Post #4050164

2026-07-24 00:38 UTC

I've not received a legitimate answer to the following #AI question: AI frameworks (Openclaw, Hermes, etc.) seem well-suited for small to large businesses looking to scale to sizes that (maybe) can make existing employees more productive or (probably) allow management to reduce headcount by implementing skills/agents to replace employees. Are there legitimate (as in viable, not in terms of legality) use cases of AI that allow individuals to ramp their own individual skills and toolsets to combat, impede, or poison corporate AI use? Failing that, is there any case where individuals can use AI frameworks to better themselves without falling into the academic trap of being an academic just to increase academic power instead of producing actual results? Honest to god, I would spin up my own small AI instance, but I don't have a business use for it and I don't want to contribute to the rise of skynet. Is there a point?

Replies (2)

@AmbianceAsunder@infosec.exchange 2026-07-24 01:08

@clankgy1@infosec.exchange you bet your ass there is. Artists can poison their images, and others can post content that is also posioned. I’m AFK rn but I can post more in a bit

Open ##4050684
@AmbianceAsunder@infosec.exchange 2026-07-24 02:05

@clankgy1@infosec.exchange https://nightshade.cs.uchicago.edu/whatis.html https://www.anthropic.com/research/small-samples-poison https://arxiv.org/abs/2510.07192 https://arxiv.org/pdf/2506.06518 I’m sure there’s loads of tools out there but, a word of caution, this AI shit is also a festering breeding ground for malware. It’s the new hotness so naturally tools pertaining to it are an easy market to inject malware into. Be vigilant, read the code and if you don’t know how, now’s a great time to learn. Hopefully others can make recommendations too. I think this is an important thread to pull Tag bomb: #NoAI #infosec #redteam

Open ##4051681