Gregory's Server|
65 comments
  @angelastella Peak LLM was autocomplete on your phone. It’s been downhill from there. @paco You are not impressed by these awesome impositor simulations built on top of LLM? I am impressed. Deeply impressed. I just don't know why we need even more impossitors in this crazy world.  And it ran fine with modest resources. I find this trend away from frugal software extremely disturbing.  @angelastella @paco  @angelastella @eniko @gabrielesvelto @cederbs Plastic is not a problem. Consumer culture is a problem. If we replaced every plastic bottle with an aluminum can, and retained the culture of buying and discarding, it would be even _worse_. Plastic uniquely facilitates consumer culture because it's the only material _cheap_ enough to use once and throw away. @pixx @eniko @gabrielesvelto @cederbs Aluminum would force us to recycle it, and that can actually be done given enough energy. (Which we don't have yet because the fossil fuel industry has been sabotaging nuclear for the last half century, but that's another thing.) @angelastella @eniko @gabrielesvelto @cederbs If we used aluminum at the rate we use plastic, I don't think that we'd be able to do it cheaply at all? (And: plastic recycling is also entirely _possible_, for the most common plastics, both technologically and economically. The biggest problem is that it requires sorting/collection, which makes it a lot harder+more expensive; aluminum can just get melted.) @angelastella @eniko @gabrielesvelto @cederbs (That said, currently, recycled aluminum is extremely freaking cheap. the co-op near me sells 100% post-consumer recycled aluminium foil, for instance :) @angelastella @eniko @gabrielesvelto @cederbs (unfortunately, the labelling says it saves 14,000 _watts_ of electricity vs nonrecycled aluminium foil, which makes it hard to trust any of their claims /me finds contact info to explain the difference between energy and power to a company that probably won't care @pixx @eniko @gabrielesvelto @cederbs Aluminum: the cost of energy would set the point between reuse and recycling. Plastic: I understood it was possible but not particularly efficient. @angelastella @eniko @gabrielesvelto @cederbs (Disclaimer: Speaking from memory, it';s been a while since I researched this) The tech for recycling plastic is pretty efficient, the problem is that you need different methods for each type of plastic, and automated sorting isn't good enough. Which means recycling plastic effectively requiers paying people to sort out the various types and put them in the right places, and if they mess up, it can contaminate the entire batch. @pixx @eniko @gabrielesvelto @cederbs That's it, contamination is the big issue. That and unwillingness to employ enough humans to do things right. @angelastella @eniko @gabrielesvelto @cederbs Which puts plastic recycling in an interesting place: If you do it at home, it can be cheaper than fresh material (e.g. if you grind up PLA from a failed 3d print), by a _lot_. But it scales _really badly_. Aluminium recycling is just recycling: people give you cans, melt them down and make aluminium to sell back to the can factory. (Which is still dumb, since... it's already a can.) plastic recycling, you receive a billion different materials! @angelastella @eniko @gabrielesvelto @cederbs Thing is, PET recycling is easy. HDPE recycling is easy. PLA recycling is easy; ABS recycling... I think is easy too but I'm not sure, actually. But if someone gives you a trash bag full of "plastic," and you have to figure out what's in it?? The PET recycling can be profitable. The HDPE recycling can be profitable. The PLA recycling can be profitable. Plastic recycling can still struggle. That said, plastics recycling _has_ been scaling, slowly! @pixx @eniko @gabrielesvelto @cederbs Let's hope it gets better. Here in Argentina there were attempts to integrate garbage recyclers from the informal economy into the process but I don't know how far it went.  @pixx @angelastella @eniko @gabrielesvelto @cederbs This year, 3 plastic recycling companies went bankrupt in the Netherlands because plastic from the fossil fuel industry is cheaper (thanks to tens of millions in subsidies). @src_esther @pixx @eniko @gabrielesvelto @cederbs Sorry, every time I read or hear about oil subsidies I think about the US military (also a great customer). @src_esther @pixx @angelastella @eniko @gabrielesvelto @cederbs Do you mean the € 40 billion per year in tax cuts for the fossil fuel industry here in NL? Or do you mean extra tens of millions for the plastic? Because the latter wouldn't surprise me at all! @Heidentweet @pixx @angelastella @eniko @gabrielesvelto @cederbs I meant the first but the latter wouldn't surprise me either.  @pixx @angelastella @gabrielesvelto @cederbs micro aluminum isn't building up in peoples brain tissues though @eniko @angelastella @gabrielesvelto @cederbs My suspicion here is that microplastics wouldn't be without consumer culture, either. The _amount of resources_ per person is orders of magnitude higher than makes sense. If we used 1% of the plastic we use today, how bad would the microplastic crisis be? @eniko @angelastella @gabrielesvelto @cederbs Sources! > Unfortunately, the majority of plastic waste is being incinerated, dumped in landfills, and released into the environment, causing significant environmental and health problems (Wang et al. 2020a), with only a tiny percentage that does not exceed 10.0% recycled in the USA 10% is recycled. 90% is _incinerated or released into the environment_. If we used 1% of the plastic we use today, we could easily manage 100% of it, no? @eniko @angelastella @gabrielesvelto @cederbs (yes I know that it's way more complicated and this is just napkin math.) @pixx @eniko @angelastella @gabrielesvelto @cederbs why do people incinerate plastic? Is that better than burying it in a landfill? @angelastella @pixx @eniko @gabrielesvelto @cederbs I know at least some plastics do burn, but I imagine the energy density is pretty low since it’s the made from the stuff that gets removed when refining fossil fuels. And the fumes are awful. But I guess using it to power something is better than burning it because it’s cheaper than shipping it to a landfill  @kranzi @eniko @pixx @angelastella @gabrielesvelto @cederbs also aluminum degrades into aluminum oxide, both naturally occuring minerals that can coexist with nature. Plastic is when you bring poisonous oils from Literal Mordor up to the surface and torture them into unholy abominations like Ken Dolls that never wanted to exist and can't degrade or be consumed normally  @wilbr @kranzi @eniko @pixx @angelastella @gabrielesvelto @cederbs Just don't Google "environmental damage bauxite mining", or do. @geoffl @wilbr @eniko @pixx @angelastella @gabrielesvelto @cederbs of course! I dont think it is "good", i only think it is better! @pixx @angelastella @eniko @cederbs plastic is cheap enough only as long as other materials aren't recycled in meaningful numbers. In Italy for example recycling of cellulose-based packaging is now at 92,3%. This makes new cellulose-based packaging so cheap that it's displacing plastic in the food sector which has historically used lots of it.  @pixx But throwing away aluminium cans after a single use would wreak way less havoc than throwing away plastic baubles after a single use. @riley @angelastella @eniko @gabrielesvelto @cederbs Maybe, but if we did it at the same scale as plastic? Aluminium mining is _horrible_, environmentally. Whether or not it's worse than oil drilling / plastic production per bauble, I don't know and am not asserting; rather, there is _no_ material that, if used the way we use plastic, would not be an unstoppable _crisis_.  @riley @angelastella @eniko @gabrielesvelto @cederbs And don’t forget spam. So much spam. @riley @eniko @gabrielesvelto @cederbs Well, we also devised efficient technologies for such non-problems as killing and torturing people and destroying whole cities. @riley @eniko @gabrielesvelto @cederbs Serves me well for trying to reason by analogy in public. @riley @eniko @gabrielesvelto @cederbs My by now almost forgotten point is: if I don't recognize a problem as such, I don't need any solutions.  @eniko @gabrielesvelto @cederbs Has somebody maybe asked (nicely) some AI how to deal with the plastic? Surely that can't go wrong. @steve @eniko @gabrielesvelto @cederbs we already have the hero we need. @eniko @gabrielesvelto @cederbs the difference between AI and plastic is that we at least thought plastic was a good idea when it was new @me Well, for one, plastics screwed over the notoriously evil Belgian king who butchered zillions of Congo people in the pursuit of rubber, so they're not all bad. Or they would have, had the world community not gotten its act together and very politely told the rich perpetrator of genocide to please consider doing it a little bit less, and not get caught anymore, lest Belgian Congo get nationalised. (Spoiler: they had to nationalise Belgian Congo.) @riley @eniko @gabrielesvelto @cederbs well AI screwed over... umm... uhh... the working class.  @eniko @gabrielesvelto @cederbs but we've still got the old plastic to deal with! at least space out the plastics a little bit |
@eniko @gabrielesvelto @cederbs
Except plastic has real uses where other materials won't do. Problem is using it for everything once and discarding it. LLMs and other generative remixers are a solution in search of a problem.