I listened to a guy on a podcast saying that Clarence Thomas was in the Black Power movement when he was young, and that kinda informs his decisions now. Thomas is very pessimistic about black Americans ever gaining equal power in American politics, and thinks black people should focus on things they can control instead (family, business, etc). I guess it's kinda like an ethnic/right-wing version of "dual-power." Also, like a lot of leftists I see on here that seem to have given up on electoral politics.
U.S. Libertarians are typically non-authoratarian right-wingers; unless the authority is business.
U.S. libertarians can be fairly radical (not conservative) in opposing pretty much everything the state does outside of operating a legal framework conducive to business. They support near zero business regulations, and as few laws as possible. A U.S. libertarian would typically be against stuff like bathroom laws, public schools, and the EPA. They typically revere Ayn Rand, and often debate age of consent laws.
Also, there is now a Mises caucus within the Libertarian party that is fascist (not conservative). They are the ones who invited Trump. TBF, the Republican party is fascist and not conservative anymore either.
I don't think anyone I worked with in 2019 has the same job (they were either promoted or switched jobs). For the people working those jobs now, I'm guessing the businesses needed to raise wages to attract new talent, probably more than the ~30% inflation. In regards to family I know that work service or manufacturing jobs, they either switched jobs for higher pay, got promoted, or their union negotiated decent pay increases. Again, I'm guessing most businesses needed to raise their wages to attract workers to replace those that switched jobs; and from the "help wanted" signs I see, it looks like they raised wages more than inflation.
We're close to peak using current NN architectures and methods. All this started with the discovery of transformer architecture in 2017. Advances in architecture and methods have been fairly small and incremental since then. The advancements in performance has mostly just been throwing more data and compute at the models, and diminishing returns have been observed. GPT-3 costed something like $15 million to train. GPT-4 is a little better and costed something like $100 million to train. If the next model costs $1 billion to train, it will likely be a little better.
LLMs do sometimes hallucinate even when giving summaries. I.e. they put things in the summaries that were not in the source material. Bing did this often the last time I tried it. In my experience, LLMs seem to do very poorly when their context is large (e.g. when "reading" large or multiple articles). With ChatGPT, it's output seems more likely to be factually correct when it just generates "facts" from it's model instead of "browsing" and adding articles to its context.
Statistically, some working mothers will answer the phone and finish the survey. If the survey was done correctly, and sampling bias and all that is accounted for, 1800 respondents is plenty to get a good representative picture of Americans. Surveys can work very well; even when using non-ideal survey methods.
Anecdotally, this survey seems to align pretty well with the people I know. E.g. people complaining about a pie costing something like 30% more than 5 years ago, while their income doubled or tripled in that same timeframe. Help-wanted signs in my area seem to be advertising wages around 50% more than 2019 as well.
Republicans almost never fail to fall in line for their party. This is probably why the party is so politically effective. I wonder if it has something to do with their worldview and belief in hierarchy. Or, maybe they just don't want to be exiled and unable to continue their grifts.
you can buy a house even on a median salary or lower.
I think think this very regionally dependant. Median household income and house cost in Austin, for example, is ~$70k and ~$650k, respectively. I grew up in a very small rural town, very far from any cities, and even though houses were much cheaper, they were still unaffordable to most people unless they could land one of the few available union jobs (most jobs available were in manufacturing and paid near minimum wage).
Things may be getting better, slightly, for the median person, but inequality is soaring, and a more dangerous problem IMO. Money is power, so inequality is a direct threat to democracy. It's also inequality, not poverty, that has the largest effect on crime rates, and social decohesion in general.
I guess most of what I've heard from her wasn't "bad," but it wasn't "good," I'd describe it as just "uninteresting." It'd probably start annoying me if I had to listen to a full album of hers, because it's not the type of music I enjoy at all.
I believe he's a devout Christo-fascist. I think only a special kind of devout Christian could come up with the porn accountability scheme with his son. He's a "lying for the lord" kind of person.
Legal immigration in the U.S. is heavily restricted with extremely low caps that aren't even high enough to support our agriculture industry. I personally think this is on purpose so that industry can easily underpay and exploit undocumented workers. And yes, the economy would collapse without the all the undocumented workers; the "age" of our population is quite high. Here's an article from a very conservative think tank (which I normally wouldn't trust) that describes the situation: Why Legal Immigration Is Nearly Impossible.
Anyways, it's futile trying to stop immigration. Desperate people will take desperate actions to improve the lives of themselves and their families. It will only get worse as the climate changes causes drought, famine, and otherwise unlivable conditions in parts of the world.
IDK, I think it's going to be very close. Quite a few people I know who disliked Trump over Biden in the last election now favor Trump. Many are in demographics you wouldn't expect. Most never vote though.
One hypothesis is that having more tokens to process lets it "think" longer. Chain of Thought prompting where you ask the LLM to explain its reasoning before giving an answer works similarly. Also, LLMs seem to be better at evaluating solutions than coming up with them, so there is a Tree of Thought technique, where the LLM is asked to generate multiple examples of a reasoning step then pick the "best" reasoning for each reasoning step.
It was probably trained on this puzzle thousands of times. There are problem solving benchmarks for LLMs, and LLMs are probably over-trained on puzzles to get their scores up. When asked to solve a "puzzle" that looks very similar to a puzzle it's seen many times before, it's improbable that the solution is simple, so it gets tripped up. Kinda like people getting tripped up by "trick questions."
A lot of industrial produced food is cheap because of child, forced, and otherwise exploited labor (undocumented workers, for example). Heavily mechanized farming (mostly used for grains) is cheap because of the vast amount of fossil fuel "energy slaves" used. And that's only cheap because the costs are externalized.
Anyways, growing your own food can definitely be cheaper than buying it. Of course, not if you start plants under lights, build raised beds and fill them with purchased soil, buy organic pelletized fertilizer, or stuff like that. It can be nearly free to grow your own food (if you don't count the cost of your own labor) by saving seeds and intercepting materials from waste streams (wood chips, lawn clippings, manure, used coffee grounds, etc) to "feed your soil."
TBF, compression is related to ML. Hence, the Hutter Prize. Thinking of LLMs as lossy compression algorithms is a decent analogy.