I agree with Yann that our current crop of languages aren't well suited to deep learning, but I'm not sure it's a Python specific problem. I don't think whitespace, threading, syntax etc are the barrier.

It's much more that Python is fundamentally an imperative language, and deep learning doesn't fit into either the imperative or functional category, I really think DL deserves its own category, designed from the ground up for manipulating tensor data structures of unknown shapes.

I haven't come across any language that encapsulate strong typing for tensor shapes, so you end up tracking shapes with pen and paper while debugging your code (and not everything is caught at compile time). Named tensors would be a good start (more information here http://nlp.seas.harvard.edu/Na...) but that wouldn't solve everything. People can't agree on computation graph vs a forward/back functional approach - PyTorch prefers the latter, TensorFlow originally preferred the former (but now seems to be migrating to the latter).

Julia does seem to be the most promising, but that may just be "the grass is always greener" speaking. I haven't played with it myself.

It's a "gated" version of podcasting in the same sense that Medium is the gated version of blogging, yes.

BigCo (Spotify/Amazon/Patreon/etc) have only just twigged about the huge untapped potential for podcasting. I guess YouTube/Twitch caught fire in the video sphere, but podcasts were kind of overlooked as a second-rate citizen.

It definitely feels like a huge number of people listen to podcasts while travelling/doing housework/exercising - i.e. when you're not directly in front of a screen. If Amazon offers a "gated" (and profitable) distribution channel for podcasters in the same way that YouTube/Twitch did for streamers, that's not inherently bad.

Access to listeners is the biggest obstacle for content creators - I'm sure Amazon would be more than happy to boost the market and skim off the top. I assume they will eventually push into audio ads, too - coming full circle to reinvent the radio. There's a natural profit alignment there between Amazon and the creators, so it makes a lot of business sense to do this.

If there's one lesson I've learned about large organizations, it's Hanlon's Razor - never attribute to malice that which can be adequately explained by stupidity.

I lived in China for a number of years, and this has been going on for a long time now (at least, with my ISP China Unicom). Absolute PITA, but that pretty much describes most online experiences in China (with the exception of Taobao, which is head and shoulders above Ebay).

Actually you're probably right - I "know" it's 2016 but it doesn't really feel like it's any different from 2010 for me. I'm getting old.

Slashdot has turned into an incredibly angry, toxic environment over the past 5 years (which is pretty visible in this thread and, IMO, part of the reason why the site was/is suffering), so I dug up my old account and logged in just to give you some positive feedback. It looks like your heads and hearts are in the right place and it would be great to see Slashdot restored to its former glory. SourceForge could definitely become a major competitor to GitHub (for one, I think GH is too expensive) so that's one avenue to go down.

I'm already returning to Slashdot more regularly and it's only been a week since you took the reins - so keep it up.

For some reason I originally read this as "Write them on boobs".

To be honest, I think that's a pretty surefire way of getting people to read about your idea!

I disagree - in my opinion, the calorific content of food as measured a calorimeter is not the same as the calorific content that your body "sees" when you ingest food. The digestive process is affected by the type of food you eat, the hormonal response it produces, and so on. If your body can't break down food, it can't utilitise any of the energy stored in it.

In other words, 100 kcal of sugar is not the same as 100 kcal of red-meat protein. Compare two people on a 2000kcal diet (both taking multivitamins to take general nutritional deficiencies out of the equation). Put one on a icecream, soft drink, doughnut and cookie diet. Put the other on meat, fish, chicken and vegetable diet. There will most likely be a noticeable difference in physique after 3 months.

The other thing about HFCS is that it offers no nutritional benefits. At least with other sugars, your body can use them to replenish its glucose stores. HFCS just gets converted to fat.

True, but this is why modern bodybuilders also concurrently take anti-oestregens (aromatase inhibitors).

It's not just a case of banging up some testosterone and away you go. There is a finely balanced stack of supplements all working in conjunction with one another.

Why did the teacher hit him on the head?

Many things I have read so far point towards insulin having a marked effect on protein uptake post-exercise.

e.g. http://www.jssm.org/vol3/n3/3/v3n3-3pdf.pdf

"Insulin increases muscle amino acid uptake and protein synthesis and reduces protein degradation."

Could you please provide a reference for this? I'm not challenging you, I'm a bit of a gym junkie and have read a great many articles on the role of insulin in protein synthesis and weight gain.

I think I'd find this information very interesting.