noosphr

12 ms·

211.
by noosphr 5mo ago
>Without all that might be $175/user/month but you're not shipping apps with just vi and bare gcc. You're right, Linus uses Emacs.
212.
by noosphr 5mo ago
Violence is defined by the World Health Organization in the WRVH as “the intentional use of physical force or power, threatened or actual, against oneself, another person, or against a group or community, that either results in or has a hig
213.
by noosphr 5mo ago
And yet we don't let people dump lead in the drinking water any more.
214.
by noosphr 5mo ago
Killing you slowly is very much violence.
215.
by noosphr 5mo ago
You should work in the Central African Republic coltan mines if you think anyone has a leg to stand on.
216.
by noosphr 5mo ago
Help us Xi Jinping, you're our only hope.
217.
by noosphr 5mo ago
They may be better than humans at reasoning but they are substantially worse than the first generation logic programs from the 1950s.
218.
by noosphr 5mo ago
To this day frontier models think that A and not B means A and B when the sentence gets pushed far enough back in their context window. The context length that model can reason over without obvious errors is much smaller than the advertised
219.
by noosphr 5mo ago
There can never be too many modifier keys: https://xcancel.com/octonion/status/1341113219142828039
220.
by noosphr 5mo ago
Maybe. Just means that the conventional wisdom was wrong and substantially over training can be a good thing. No one I knew at the time suspected that, including the people who wrote the paper.
221.
by noosphr 5mo ago
>We show that a variety of modern deep learning tasks exhibit a "double-descent" phenomenon where, as we increase model size, performance first gets worse and then gets better.
222.
by noosphr 5mo ago
>For example, getting good performance on a dataset with deep learning also involves a lot of guesswork. But, if your training loss is way lower than your test loss, you're in the "overfitting" regime, and you're wast
223.
by noosphr 5mo ago
We support extreme portability for sufficiently large values of two.
224.
by noosphr 5mo ago
The if in there is doing a lot of heavy lifting.
225.
by noosphr 5mo ago
This isn't an indictment of how good AI is but how poor our tools are. We had gui makers in 1996 that made slop which allowed you to iterate in real time. They didn't need a datacenter worth of compute and a nuclear reactors worth
226.
by noosphr 5mo ago
That is a problem with US society. One that deep fakes can solve by allowing plaslsible deniability for any image or video.
227.
by noosphr 5mo ago
And to a good Stalinist it was the dignity of the USSR. Something millions died to protect in the great patriotic war. Again, symbolic violence is worse than real violence to a lot of people.
228.
by noosphr 5mo ago
Go ahead. It's not like I can stop you from imagining it. Adding a GPU to the mix does make it any worse given that you already thought about it.
229.
by noosphr 5mo ago
https://en.wikipedia.org/wiki/The_Right_to_Read You will have no compute and you will like it. The top comment here is about arresting the children for distributing child porn. We really should have kept the normies ou
230.
by noosphr 5mo ago
Back in my father's day you'd get executed for drawing a dick on Stalin's forehead. It's nice to see people still care more about symbolic violence than actual violence.
231.
by noosphr 5mo ago
I guess the google bot is now on by blocked list. There should be a public corporation, like Wikipedia, that starts scraping the Web and provides an api for anyone to access.
232.
by noosphr 5mo ago
That project had nothing to do with the freenet that ran after 2006 either. It's not the first ground up rewrite with major breaking changes using the same name.
233.
by noosphr 5mo ago
>The tool-call history collapse is a problem I'd pay real money to have solved cleanly. It's general attention collapse and it happens everywhere once you start noticing it. The simplest example, which even frontier models fail
234.
by noosphr 5mo ago
>There are jobs, but not the ones most people want. There is no such thing as a shortage in labor markets. There is just people not willing to pay the correct price.
235.
by noosphr 5mo ago
Our brains are becoming more plastic by volume than neurons. I'm not sure we have a leg to stand on when lecturing the past.
236.
by noosphr 5mo ago
Germany's current political course is a replay of the Weimar Republic. Including the people wagging their fingers about the dangers of fascism. You'd think the people telling us not to repeat history would read a history book.
237.
by noosphr 5mo ago
Politics junkies are the least human people around.
238.
by noosphr 5mo ago
I can build a capacitor bank in my garage that produces more power than a nuclear power plant. For a millisecond.
239.
by noosphr 5mo ago
We have no reason to expect anything of the 'rate of discoveries' because that is completely independent of anything to do with productionizing them. For all we know it becomes negative because all the people who understood how to
240.
by noosphr 5mo ago
Only so long as you can keep inventing the next sigmoid in the stack.

More ›