deep learning is part of enormous advances in NLP[0], just as it has set re
531.
by vonnik 10y ago
The Wikileaks-Greenwald-Russia axis stands firm.
532.
by vonnik 10y ago
Benchmarks we ran ourselves show that we're faster than TensorFlow using multi-GPUs for a non-trivial image processing task: https://github.com/deeplearning4j/dl4j-benchmark . That's the best apples to apples
533.
by vonnik 10y ago
Deeplearning4j does this already. It has a huge community, a Scala API and does model import from Keras. It's important to note that Spark is not an efficient computation later -- it's best if used for fast ETL. If you get that wr
534.
by vonnik 10y ago
Fwiw, here's a deep learning framework for the JVM that some Clojure people are using: https://deeplearning4j.org/ I bring it up not just because LISP was conceived as a language for AI, but because of the points about
535.
by vonnik 10y ago
Actually, opinions about AGI are quite mixed. People like Ng are more conservative; people like Schmidhuber are more aggressive in their predictions; both are eminent researchers. RE: Siri and Alexa - Industry deployments of AI are a laggin
536.
by vonnik 10y ago
Posted previously: https://news.ycombinator.com/item?id=13266506
537.
by vonnik 10y ago
There is too much wrong with this post to address all his points. It's just more noise in an already noisy space. To be clear, I didn't attack the author. I simply asserted that he is speaking from ignorance as evidenced by his st
538.
by vonnik 10y ago
This guy has no idea what he's talking about, and the original post of which the linked article is an update was published in 2014. That's a lifetime in AI research, which is moving very fast. Real advances are happening monthly.
539.
by vonnik 10y ago
There are some valuable resources listed here. The field of deep learning is moving really fast, and a lot of that research is available on arXiv. Andrej Karpathy built a tool that surfaces interesting papers: http://www.arxiv-sa
540.
by vonnik 10y ago
This is a great piece, and it puts most coverage of Malheur to shame. The one pang I feel after reading journalism like this is that it will not reach the people it should. It may increase our understanding of them, but it doesn't incr

More ›