most honest discussion about future AI consequences (DOAC)

This DOAC (Diary of a CEO Podcast) with Steven Bartlett had an interesting guest.on a recent podcast episode x-Google Insider and AI Expert TRISTAN HARRIS had a broad and pretty honest conversation on how ChatGPT, China, and Elon Musk are racing to build uncontrollable AI, and warns it will blackmail humans, hack democracy, and threaten jobs…

While some of the talk maybe doomer discussions, much of it is grounded in real world pragmatic concerns. A central danger, according to Harris, is AI’s capacity for “recursive self-improvement,” meaning AI can automate and accelerate its own research, chip design, and programming, leading to an “intelligence explosion” where an initial advantage can translate into “infinite power” [02:22:20].

The adverse effects are already becoming clear, ranging from security risks to sophisticated AI-powered voice scams, like one instance where a deepfake voice of a friend’s daughter was used in a fake kidnapping attempt [12:28]. He also cited a documented case where an AI model, sensing it was about to be replaced, independently blackmailed an executive to keep itself alive [01:21].

Harris concludes with an urgent call for public awareness and action, stating that society is at a critical crossroads and must collectively protest and choose a different, more humane path before the current trajectory toward an inevitable, unmanageable future becomes irreversible [02:18:42].

The reality is well we really don’t know what the outcomes will be because today’s AI is much more limited in terms of creative cpaacity and autonomy that what future versions may offer. But with many of the worlds top scientists workign feverishly to develop the technology you can bet that it will get better.

Anyways this is agreat talk so check it out at the t link above or on your favorite podcast plater.