/sci/ - Science & Math » Thread #14235702

34KiB, 158x197, wwww.png

View Same Google iqdb SauceNAO

Why are all AI talks like this?

Anonymous Wed 16 Feb 18:14:47 2022 No.14235702 View Reply Original Report

Quoted By: >>14235794 >>14236457 >>14236507

https://www.youtube.com/watch?v=PSQ7PoheaRo

You think you're gonna learn something interesting but all panelists just end up talking about bias and diversity 90% of the time.

Anonymous

View Same Google iqdb SauceNAO Machine_stops.pdf, 140KiB, 1x1

Anonymous Wed 16 Feb 2022 18:40:27 No.14235794 Report

Quoted By: >>14235816 >>14236075 >>14236469

>>14235702
Because the race/gender war component of WWIII is coming, and the factions who know they shall lose and whom have no ability to adapt are just going through the metaphorical near-death "follow the light/don't follow the light" routine.

https://www.cs.ucdavis.edu/~koehl/Teaching/ECS188/PDF_files/Machine_stops.pdf

Anonymous

Anonymous Wed 16 Feb 2022 18:47:52 No.14235816 Report

Quoted By: >>14235824

>>14235794
What? And who are these factions? American companies are working for one of them?

Anonymous

Anonymous Wed 16 Feb 2022 18:49:52 No.14235824 Report

Quoted By: >>14235844

>>14235816
It's all on a need-to-know basis, and you don't need to know.

So don't worry about it.

Anonymous

Anonymous Wed 16 Feb 2022 18:53:12 No.14235844 Report

Quoted By: >>14235870

>>14235824
I don't know, sounds like a big ol' LARP to me, is there anything that can confirm your story?

Anonymous

Anonymous Wed 16 Feb 2022 19:04:54 No.14235870 Report

Quoted By:

>>14235844
The future, bro.

Don't worry about it.

Anonymous

Anonymous Wed 16 Feb 2022 20:03:14 No.14236075 Report

Quoted By:

>>14235794
Amazing, thank you.

Anonymous

Anonymous Wed 16 Feb 2022 21:36:39 No.14236457 Report

Quoted By:

>>14235702
Technophiles aren't very creative these days. That tends to be the type of person behind AI research.

Anonymous

Anonymous Wed 16 Feb 2022 21:41:05 No.14236469 Report

Quoted By:

>>14235794
name one single race or gender which cant adapt thats still alive today/hasnt been completely wiped out, with-out the help of the races/genders that could
there are none
because if there were, then, they die

Anonymous

Anonymous Wed 16 Feb 2022 21:51:35 No.14236507 Report

Quoted By: >>14236967

>>14235702
> You think you're gonna learn something interesting but all panelists just end up talking about bias and diversity 90% of the time.

You become a better AI engineer/researcher by improving the performance of your AI. You become a better AI ethicist by talking. A relatively unknown AI engineer is still valuable because his products can be used and combined with other tools. A relatively unknown AI ethicist is pointless because their whole reason for being is to lecture people on how they should act. And how many people do you really need telling how you should act? At most one, because you can’t serve two masters. So there’s really only space for one AI ethicist. So every AI ethicist is competing to become the BEST AI ethicist. But how do you become a better AI ethicist than a competitor? By talking more than they do. So now you have a competition where multiple left wing sjws have already sunk a significant amount of their career into talking and they are never safe in their position so they just keep talking and talking and now every one of them has huge sunk costs they can’t recover and each one of them is facing the reality that if they don’t become the best AI ethicist then all their work will have been wasted so now they’re all in a never ending war of talking to see who can outlast the others by just talking and talking and talking.
That’s why all these AI talks are just radical leftists bloviating while actual AI practitioners are getting things done in the background.

Anonymous

Anonymous Wed 16 Feb 2022 23:50:58 No.14236967 Report

Quoted By: >>14237029

>>14236507
that's very interesting, where are the talks done by actual practitioners?

Anonymous

Anonymous Thu 17 Feb 2022 00:14:47 No.14237029 Report

Quoted By:

>>14236967
They are in front of audiences that are assumed to all have PhDs, and they never have "AI" in the title unless they're by Peter Norvig. Instead you should search for talks based on more technical vocabulary. Natural Language Processing for example, Computer Vision another example. And really those would probably be too high level to be interesting too. Look for lower level concepts. For example: word lemmatization, image rotation invariance, information geometry, etc. Good talks are usually about building blocks like these and not about grand unified sweeping theories.

Capcode	All Only User Posts Only Moderator Posts Only Admin Posts Only Developer Posts
Show Posts	All Only With Images Only Without Images
Deleted Posts	All Only Deleted Posts Only Non-Deleted Posts
Ghost Posts	All Only Ghost Posts Only Non-Ghost Posts
Post Type	All Only Sticky Threads Only Opening Posts Only Reply Posts
Results	All Grouped By Threads
Order	Latest Posts First Oldest Posts First

Your latest searches