i think no one cares. it’s a “you know what you’re signing up for” sort of thing akin to the military, oh and you’re compensated hundreds of thousands of dollars to do so
I’m sure my values are just different, but it is hard to hear people complain about sports CTEs. I can’t get this itch out of my brain that it’s more about signaling that you yourself aren’t the brutish sort that would approve of such violence.
But then again I’m generally suspicious of people strongly condemning the free, legal, and informed choices of others.
And yet, when I watch videos of these guys at 50 barely able to speak, it makes me sad, so I support changes that reduce the likelihood of CTE without sacrificing the nature of the sport. Which yes, is inherently violent.
i think there is room for word vomit overexplaining every little nuance, detail, and caveat as a style of writing especially if one is trying to be through and “rational”, but I think it takes much more skill to condense the meat of a subject
the common unifying theme of most tech companies over the past 15 years has been rent extraction. whether posing as middlemen who VC’d their way to monopoly power ruining both sides of their target market, or taking a common good without attribution (attention, knowledge) tech hasn’t been much a game changer beyond digitizing something and charging everyone more for it
It's surprising to me that you would need to wait 30 minutes for a bus. Are you in a very rural area? Was it outside of normal commute hours? Or is your city's public transportation network wildly underfunded?
There are three buses that go from the main depot to my office. They all arrived in a line back to back to back, picked up zero passengers, and went their merry way.
I watched this happen from one block over, waiting for the stoplight to change so I could cross the street and catch any of them. Unfortunately, since they were all moving in a blob, the next blob was going to be around in 30 minutes instead of the average 10 minutes one might expect from a non-blob arrangement.
But that part of the story was anecdote. The more relevant part was that in the event of disruption of the infrastructure solution, modern cellular radio and digital technology let us ad hoc book a one-off service solution on the fly, and that is so convenient that the people who facilitate it can get away with some massive wealth extraction as a result.
Are you European or American? To a lot of America, that is considered a good numbers. Its worse in many parts of this vastly large country. The country has chosen to invest heavily in road networks. It appears that you can have one or the other but not both without great expense and reform.
I like to remind my European friends that this is what the US looks like placed on top of Europe: https://i.imgur.com/HkRZfKd.png
Its not a perfect illustration but it drives home the point.
That surface area map is a seriously bad argument: We are talking about bus lines here. Local transit just requires local lines. This isn't about train lines crossing through the whole country.
And how come China does a much better job than the US while they have the same surface area? Why aren't you overlapping those two maps?
In our city, what happened was a demographic shift that the bus network never adapted to. The city had a "hub and spoke" model where the vast bulk of buses went into and out of downtown; it's optimized to get suburbanites to their jobs and back home.
But as the city transitioned from a manufacturing / industry / banking to a service economy, the jobs spread out. If your primary business need is an office park to answer phones or crunch numbers or move lightweight product, property in the suburbs is cheaper. So now you have a situation where someone lives in the two o'clock suburb and their job is in the three o'clock suburb... And there's no bus from here to there, there's a bus from two o'clock to the center and back out to three.
... or they could use a car (carpool or owned), which will cut the travel time by a third.
Retooling to the modern needs is going to really goose the cost (especially since nobody wants to tear down the current model at the same time; the people who rely on it do rely on it to get to their jobs and back, still; their job isn't heavy industry anymore, but people still work downtown!).
> the best alternative to rigorous specification is human review. human review doesn't scale well to the volumes of output produced by language models. to make matters worse
when the business model is selling more tokens you get such per serve ice times that lead to “more” thinking, engagement baiting, fluffy narratives, and straight up dark patterns
straight up lying and doomer boomer mentality, sad and gross.
very consistent correlations between migrants being net positive economic providers, more law abiding then native citizens, and just straight up regurgitating racist myths
mass deportations won’t solve anything economic when it’s the same admin hamstringing prices and rates with tariffs, war, and defacto support of the oligopoly
i really wish basement dwellers like you though of something original now and then
i’m not trying to dismiss this article entirely as i did get value around from discussion around verifying
but this coming from a VC firm invested in token producing companies and quoting the “genius” behind openclaw makes me think this is just an psyop of over engineering to get ppl to blindly spend tokens on the LLM slot machine
i have unlimited tokens being a large corp so i’m a bit detached from billing and even general best practices for promoting
but the incentives of these companies to become profitable at any cost slipping into entire new types of dark patterns around token based billing seems gross
- charging for injected prompts and cot tokens
- changing default thinking effort to be higher
- training models to give longer winded answers that don’t say anything more of substance
- refusing to fulfill a request and still charging you
i wonder if you could ever just charge based of each user message and it so how breaks even across short and long replies
reply