stephanheijl

I'm loving this, I've been using scaleway as a host for a few months and they've provided great value during that time. As a Dutch national it would be great for my latency to see scaleway add some capacity to Amsterdam.

Their low end offering allows me to get up and running quickly with a new machine and experiment, without touching any existing instances or firing up a VM myself.

It's a shame transferring machines to the new region apparently isn't supported, but since I'm running on docker it shouldn't be too hard to recreqte my current instances.

While I support the general idea, it irks me that this concept also includes tracking of my browsing history ("distributes the money from your subscription to the sites you visit based on the time you spend with them"). I would like to see more transparency with regards to how this information is used and how much of it is actually exposed to the developers of this plugin.

Specifically in the context of the usage (that is, transferring USD), the fees were about 6 times what they would have been if he were to use a bank account. In the big picture I do agree with you, though. 1% fees on individual services are not a lot, even when they stack across different services.

While I agree that those fees are extravagant and the underlying process should have been much easier, it probably would have been prudent to investigate the process first, before starting it. I can't imagine being unaware of fees associated with using Coinbase before perusing their services.

Very extensible API, accessible and widely used programming language (Python), the ability to use both Theano and Tensorflow as a backend and easy to implement non-linear neural networks (where data is split and merged at will) all contribute to this. Using Keras means you will almost never need to implement some custom layer or function, whilst sacrificing very little performance-wise.

This is true. Stars in this context are also more indicative of projects that are attractive to a more mainstream audience. Examples like Deep Dream and Neural style, where laymen visit the Github page and star the project because it's "cool" are prominently featured, while projects like Torch and Theano have had massively more impact on the deep learning world today.

While I agree with the premise that open source could, in theory, serve the long tail better, I think the barrier to entry of pharmacological research is far higher than that of software development. It should be noted that this is not without reason. Controlled, double-blind studies are necessary for development of medicine and they are hard to do without a substantial amount of funds.

This barrier to entry may decrease as biological models become cheaper and more accessible, but even as a hopeful bioinformatician, I am skeptical of this happening in the near term (10-15 years).

To be completely honest, if you're not using the tool Let's Encrypt is creating with the specific intent of making configuration easier, it sounds like a moot point to state that it "might be quite hard to configure it". I'm not saying that your use case is invalid, but starting your comment off with stating that it is hard to configure might throw people off.

I think that when the articles refers to not being universally accepted, it means that Bitcoins do not constitute legal tender. Their value is not guaranteed by any government (which is also one of the reasons the coin is, at least at this point, inherently less stable than most fiat currencies in the world).

The point that was being made, if I am correct, is that you can't convict someone of money laundering if it doesn't involve money, not that the transaction these people hypothetically would've had was OK. If the accused had been trading in, for example, sheep (which also do not constitute money in the United States) the ruling would probably have been the same.

I'm looking forward to seeing this in stores, as I've wanted to build a new machine learning rig for some time. The GTX 1080 just didn't seem like it would do the trick, with ostensibly limited software support and all.

I'm specifically wondering about FP16 handling though. Single precision FLOPS are never mentioned in the blog, nor on the NVidia page. It would be a shame if the FP16 units on this card were gimped in the same way as the GTX 1080...

While this is very intesting, the storage mentioned does seem somewhat small to me. The article compares the area covered by the text to be a little smaller than a HIV (Human Immunodeficiency Virus). However, the genome of HIV apparently stores approximately 9.2kb (kilo basepairs, not kilobytes, so ~18KiB) of information. It also doesn't need to be kept at -196C. Are there any viable tactics for atom-based storage that could rival this density?

So this is basically a repost of the Slideshare on this topic (https://news.ycombinator.com/item?id=10919744 < previous HN thread.) without the pretty landing page around it. It does look like some 5000 new LOCs went into it since then, so that 's positive.

I also like the honesty of this report, mentioning different methods that are similar (even another project called lda2vec) and the disclaimer that this is probably not for everyone. The Python code does make it more accessible however, so I could see myself at least reusing concepts that are implemented here. On the whole this looks like an interesting algorithm for the stated purposes.

Sounds like some of these companies aren't even implementing the automation part.

> But she and another former employee, Alex Gioiella, said the only automated part of the service they saw was the occasional marketing text message.

You'd think that the human workers would at least get served up suggestions by the system if it were actually learning, to make sure the algorithm got "gold-standard" feedback on its dataset.

Especially requests like those for delivery food could be standardized pretty quickly, so that workers would only be handed situations where the "AI" (if there even was any machine learning involved) was uncertain as to what needed to be done.

> But usually, the Hero said, the requests were for pizza or Chipotle delivery.

9/10, but I was already familiar with the algorithm. Useful features to find the nature of the "artist" for me included the following:

- Is the style of the painting applied uniformly over the whole image? Humans will, for effect, leave parts of a painting less emphasized and developed, whereas this algorithm will generally apply a style to the entirety of an image.

- If a painting has a surreal style, the subject is generally also surreal. Human painters distort shapes and forms. This is not done by this algorithm.

- Humans will add contrast to make objects stand out, even if colours are similar to their background. This is something the computers haven't yet completely figured out.

Still, this is a very impressive algorithm.

This page refuses to load if you are blocking tracking web services. If you are using AdBlock or Ghostery, you will need to pause them, or at least disable blocking of Disqus. A tad weird, because the Disqus chat only shows up in the final tab of the entire page...

Now that this is working again, we can see the work-mac device constantly trying to connect to fqtag.com, a site linked to malware/adware behaviour.

Good news everyone, the queen will now be able to alert you if your computer is infected! /s

SIGFOX has certain limits (outside of theoretical and practical bandwidth restrictions) placed on it. radio-electronics.com (http://www.radio-electronics.com/info/wireless/sigfox/basics...) states up to 140 messages of 12 bytes per object per day, with a throughput of up to 100 bytes per second.

The Things network, on the other hand, seems to supply its services openly and freely without limits, using the LoRaWan spec. Semtech (http://www.semtech.com/wireless-rf/lora/LoRa-FAQs.pdf) says it provides between 0.3 and 11kbps of bandwidth, so in theory your speeds can be anywhere from 3x to 111x higher.

It looks pretty cool, but it slowed my browser down to a crawl when I had it open in the background. That seems like something that needs optimizing.