Does it work similar to the noise reduction on flagship phones? Then it does create a feel of artificialness when looking closer, with a tendency for artifacts.
- 0 Posts
- 17 Comments
The key question is if total costs along the pipeline, from requirements definition down to the final quality controlled fully debugged product can be reduced, at real LLM costs (not with the currently vastly subsidised costs).
That is right, it is a tool. But how useful will it be as a tool once it will be sold by token at real costs, where every mistake that tool makes costs money and we are talking here maybe about 10 times higher costs than people currently pay for Claude, at the minimum.
Add to that the question how the use of LLMs affects the career pipeline from junior dev to senior dev.
There not so many tool analogies where the tool is especially good at making things look good, even if they aren’t when you dig deeper.
If it won’t be illegal, and under the current US regime it sure as hell won’t be illegal, it will come, it is just a matter of time. It will be interesting though if they’ll still do it even where it likely will be illegal, like in the EU.
Funny how you forgot to list the results on the question of “impact of elections” or the one about political pluralism. But in either case. Comparing those results between entirely different countries and systems of government is rather difficult to begin with. After all, this is about perceptions, not reality. It would be interesting to see how many people would agree with “My country is democratic” in North Korea.
The claim that ethnic Uyghurs have absolutely equal rights before the law compared to a Han Chinese living in Xinyiang is pretty detached from reality. But even if they had, that doesn’t mean that the law isn’t biased against them to begin with.
But the clearest indication is a >90% satisfaction of people with the federal government. Such country is either utopia, in a massive economic uprise … or not a democracy. China on a Beijing level has a “congress” that is functionally as meaningless as a legislative could be. It is so large that it is by design already pretty impossible to be a functional parliament, and anything but a rubber stamping institution. And so the records also show that it isn’t much more than that. Power is increasingly centralised in one person, de facto. There is not much left of Deng Xiaoping’s reforms on preventing power the concentration of too much into one single person. Xi has increasingly hollowed out the system of Collective Leadership. Naturally, elections or citizen’s opinions on any of that had very little impact on any of that.
What China is doing with the Uigurs is not merely Western propaganda. China is also obviously a dictatorship. No sane person would challenge that. The surveillance state is also hardening and Xi is aiming towards a neo-maoist trajectory in so some regards but without the stone age communism of the Cultural Revolution.
The West, especially but not only the US are moving towards a hardened dictatorship as well but that doesn’t change the situation in China.
All of that is quite distinct from the question if China is a benevolent dictatorship. I would say partially. Sadly a lot of what made China that is currently being dismantled. The previous safeguards of all the power consolidating in one individual have been broken. There are good reasons why those safeguards were previously created. So that development is troubling. Absolute power tends to corrupt absolutely. In those few cases where it doesn’t the reign can lead to a golden age. But if the guy at the top sucks it is incredibly hard to get rid of him (or her), much harder than in a democracy or even non absolute dictatorship.
I have the impression you don’t know how rounding works. The two temperatures you quote would be different even with only full numbers.
In any case, for the case of body temperature, you do exactly that, use as many positions after the comma as you want. (Usually more the one is pointless though as you can reliably measure that anyway)
In physics maybe but plenty of scientific fields where your temperatures are commonly not superfreezing. Using K just bloats up your numbers then, without any benefit. In case you ever need K, just add 273.15 to the temperature in °C. The only difference is the base, the increments are the same. Fahrenheit is much less compatible with Kelvin.
Why would anyone care about -17°C? It is an arbitrary number without any relevance. The only relevance it has to you is if you think in Fahrenheit where it is an arbitrary zero point. Not even 38°C is a number you frequently hear used, unless its seriously hot and it happens to be the ambient temperature. Human body temperature is more relevant, but it isn’t a round number in either of the measurement systems, nor is it identical between individuals either.
That “higher fidelity” argument just makes me wonder if some people don’t know the decimal system. 22.7°C, there you go. Most people don’t need that level of precision but it if they do, they simply add a position after the comma and are done with it.
That “better reference to human interaction” argument just doesn’t hold any water though. The claim that using imperial means you are closer to nature is ludicrous and also horrendously US defaultist. Much of continental Europe was fully metric when people were still so much “closer to nature” and barely anything “was computer” yet, except for some room filling mainframes. Yet people here had no issue with all those metric units.
While imperial is absolutely atrocious at engineering and at scientific applications, SI units work perfectly fine for human reference interactions. Are there tiny differences, that give maybe imperial an edge in some circumstances? Possibly? In a way that it actually matters? Hardly.
This is certainly the case for °C vs °F. Anything finer than °C is below the precision of everyday thermometers and also hard to percept. While increments larger than that can be easily measured and are also perceptible. All relevant environmental temperatures are merely 2 digits, with boiling water at 100. That’s perfectly adjusted to human interaction and reference. Most people don’t need finer “granularity” in everyday life but if they do, they simply include the first position after the comma. This is optional and completely frictionless “ganularity”, when you need it.
I am not saying that Fahrenheit is necessarily worse. It is one of the few imperial units that don’t suck. But it is also not meaningfully better either, just different.
Jiral@lemmy.orgto
Today I Learned@lemmy.world•Energy efficiency of various transport modesEnglish
21·26 days agoThis table is heavily biased against transit. First of all it is based on data from places that have by far and large underinvested inefficient underused transit in cities built for cars and not for transit. Secondly MJ/passenger/ distance is itself heavily biased against transit. Distance travelled is of no value in itself. Getting to places of interest is of value. Transit journies are on aversge shorter because transit oriented corridors allow for more compact urban layouts. Having to drive twice as far because cars need so much space, adds no value to going to the super market.
Maybe that baby is from Lahti?
You probably don’t understand German (well, even if you do, possibly won’t help you either). A pity, this song is a perfect fit of exactly that kind of person your comment describes: https://www.youtube.com/watch?v=hDeFYKmgePQ
As it happens, Austrians do know quite a bit about this kind of people. You may or may not guess why.





It doesn’t get more target group than that. Also, one should terminate using this service as it is doing the opposite from what it claims.