Yup. Plus the interactions you'd expect for instance in terms of matching style of voice in a normal discussion are missing. That being said it still sounds pretty impressive.
A poor person cannot afford to miss a day at the job, or not care for their children, when it would just be another day not working for a rich one. Not to mention the difference in life expectancy between the richest and poorest populations are 10-15 years. If anything jail sentences should be longer for wealthy people!
Right, the uppercase conventionally means 1024 so it should be kB or ko for 1000.
There are some odd rules, like L for liter because of readability or some other reason. I've also learned in (French) school that we shouldn't use kL, that it doesn't exist. It does exist, but we apparently don't want to use it and use hL instead. Some people thing it's a linear unit and not a cubic unit. We count liters like we count potatoes and we use m3 for volume math.
One of the (maybe more) obvious downside being the increase in mutations this brings, in the order of 1 full generation of mutations for every decade the fathers are older for instance. There are plenty of studies on these issues, notably paternal age genetic disorders and "selfish genes", as well as increase of autism, schizophrenia, mendelian disorders, ....
Increased mutations/mutation rate is not necessarily a downside. Mutations aren't inherently bad, they can be beneficial, and the vast majority actually have no effect whatsoever, as the majority of our DNA is never used. But yes increased mutations means increased chances of bad mutations happening even if it also means increased chances of good mutations happening. Anyhow, it may put evolutionary pressure for longer lifespans. Studies involving fruit flies, and selective breeding them as old as possible for successive generations, show that can be the case. And other studies involving natural selection show that high mutation rates are a good thing (ie they help species survive), when environmental change is high (but high mutation rates are detrimental if the environment remains stable).
I remember seeing a talk by Steve Jones[1] where someone asked a question like this and he said the human species has basically not evolved at all for I forget how long he said but it was at least hundreds of thousands of years. He said specifically if you took the children of someone like this dude[2] and put them in a modern school system they would not perform noticeably differently in any way from a modern child as long as they had decent food etc all the other benefits of modern society.
[1] https://profiles.ucl.ac.uk/7056 (emeritus professor of human genetics and evolution at university college London and the author of a fantastic book on the subject called "In the Blood")
South in French is sud, while noon is indeed midi. However, the two words are the same in Latin (meridies, lit. midday, as in p.m. = post meridiem, etc.), so probably in some of the other Romance languages they do match?
Although the article mentions that this research had few practical applications for beer fermentation, in 1888 Pasteur work with the Champagne makers Joseph Perrier led to critical improvements in bottle explosion rates.
That’s indeed its interface, but there is still a huge training phase during which the model was taught using mostly English, and it explains why some other languages interactions feel like they are some direct translation of English expressions (not to mention higher ways of thinking). Most of the thinking indeed is done during training, and then encoded in the weights which act like a huge cache of thoughts, but there is still a lot of learning on how to use the context when you emit this character (token), the attention layers, which is also a form of thinking that uses whatever new input you are giving to the model and applies a series of transformations (granted, finite in current models) to it.
But maybe the dataset for said language is just too small to produce correct content at all, and that's why ChatGPT is bad, not because "ChatGPT is thinking in English".
Except when it default to answering in English, ChatGPT sounds like french when I talk to it in french.
The bigger the dataset, the best is it's capabilities, training it exclusively with a language don't mean it will be better at this language.
I really like the default choices (floor as a “function”, lots of latex heritage, …). The first example is a bit weird though: the word “The” is not left aligned as it would be in a typesetting system, taking perception into account, but merely bounding box aligned with the word below. In practice the beginning of the sentence looks slightly off to the right.
While we’re nitpicking, the space between the heading and the body in the example (ie. the line height of the heading) is too narrow and makes the first line look cramped.
reply