I've never used arch, but I do love how nixos never breaks, so I can't really let go of that anymore. i do use docker/conda when something is only available that way.
I started using Linux (nixos first, now back to Arch) more frequently one year ago and I have a nice boring KDE desktop. Everything works pretty well and the UX is way ahead of OS X.
In the past I was a full time user of KDE3/4 but I've used plenty of OpenBox, i3, sway et similia. In the last 8 years I ended up mostly on Mac, albeit I've installed Arch Linux on a 2015 Macbook Pro (and it was glorious).
Funnily enough, I realised I'm at that age where I'm working only on my desk, so I don't really need to waste lithium batteries for a device that's always plugged in. And I'm tired of the Macbook wasting CPU cycles and getting as hot as a volcano on proprietary crappy software I can't change.
Apple really needs to start hiring some software engineers; the hardware is pretty good but the OS is terribly bloated.
It's a shame we still don't have non-Apple decent laptops, I hope things will change thanks to the shift to Arm and cellphone makers (who did a pretty good job with the quality of some Android devices).
The datasets should not be used for knowledge but to train a language model.
Using it for knowledge is bonkers.
Why not buy some educational textbook company and use 99.9% correct data?
Oh and use RAG while you are at it so you can point to the origin of the information.
The real evolution still has to come though, we need to build a reasoning engine (Q*?) which will just use RAG for knowledge and language models to convert its thought into human language
You use formal verification for logic and rags for source data.
In other words - say you have a model that is semi-smart, often makes mistakes in logic, but sometimes gives valid answers. You use it to “brainstorm” physical equations and then use formal provers to weed out the correct answer.
Even if the llm is correct 0.001% of the time, it’s still better than the current algorithms which are essentially brute forcing.
I’m still confused as to the value of training on tweets though in that scenario?
If you need to effectively provide this whole secondary dataset to have better answers, what value do the tweets add to training other than perhaps sentiment analysis or response stylization?
Nice thought experiment.
I absolutely hate exceptions and I'd like a language without exception.
They're the goto of our time.
When we have Maybe/Option and Effect/Result, there is really no reason to throw exceptions and having to mentally track where that is being handled.
I'm a bit worried about algebraic effects becoming more popular (and influencing frontend JS - which is already way too complex) because it's promoting throwing "exceptions" to control the flow. All of this to avoid the coloring problem in async/sync? Absolutely not worth it imho.
Those are not the same kind of subsidy. there is quite a lot more ambiguity and uncertainty in the type of subsidy because it is claiming mispricing of externalities (i.e your usage caused x harm and you didnt pay for x harm) rather than the variety the renewables industry mostly gets which is a direct injection of cash through grants and various price supports.
Much of the forced and child labor used in solar panel production, mining, and transport, as with everything else in China's production systems, goes completely unacknowledged and unaccounted for in these navel gazing expeditions.
With all those corners cut to mass produce cheap materials like batteries and solar panels and computers and so forth, China is able to undercut production that utilizes ethical and sustainable production. I don't think you can look at China's production "costs" as legitimate data, and the problem doesn't seem to be one that anyone outside of China will ever fix. The extent of the influence the rest of the world has over the problem lies entirely in our ability to not do business with China.
When you look at all the savings you get when buying electronics and solar panels and infrastructure related products coming out of China, you're getting a human suffering discount. I don't think it's a good thing to include those numbers when considering long term things like fossil fuel dependencies and so forth - let's not bake in the human suffering discounts and at least try to price in human rights and humane labor practices.
It turns out a lot of things are way more expensive when factory workers and shippers and everyone in a supply chain get paid fair wages and work fair hours.
This isn't to say anything in favor of fossil fuels, I just think the immediate plight of China's factory workers might be an important factor relevant to the actual costs in play.
Do you know of any credible sources/references that have attempted to quantify that "human suffering discount?" I'm genuinely curious.
My impression is that China simply has the Solar industry established. It's a bit like semi-conductors in Taiwan, the knowledge, practice, and facilities are present. It takes a significant investment to build factories. To that extent, it also raises the question what percentage of solar panel costs is labor, vs capital investments, vs raw material.
I do not discount the statement/concern. At the same time, I don't think the other end of the spectrum is true where we could say "we could do it too if we also used child labor." I do think it's the case where Chinese manufacturing capacity of solar panels is simply superior compared to any other country. It's a massive capital investment and commitment to enable that much production capacity. At 80% the global production of all solar panels, China has an army of solar production factories.
From what I could find (none of which was satisfying conclusive), it looks like there is a very considerable raw material cost for solar panels, and the manufacturing process is also complicated. [1][2]
Yes, things would be more expensive if workers had more rights. I don’t see how that is an argument against or in favor of any particular energy source.
Further, if workers’ well-being are your main consideration (admirable), we should be moving away from coal as quickly as possible.
> if workers’ well-being are your main consideration (admirable), we should be moving away from coal as quickly as possible.
I'm pro renewables and I'm happy to see the back of coal . . .
however ...
Australia is the second biggest biggest exporter of coal, uses no child labour, is heavily mechanised with a small number of workers compared to tonnage moved, has excellent worker conditions in terms of safety, paid overtime, holidays, etc.
You point appears to be based in some Appalachian romance notion of tunneling out coal with pickaxes and coal carts pulled by children.
I find this an interesting statistics, in 2023 there were 36.5k total people employed in the US coal industry. [1] It's simply just not a lot of people in the grand scheme. That speaks to how industrialized coal production is in the US - it doesn't take that many people to do mountain top removal and drive heavy machinery.
FWIW & for comparison, Circuit city at its peak employed 40k people, that's more people than the US Coal industry employs today [2]
Proponents of renewables often cite the price and use it as an argument against nuclear energy. Nuclear energetics uses highly regulated local labor, giving it an inherent disadvantage against unregulated foreign labor. If renewables rely on underpaid or even child labor, it's not sustainable nor realistic, the numbers in that calculation have to be updated and the decisions reconsidered.
The entire U.S. tax code is one big Central Planning committee. Who should get subsidies? Who should get taxed? The capitalists who celebrated the luxury of American supermarkets over Soviet grocery stores failed to mention the significant farm subsidies given then (and still given).
There’s an old joke of a woman who agrees to sleep with a man for a million dollars but refuses for a dollar (“what kind of woman do you take me for?”), the punchline being that it’s no longer a question of principles but of price.
I guess the point here is that as soon as some amount of subsidies is acceptable, it’s no longer a showdown between unbridled capitalism and a command economy. The question really just becomes one of degree.
Climate change itself is an unintended consequence of pricing the well-being of the commons at zero.
There is a fair discussion regarding regulated vs free market to be had, but people at least need to understand market failures first (externalities being one of them).
Again, literally not true because of how big sites that aren't steam for those games are, and how many games are being made and how much they sell on steam.
Yes obviously no one is saying elden ring will have pornography, but there are a lot of games that did sell or would sell more with more romance and to a stronger extent sex stuff in them.
For example take an RPG, like Witcher, Cyberpunk, etc. They skirted the line but it's obvious that people want to get to that point in the in-game relationships, it is what it is, it might offend certain people but fact of the matter is, I don't think the people offended are a majority.
You can always have that content as a DLC and sell it separately and or to block it from countries that don't allow it, but let's not pretend like it doesn't sell.
Sure, but that's digressing from the key point of the grandparent post about the visual novels - that this median number of "game" revenue is meaningless, because steam sales conflate to entirely different markets i.e. proper games and visual novels, and since the latter outnumber the former but each generally get far less sales (as the market demand is so differently structured) the number isn't informative about the revenue of actual games; and if we want to reason about the game market financials, we have to split the porn in a separate category, as it's too different (money- and sales-wise, it's not about prudeness) to be treated as just one of the genres.
I kinda do. I just want them to be more than a fade to black implication or awkward not even colliding meshes with low res genetalia.
It's a landmine, so no AAA dev is doing any of that, and it's really high tech to do convincingly so few indies can pull it off. That's why 2d works like VN's tend to be the best compromise. a gif loop or a few still frames and a seductive voice can be just enough to sell the scene.
> they're completely different products: visual novels are closer to a slideshow than a game
most western ones. They're "expensive" to westerners, but if you ever buy a decent Japanese VN you can understand the appeal. full voice acting, multiple different scenes and poses, an actual attempt at a narrative (we can argue all day if it's good, but so many western ones have shifted to "haha quirky meta" as a substitute for substance), good music, etc. Some even have some surprisingly addicting gameplay, which makes you question why it's categorized as a "visual novel". But that's marketing for you.
I don't blame you for your opinion, but I assure you it can be done right.
Competition doesn't have excellent games. There aren't that many excellent games released per year, if you have an excellent game just getting noticed is enough.
2. You can't trust LLM for anything. They are a Large Language Model, you could consider LLMs having knowledge to be a bug. Their architecture is not meant to reason.
3. ChatGPT and co really needs a feedback loop with compilers to drastically improve(http://devin.ai/ is working on that)
4. If you pick a popular language you get good results because they trained it more with those. JS and Python are ok, Rust or Haskell are never correct
reply