It has a strange kinky porn vintage vibe…
Oh, just look at the fingers…
ETA
2 Likes
The various youtube AI videos were fun for about a month… honesty, now they’re just boring, because they all kind of look the same, and aren’t very creative.
4 Likes
Leaders of the world’s most prominent AI companies are being recruited for the Homeland Security Department’s new advisory group.
Officially titled the Artificial Intelligence Safety and Security Board, the group will counsel US government on a wide range of AI-related issues. The guidance would concern not only national security but also electricity consumption (which is a problem on the horizon for AI) and manufacturing, according to the Wall Street Journal.
[…]
3 Likes
2 Likes
4 Likes
Please, take it away!
10 Likes
Europol’s report last summer said it was mostly used for crimping!
2 Likes
3 Likes
Criming autofuckup!
3 Likes
But yeah… criming is the biggest use of AI, no doubt.
4 Likes
Crime is the old fashioned word for “disruption”!
Like does tech have a business model that isn’t crime any more? Break the law until you are too big to prosecute.
5 Likes
5 Likes
7 Likes
4 Likes
You can tell AI is great because all its proponents lie constantly about what it is, what it does, and what it might be able to do. At least with crypto we were spared the inane analogies in favor of it being magic.
8 Likes
Ai is like when a comedian steals a whole load of other comedians’ jokes and puts them into a blender until they come out beige.
6 Likes
Mmm, beigeness.
3 Likes
The giant sandslugs in Dune were a bit ick.
ETA @anon15383236
That song was in my head when I wrote that!
4 Likes
I just (virtually) attended an ‘AI’ meeting in my field. ‘AI’ in quotes because everyone has moved on from the ‘just throw your data in there and magic happens’ to a specific application of neural nets to particular part of our overall problem and hoping they work.
There remains no reason to believe that the “Curse of Dimensionality” is handled well by neural nets. That curse, in short, just means that any problem over 3 dimensions gets basically impossible to do rigourously (without being really, really careful, which neural nets are not). Since “dimension” here usually means, roughly, ‘an input to your problem’, there tend to be many more than 3.
This question was avoided during the entire two days of discussion. That’s disappointing, because mathematicians are professionally adverse to lies, so it tells me people really, really don’t like the answer to the question.
6 Likes
