BrianKrebsLately, I've been a wee bit fixated on a vaguely dystopian thought. It's still kind of a squishy idea in my head. But it has to do with AI's potential for inherited memory, the future of reputation, and the right to be forgotten.
Alan K. Martinez@briankrebs The right to be forgotten is interesting. I'm not experienced or maybe even smart enough in this subject area but thinking about the Right to Be Forgotten, if someone or some entity would like to be forgotten how is AI data scrubbed from various sources to remove that entity? The number of sources could also make that super difficult to manage as the data all funnels into AI data sets. interesting... or am I thinking about this all wrong?
@akmartinez IDK. Was reading a great article the other day that I will have to find the link for about efforts to make LLMs "forget" dangerous information. This is really concerning. You're asking systems designed to gorge on huge amounts of information to ignore some of it. I don't think that's going to end well.
LostInLibrary@briankrebs @akmartinez they are still programmed by and using source material (and exclusion lists) created by biased /imperfect humans. None of this will end ‘better’.