vletal

What if black holes are NaNs in the simulation we live in. Maybe it was written by a phd student as an experiment for his thesis.

You just arrived to infinite density using a concept of infinite time. Now you need to justify existence of infinite time.

The argument is that there is no infinite amount of _anything_. That is a fictional concept of 'too many parts to count'.

Depending on the task.

When you are trying to solve a crime, you only know that the murderer was a nurse, it's very important to assume a valid p(murderer|gender) as well as p(gender|nurse).

On the other hand, leaking p(gender|nurse) into a candidate scoring algorithm would be a no-no.

What people seem to ask for is very interesting actually. To both learn the underlying statistics AND learn not to use it explicitly in speech at the same time. Assuming next token prediction is the learned function, these two feel a bit contradictory.

It's justs big statistical model. Statistics say that 75% of paralegals are woman and 40% of lawyers are man.

It's completely valid to assume that in 'paralegal married a lawyer because she was pregnant' the lawyer is a man.

That's not a negative gender bias but rather a good educated guess, is it not?

It's not like the model would say that woman can not be lawyers.

Is it just me, or the accompanying text does not correspond to the audio?

Spanish speaking people getting offended by the name "cuLitho"? nVidia was like "sin penesar".

Hard to measure these days. The training sets are so large they might contain leaks of test sets. Take these numbers with a grain of salt.

In Czechia we have group called 'ztohoven' (can be roughly translated to both 'getoutofit' or '100shits') who replaced a webcam facing mountains with their own live stream of fake nuclear explosion.

I remember seeing that aired live on the national TV. Fun times.

https://www.nytimes.com/2008/01/24/arts/design/24abroad.html

Why it works here is that the "analog" representation is not influenced by noise, because it's simulated on digital hardware.

On the other hand, why we use digital hardware precisely because it's robust against the noise present in the hardware analog circuits.

I agree with many of the concerns outlined in the post. On the other hand the humans depicted in the example dialog behave so dumb they could come from a C level soup opera. If I was a bit more close minded I can see myself stopping listening at that point. Undermines the message.

I can see this being useful for old MacBooks with little RAM but I'd never used it.

Closed apps running in background are like services. I do not need the window to be rendered in the background for messages or chat apps. I open them only when needed. Or having a large-ish apps I use on daily basis preloaded in compressed memory. It's just a sensible workflow.

I'd not mind the structure and flashbacks if

- the first act was not so slow and boring, failing to actually introduce the characters - the finale was not so dumb...

At first I felt a little bored. At the end I was disappointed. Only the middle of the movie was kinda fun.

It was low brightness. Thx.

On mobile there is no visible difference in the interactive example on the web. Before/After looks exactly the same (iPhone 13).

Maybe use a zoomed example on mobile?

What?! The 2016 full-self driving video where Tesla drove from home to office was a fraud?!

I thought Musk got spoiled under the influence of his success. He might have been a bad apple all along. :(

Awesome work. Curious whether the whisper model could be ported to tinygrad and how the performance would compare to your implementation.

Any evidence?

Yeah, sure. The issue is with temporal consistency. Meta and Google have some successes in that area.

https://mezha.media/en/2022/10/06/google-is-working-on-image...

Give it some time and SD will be able to do the same.

Surprisingly consistent given how verbose it is.

Sad that iPads do have open bootloaders. Id be happy using my M1 iPad Pro from time to time.

Did that chicken got recently laid off from Twitter?

I'll happily build them a model which estimates it.

Heck, it could be even better, since your "mental" age based on your interests in movies could be more suitable for delivering ads then your actual age.

Awesome! Although I wish to see the intermediate denoised image instead of the progress bar. Just a suggestion.

Or. Maybe, just maybe, he used it to avoid paying taxes like other billionaires do.

https://www.nytimes.com/2018/08/03/business/donor-advised-fu...

What is there problem with small vehicles?

Imagine you are a history or philosophy teacher in 2100.

How cool it is to discuss these kind of issues? What do you think about the "erasing open source community" argument from the historical perspective? What does it have in common with industrial revolution?

Even though the real life implications are real, I find it fascinating and not so simple to unravel.

Economy of scale. You make little from a large population. Or you can go Nestle level evil and straight out exploit the most in need [1]. Plenty of ways for BigCo to make money of developing nations.

[1] https://en.wikibooks.org/wiki/Professionalism/The_Nestlé_Inf...

Exactly. Russia's #1 gas importer is EU.

Maybe internally, but we have to remember how long it took them to return back to the old keyboards without ever acknowledging it was a horrible mistake.

It does not ban anything, it provides opt-in warnings to protect kids from obtaining or sending nudes.