(02-07-2025, 01:52 PM)Michigan Swamp Buck Wrote: Ninurta,
I had the same thoughts on open source, plus, this data breach from its use makes it sound shoddy.
Perhaps this has merely shown how inflated the market value of AI has become when an unsecured open-source program can toss a monkey wrench into the works.
Actually, this makes me reconsider the idea of using my own AI network. I wouldn't mind trying an open-source AI app at home, one that would have my personal interests in its artificial mind.
The AI itself and the data breach should probably be considered as two separate cases. The AI would live in code on the website, but the stolen information would probably live in an unsecured or too lightly secured database separate from the AI to track logins when folks create accounts.
There's also the possibility that it was a honeypot meant to draw folks in with a particular interest in AI in order to mine their information for whatever reason as a targeted audience, and so the login info database was purposely insecure.
But in any event, the AI code is likely discrete from the login database that was stolen. As you probably already know, open source code is often published so that folks can specifically peruse the code for weaknesses, and it's unlikely that they would purposely post weakened code in the clear if they were trying to storm the market.
.