I would believe him. Your right he has being right.
This AI stuff is getting concerning.
From these same Movie Studios recruiting voice actors and these same companies then not letting the voice actors know that their voice will be stolen and reused by an AI software..
If the boogey man of AI gets people to take data security and privacy seriously, that'd be fantastic. I can't see it happening though.
The amount of data collected, and the tools for that collection, have expanded massively since Snowden made the US Government's actions explicit. Yet it's not a fundamentally new development, and people are still broadcasting themselves and seeking attention online.
Anyone posting their videos publicly on youtube or social media probably already has enough material for someone to "
deepfake" porn,
extortion videos etc etc.
AI is being massively hyped, but even the more
pragmatic interpretations of LLMs as tools for abstraction of programming and data manipulation, means that data tasks which previously required specialised knowledge and skills to implement, will effectively be accessible by anyone online.
The fact that the "superalignment" team (Sutskever, Leike, Aschenbrenner) has mostly either retired or been fired (tensions supposedly over Altman's plan to involve the Saudis in developing their own line of AI processors and compute capacity). Open AI has already dropped their stance against developing military implementations earlier this year. Now Altman is talking about transitioning to a commercial model and ditching the non-profit oversight, and then introducing the ex-NSA director to their board, which says it all about the nature of that commercialisation..