Interesting Links for 30-10-2015

[andrewducker]

• How industrialisation broke bread, and the people trying to fix it.
  • (tags: food bread industrial technology history USA viaHugh )
• This new 'sexual consent and tea' video from Thames Valley Police is brilliant
  • (tags: tea consent video )
• So you want to abolish time zones
  • (tags: time viafanf )
• Study of self-driving cars shows other drivers are good at

Comments

[steer]

Why the EU's "Human Brain" project is a waste of money

We should never ever listen to neuroscientists on these matters. Fifteen or twenty years ago they were giving this exact same rant about neural networks research (computer models of extreme simplifications of neurons). They complained that neural networks ignored essentials of how neurons actually worked. They were right, those models did, and it completely wasn't the point. Neural networks are now a completely standard part of the toolbox for a number of eminently practical learning projects. Developments of neural networks are now just a commodity solution for a class of pattern recognition problems. Loads of captchas you do are simply helping train neural networks and it's a standard technique for OCR and voice recognition.

End story: I would expect projects like that one to produce useful tools that will be helpful for society. I expect they will not help neurologists in their research.

[andrewducker]

That makes a lot of sense. Thanks!

[momentsmusicaux]

I wish articles like that NY Times one on wheat had a tl;dr summary. I can't deal with factual stuff that's written like a story :/

[andrewducker]

1) Due to technological advances making it possible, the interesting bits of wheat (which include most of the flavours, but also mean it spoils after several days) were first stripped away and then bred out of it. Along with any of the "goodness" that was in it, because that doesn't make money.
2) People are working to bring that back, by breeding interesting flavours and encouraging local bakeries/restaurants to use them. They've found that local soil, etc. makes a real difference, and it's very exciting if that kind of thing excites you.
3) They're grappling with how to encourage big business to use it without losing its "soul". How do you encourage "local" while simultaneously making something mass market?

[momentsmusicaux]

:)

Thanks!

It's not so much the length I mind, but the rambling story aspect of it (which I realize is a bit fucking rich, coming from me!!!). Non-fiction pieces seem to be written like fiction -- 'It was a Saturday like any other, yet Steve Floofenflaffen was on the cusp of a discovery that would revolutionize wheat.'

[skington]

FWIW, some of the bits I liked the best out of that piece were the story bits - things like how our hero was fed up producing nothing but white-bread wheat, so moved to another job, and was surprised to find out that his clients were growing wheat for crop-rotation reasons, so he got to do fun stuff with wheat after all.

Study of self-driving cars shows other drivers are good

[cartesiandaemon]

So hopefully, they've reached the point where they have fewer brain-farts than humans, but there's still gains to be had in "behaving how people expect so other drivers don't make mistakes". I just hope it's possible to make them seem more predictable (slowing down in advance?) without needing to drive more dangerously (don't really want them to run red lights just because people driving behind can't break in time if they stop...).

Re: Study of self-driving cars shows other drivers are good

[naath]

I don't think you *can* eliminate the "driver behind is an idiot/arsehole who runs into you because they are not looking/stupid/cross" type collision. I'm not especially convinced these cars are being any less predictable than human drivers, but then I think human drivers are frequently unpredictable idiots, and maybe human drivers have some better model of other human drivers than I do, as a non-driver.

[kalimac]

Time zones: Whoever proposed eliminating them? I hadn't heard of such an idea. As the article proves, if we eliminated them we'd just have to reinvent them.

However, it's not as if the current system isn't confusing. I was once phoned by a friend in Australia who thought that California was 6 hours earlier instead of 6 hours later and consequently reached me in the middle of the night.

Speciesization of dogs, wolves and coyotes: Since species evolve from each other, we have to draw artificial lines somewhere to distinguish them. The dog and wolf are a formerly single species that's in the process of separating into two right before our eyes. The creation of a new species combining them and the coyote is exactly what I would have expected from everything I read about species explosions in paleontology classes 40 years ago.

[andrewducker]

Software developers. It's been a rare year where you don't see a software developer or six publish a rant about how stupid and confusing time zones are, and how they don't work properly, and they're fiddly, and they aren't in nice straight lines, and different countries/states/towns all do things their own way, and couldn't we just get rid of them so that my code can be nice and simple and bug-free?

[kalimac]

Where the software developers have a legitimate beef is the way time zones keep moving around and changing. Not to mention the date range of DST. If these things would just stay put, we'd be fine. But eliminating time zones altogether is the wrong solution, like chopping off your foot to solve your gout.

[andrewducker]

Having the same DST on the same dates in all countries would certainly simplify things. When I'm God-Emperor...

And yes.

We've found the mechanism which causes the body to retu

[cartesiandaemon]

"We've found the mechanism which causes the body to return to a "set weight" (in rats)" Ooh, that's really interesting. I'm not sure I can easily even parse the abstract though, anyone able to translate to how definite that actually is?

RE: We've found the mechanism which causes the body to retu

[alitheapipkin]

That is one dense abstract, I can't follow it tonight either! It's something to do with hormone responses that is *consistent* with the idea of a set weight is about as far as I can gather.

RE: We've found the mechanism which causes the body to retu

[andrewducker]

See my reply, where I've done my best to translate based on very-little domain knowledge.

RE: We've found the mechanism which causes the body to retu

[andrewducker]

Some animals and humans fed a high-energy diet (HED) are diet-resistant (DR), remaining as lean as individuals who were naïve to HED. Other individuals become obese during HED exposure and subsequently defend the obese weight (Diet-Induced Obesity- Defenders, DIO-D) even when subsequently maintained on a low-energy diet.

Sometimes people are stable at a low weight level, even if they eat a lot. Sometimes people who eat a lot put on weight, but then become stable at the high weight.

We hypothesized that the body weight setpoint of the DIO-D phenotype resides in the hypothalamic paraventricular nucleus (PVN), where anorexigenic melanocortins, including melanotan II (MTII), increase presynaptic GABA release, and the orexigenic neuropeptide Y (NPY) inhibits it.

There's a bit of the brain that we think mediates this, a chemical whose release is associated, and different other chemicals cause more or less of this mediating chemical to be released.

After prolonged return to low-energy diet, GABA inputs to PVN neurons from DIO-D rats