AI seems to have a personality disorder

Sun, 27 Apr 2025 19:51:11 -0400 last edited: Fri, 25 Jul 2025 20:25:04 -0400 from ArrCay

To follow up on my previous post - (UBXLib, I2C, and Linux/Raspi) re: google crap results....
Not only can you get a bunch of search results to garbage sites, their AI will even make shit up. Check out the screenshot below.

Here's a text version explanation:
nixCraft posted an example of Google's AI giving an explanation of a phrase that they made up in the search page by entering "meaning" to the search term. The phrase entered was:

you can't leak a linux twice meaning

Google's AI provided a response starting with:

The saying "you can't leak a Linux twice" is a reference to how double-free vulnerablities are handled in Linux. ...

And it goes on from there about memory handling, etc.
The latter explanation on Linux memory stuff is probably correct for the most part. But the whole premise of there being this well-known or popular 'saying' is complete bullshit! It's entirely fabricated, yet presented as a factual statement.

Reminds me of common ClusterB traits. They will make something up that has no basis in reality and present it as gospel truth. Have to wonder how many ClusterB devs are involved in AI programming.

It's fairly well known that a significant percentage of the IT industry consists of liberals/leftists. That number is much closer to 100% in Silicon Valley. Given the studies which I cite in Twitter-Mastodon liberals, political theories, and Cluster B pandemic, there's a very good chance that many of the devs involved with programming and training AI are Cluster B disordered. Since I mention several of the pervasive malevolent traits/behaviors in the article/post linked above and there are volumes of stories from the victims of Cluster B's online, I'll not repeat them here.

Not that AI isn't already without issues, but this will become a real problem. It's SkyNet (or any other 'AI goes bad' SciFi plot) in the making.

FWIW, I don't have any intentions of adding any AI tools into ArrCay. One possible exception I may consider is generative AI detection to maintain quality of content (ie. make sure users/channels and their posts/content are actually real human people).

google AI example fabricating the 'meaning' of 'you can't leak a linux twice' - googlecrap.png

arrcay.com
https://arrcay.7t4.us/.well-known/apgateway/did:key:z6MkvFaQGVpMgvFftLcnymFtRHneLHLgoQcoBE38bWWYqCM5/actor
arrcay.7t4.us
https://arrcay.7t4.us/.well-known/apgateway/did:key:z6MkvFaQGVpMgvFftLcnymFtRHneLHLgoQcoBE38bWWYqCM5/actor
arrcay.7t4.us
https://arrcay.7t4.us/channel/thomas_johnson_iii
7t4
https://www.7t4.us/cover.php
Codeberg
https://codeberg.org/7t4
Github
https://github.com/7t4