blueyes

33 ms·

331.
by blueyes 9y ago
The title is misleading, because this experiment start in late 2016. Sure, they plan to run for 12 years, but right now it's more like a 14-month experiment. Even when the experiment is over, it will have been done on a very small samp
332.
by blueyes 9y ago
This is also a good one: https://deeplearning4j.org/lstm
333.
by blueyes 9y ago
> And there’s some country where it was, like, referendum. It got totally crushed a year, two years ago, maybe it was Sweden. I don’t remember. Fwiw, the country that Sam is referring to here is Switzerland. http://www.bbc.com
334.
by blueyes 9y ago
duplicate: https://news.ycombinator.com/item?id=15894896
335.
by blueyes 9y ago
This is great advice. And it's great copy in itself. Most companies focus too much on what they do, and not enough on who they serve.
336.
by blueyes 9y ago
What is so bad about standardization?
337.
by blueyes 9y ago
It's interesting how Russia and Assange are backing every movement that weakens the West (Trump, Brexit, and now this): http://www.businessinsider.com/julian-assange-catalonia-inde... Russia wants to break apart the bl
338.
by blueyes 9y ago
IBM bought visibility with its ad campaigns, and it's trying to buy credibility with this sponsorship, but I'll bet you anything that the only part of IBM that MIT will use is the brand. The tech will be open-source tools that IBM
339.
by blueyes 9y ago
Actually, Tensorflow is huge in the data science community, but PyTorch is taking over ML research. TF is last year's cool thing. That's one reason why fast.ai just adopted Pytorch for their courses. Tensorflow is surprisingly low
340.
by blueyes 9y ago
IBM has almost zero credibility in deep learning and AI. They haven't hired anyone of note. They haven't produced any novel or influential research in the field in years. And yet they air these cheesy Dylan ads and the rubes fall
341.
by blueyes 9y ago
How to be a blowhard...
342.
by blueyes 9y ago
Peter Thiel's hypocrisy takes my breath away. He pushes to elect Trump and buys himself citizenship in New Zealand. He calls himself a libertarian and builds a company at the service of the surveillance state. He gets two degrees from
343.
by blueyes 9y ago
The best conmen are the ones who fool themselves. In a sense, his sincerity is beside the point. But if he really was sincere, why did he title his book The 4HWW unless it was to sensationalize his work, and draw more attention to himself,
344.
by blueyes 9y ago
For a lot of people, 4WW was a "think and grow rich" kind of pressure cooker. Sure, you may be able to do with if you combine the highly marketable skills of being a certain kind of programmer with a low-rent country. Those folks
345.
by blueyes 9y ago
Sure, but that's $2m of today's compute on today's chips. The constraints on chips and cost are moving, and Google is pushing them.
346.
by blueyes 9y ago
Finding the right life partner requires a series of filters. For me, those filters were: 1) quality of conversation; 2) ability to share work; 3) trust and reliability; 4) physical attraction. Not necessarily in that order, but long-term, t
347.
by blueyes 9y ago
I think a good analogy to teach recursion are humans, or really any living thing, since they too are functions that call themselves. That is, each time you create a new human, it's function is to create a new human, and so on ad infini
348.
by blueyes 9y ago
PyTorch is becoming much more popular among researchers, even if TensorFlow use is widespread among data scientists. Would it be bad if TensorFlow became the default library for deep learning? For companies that aren't Google, yes. Man
349.
by blueyes 9y ago
> spaCY, the leading open-source NLP tool? Sounds like marketing BS. what about OpenNLP and Stanford's for NLP?
350.
by blueyes 9y ago
Fwiw, this dates from 2016. Maybe put (2016) in the title? https://cloud.google.com/blog/big-data/2016/07/understanding... Tensorfire is much cooler: https://tenso.rs/demos/fast-neur
351.
by blueyes 9y ago
A much better media investment than OZY. Glad to see tech money stepping up to back real news organizations in America at a time when we desperately need them to do their work well. More research into the fiasco in Washington, and across Am
352.
by blueyes 9y ago
There is little to object to in this post, but there is also little to praise. While it is an accurate description of the differences between OpenAI and Fast.ai, it doesn't make a compelling argument that Elon and friends should be spe
353.
by blueyes 9y ago
They're secretive about it, a bit like Watson. If they could really beat the benchmarks, they probably would have let us know by now.
354.
by blueyes 9y ago
They've published a few research papers. Few other researchers cite them, so it's hard to argue that their research has been influential, or even significant. In 2013 they announced they had solved/broken CAPTCHA, but refused
355.
by blueyes 9y ago
Never before has an AI startup done so little with so much. That puts their total funding with debt at around $130 million. It's an utter waste. Reminds me of the $15 billion IBM has spent on Watson. The people with the money are very
356.
by blueyes 9y ago
I was actually thinking of the main deep-learning/RL frameworks: TensorFlow, Keras, Theano, Torch, MxNet, Chainer, Caffe. That's where the action is. And if you're talking about AI today, you need to be talking about deep neu
357.
by blueyes 9y ago
The only machine working over at Watson is the PR machine. It's astonishing how much IBM has spent to achieve so little. Per the Jefferies report, IBM spent $15 billion in Cognitive from 2010 to 2015, not including an additional $5 bil
358.
Watson Hype: A Reality Check for IBM’s AI Ambitions(technologyreview.com)
2 points by blueyes 9y ago|0 comments
359.
by blueyes 9y ago
What if they were all necessary conditions of success, and each insufficient by themselves? One of the main reason startups fail is founder conflict (lack of team), and many teams have trouble building something that actually works (lack of
360.
How easy is it to hack into Mar-a-Lago? SUPER EASY(propub.li)
2 points by blueyes 9y ago|0 comments

More ›