glass_of_water

I think part of the problem is that it's so difficult to actually measure productivity in software development.

I've found that it's much easier to eat healthy when I regularly do some form of cardio exercise at least every other day.

After working out I naturally tend toward healthy foods, maybe because I don't want to undo all of the hard work.

When I don't work out for a while, my appetite and digestion get messed up and I start boredom eating more.

I noticed the same thing. After consuming a soda a day for even just a few days I start to crave it. But if I stop drinking soda altogether, within a week the cravings go away, and I'm just as happy with water.

But it's not like Lyft would suddenly have 350 full time engineers developing new features. A large chunk of those engineers would be working on building and maintaining infrastructure that AWS provides.

I don't think so, since projection would be like taking the line between the center of the earth and the plane and finding where that intersects the surface of the earth. That doesn't depend on the trajectory of the earth through space.

I don't think the issue is controlled inflation. I think the fear is of rapid inflation caused by governments struggling to make interest payments, which could cause the value of any cash or bond positions you have to effectively go to 0.

Similar reason people find gold appealing I suppose.

That's a really interesting thought experiment, and I don't really know what my answer is.

The following choices assume that more people living an extra 30 years is a good thing ethically, and that there are no devastating effects environmentally or otherwise. The following choices also assume that you'll always have a monopoly on the production of the drug.

1. Assuming that the supply is practically unlimited, an altruistic actor should probably give it away for as close to free as possible. A purely selfish actor should probably try to sell to each individual for as much as they can afford (though this can be hard pricing strategy to enact). I suppose you might actually want to charge less than that even as a purely selfish actor, since you wouldn't want to somehow destroy the world economy by making everyone really poor except for yourself (would this actually destroy the economy though? I'm not sure... probably not relevant for the purposes of this thought experiment).

2. Assuming supply is limited, a purely altruistic actor will have to choose the most fitting recipients. This is a really tough choice. How do you define the most fitting recipients? Those who are most likely to contribute to the betterment of humanity with their extra 30 years? It seems like an almost impossible task to pick out who these people would be. A purely selfish actor should charge the richest as much as they're willing to pay.

If it were me and supply were unlimited, I'd like to think that I'd sell it for as little as possible but enough to still never have to worry about money again. Making $10 million in profit would probably be more than enough (though if you can make more and still help everyone, that'd be even better). If supply were limited, I don't know what I would do.

One thing I just realized though, is that even if supply were unlimited, if you had this godly power, you might choose to withhold it those you deem unworthy, which is a whole other ethical can of worms. I'm guessing this kind of deviates from the intent of your thought experiment though.

What's your answer?

If you're using macOS, you'll need to do sed -i "" ... (put an empty string argument after the -i flag). If the second to last argument on the command line is -i, I think on the version of sed that ships with most distros, sed assumes that the argument to -i is the empty string and that the last argument in the command is the pattern. The version that ships with macOS, on the other hand, always treats the argument after -i as the extension to use for the backup file.

> Note, that the more elaborate "neural net within neural net" approaches involve vast levels of computing power...

This is one thing I worry about, that even if we were to get closer to AGI, the algorithms/system might not feasibly run with our current computing paradigm.

How did you go about learning to compose?

Could you elaborate on how you find it to be a hassle? I find it interesting that the imperfection of most people's memory could be a feature.

> Time spent flicking through flashcards for vocabulary ... seems like a poor substitute for time spent reading

I've been using Anki decks in tandem with native reading material to study Japanese. Whenever I encounter a word I don't know, I add it (along with a sample sentence demonstrating usage) to my Anki deck. Whenever I don't do this I get frustrated by having to look up the same words multiple times (often times a new word is used infrequently enough that I forget it even if I look it up every time I encounter it).

Also, when learning Japanese, there are a lot of things where memorization is very useful like learning to map between English words and kanji and learning the readings of place names and people's names.

I didn't consider that; it's quite possible. Luckily the reaction is pretty mild.

From a nutritional standpoint, I agree that it would be unfortunate if people replaced healthy foods in their diet with the impossible burger. But I think part of your position is predicated on the idea that real meat (specifically ground beef in this case) is healthier. I don't think it's clear yet whether or not that's the case.

Would you be willing to elaborate on the culture change point? Do you think that too few people are concerned about animal welfare to choose the impossible burger over ground beef even if it's competitive price-wise?

Does it have to be one or the other though? What's wrong with funding projects such as the impossible burger until lab-grown meat is economically viable?

I feel like your position is analogous to saying, "I think it's better to focus on nuclear fusion. Why waste time working on renewables?" Of course we'd like to power everything with fusion, but it's not there yet economically or technologically. So in the mean time, it's probably a good idea to also fund intermediate solutions.

> Either eat real veggies or eat real meat.

Could you explain this stance? Precisely what do you find objectionable about other people choosing to eat plant-based foods that happen to imitate meat?

Edit: I see your position against the meat substitutes is primarily health-based. Almost certainly it would be harmful to one's health to consume foods such as the impossible burger frequently, but you should consider that the products that it's trying to replace aren't necessarily any healthier.

I always get this wheezy feeling like a little bit of fluid is building up in my lungs after I eat the impossible burger. I don't know what causes it; the only thing I can think of is maybe the unusually high iron content (at least compared to anything else in a plant-based diet).

What kind of foam did you use? Did you have any offgassing issues initially?

It probably depends on what you make. Nuts and store-bought seitan are indeed quite expensive, but lentils and legumes are quite cheap. Where I live (major US city), tofu is only $1 a pack.

I think that humans in the past have undoubtedly had to eat meat to survive, as do many wild animals, but now that most humans no longer need meat, there's no reason to inflict suffering on other animals.

With the vacuum technique, I imagine you would have to think more carefully about the containers you use (e.g. lids would probably pop off tupperware-style containers).

Also potentially concerning is that whenever the PoW algorithm changes, Obelisk will have a head start, since they're run by the people deciding what the PoW algorithm will be.

If the book crew is better, I'd hate to see what the crew on the show is like. I've only read Leviathan Wakes so far, but I found most of the characters to be really flat. I think the book suffers from the same issue as many other books in the science fiction genre: there's so much focus on developing the setting that the characters are underdeveloped and seem very generic.

Regarding point 2, without a government, what would prevent groups from forcefully taking over the means of production and limiting access?

Beyond how the machine works, I'm wondering how he designed the patterns.

I think the primary advantage of using this is that you get both the usability of writing imperative code (like PyTorch or TF Eager) and the benefits of having a computation graph (easier distributed training, using TPUs and stuff).

What's the breakdown of the $50k/month/customer at the very beginning? Is most of the cost the equipment or leasing the backhaul connection?

Ethereum isn't inherently better as a currency. It still has the same scaling limits as bitcoin, just lower utilization at this point.

How would you characterize the abilities of transfer learning? Not disagreeing, simply curious.

The machine learning is done in the browser with deeplearn.js, so the images aren't being sent to Google's servers.

> I agree it's troubling that Google is taking so long.

Google has absolutely no obligation to clean up after your mess.

You should be grateful for any help they and other search engines give you.