New here?
Create an account to submit posts, participate in discussions and chat with people.
Sign up
32
Nigger engineering (twitter.com)
posted 1 day ago by derjudenjager on scored.co (+0 / -0 / +32Score on mirror )
You are viewing a single comment's thread. View all
14
TallestSkil on scored.co
1 day ago 14 points (+0 / -0 / +14Score on mirror ) 2 children
The lack of screaming tipped me off. We really can’t trust anything we see anywhere anymore.
m0r1arty on scored.co
1 day ago 7 points (+0 / -0 / +7Score on mirror )
We couldn't trust what we saw before (Once we looked into it), AI just tips the hand a little easier and makes everything shown to us held to a higher state of scrutiny :)
IGOexiled on scored.co
1 day ago 6 points (+0 / -0 / +6Score on mirror ) 1 child
The old adage is that government has the technology about 20 years before the public gets it.

Anyway, 9/11 was pretty crazy.
alele-opathic on scored.co
1 day ago 1 point (+0 / -0 / +1Score on mirror ) 1 child
> The old adage is that government has the technology about 20 years before the public gets it.

The idea is correct, and government ran 'influence bots', the proper (government) name for political chatbots, can famously be attributed to overturning America's once-staunch aversion to seasonal 'vaccines'.

> Anyway, 9/11 was pretty crazy.

AI/ML wasn't involved in 9/11 or the coverup in any way.

Even Assange's famously 'odd' proof of life video from the mid 2010s still had his lapel morphing into his shirt.
IGOexiled on scored.co
1 day ago 0 points (+0 / -0 ) 1 child
Ai/ml has been involved with generating a lot of planning and ops. Imagine grok train/embedded on geopolitical takeover and queried daily by every 17C. It makes the dots and helps connect them.

I did not mean to imply that the building collapse was photoshopped.
alele-opathic on scored.co
1 day ago 0 points (+0 / -0 ) 1 child
> Ai/ml has been involved with generating a lot of planning and ops.

Why do you think so?

It is tough to prove either way, but we have some basic information we can use to rule this out before a certain time. For example, Twitter was created to provide a corpus for language models, and eventually a testbed for these bots.

The reason for the short 140 character limit? Short context length of early bots - the KV cache still costs memory, and the compute to train the bots you are suggesting didn't exist in 2009 (hence why early bots just argued) and certainly didn't in 2001. Amazon was created to provide the compute for these bots (coming from an early investor call recording - Amazon AWS is the reason Amazon exists) and AWS didn't even begin to exist until 2006.

That might be possible now, but anything done in 2001 would have been just good-old manual kikery.
IGOexiled on scored.co
1 day ago 0 points (+0 / -0 ) 1 child
Well yes they were looking for ways to do it in 1922 (Public Opinion - Lippmann) without any technology.

Stable Diffusion isn't much more than averaging and adding probabilities. I'd say that 1956 Dartmouth, John McCarthy made it evident it was an aim.

In a more "its the aliens" way, we can see remnants of the idea from ancient civilizations-- the reading of tea leaves or of bones. Gematria in the Hebe language, the scriptures as interpetable machine code. It's no surprise that occasionally the technology resurfaces. Witchcraft, sorcery, parallel compute.

Toast message