“AI, Person X has the following interests, connections, and has visited the following websites recently. Write a highly targeted ad, masked as a blog post, that you intend for them to click and optimally buy from, maximising commission. The only goal is making money.”
Now do this 100s of times per day until we extract all their wealth.
Google and Bezos Playbook 2.0
The future influences are impersonating targeted AI bots. Far superior to humans faking it and selling a fake life. AI will whisper into your ear with personalised messages.
That’s the real danger of AI. Not some bullshit take over all machines the FANGG founders float as reality. Even if it ends there, that’s many decades after the above has significant impact first. Arguably greater impact unless curl-tailed, due to its damage to human culture and social behaviour.
Imagine a world war that syphons 1-3% of an enemy GDP using these targeted AI bots, or maybe 5-10% if it targets business decision makers too. No rockets launched. No contagion let loose.
Targeting individuals on mass is the real danger. And it’s about to become cheap and extremely effective.
For the above reasons proof of work and base cost in communicating is likely the best approach to offset this near limitless abuse capability of targeting individuals on mass.
I’d imagine Bitcoin is also part of the solution here too. We need to increase the cost of mass manipulation to above a reasonable threshold that’s possible today by state actors (typically being the largest players).
And governments are going to take the divide and conquer game to the next level.
