4 ms·
It seems inevitable to me that as ML-based algorithms get better at exploiting the quirks in our monkey brains and keeping us on platforms for longer and longer
by 666lumberjack 8y ago
It seems inevitable to me that as ML-based algorithms get better at exploiting the quirks in our monkey brains and keeping us on platforms for longer and longer periods, somebody will need to develop a counter-algorithm that lives in a browser extension or (somehow) on your device and is able to recognize unhealthy engagement patterns and make you aware of them.
Actually implementing such a thing in a way that's performant and trustworthy and marketing it to people who maybe aren't consciously aware of how easily the brain can be manipulated involves a number of tough challenges, but it seems like it'll have to happen eventually.
- kurthr 8y agoYeah, just use the "performant" frustration tools invented by the telcos and great firewall... slow down the load/scroll, jerk the video, lose images, drop UI events, add typos, force reloads. Soon you'll hate that "social" interaction game. It's difficult but not impossible on HN :^D
- jon_richards 8y agoI predict that one day we'll all have our AI best friend. They'll teach and socialize us when we're young, guide and consol us as we grow, and protect us from misinformation and adversarial AIs throughout our lives. Picking the brand of AI for your kid will be the most important decision you ever make for them.
- kurthr 8y agoI suspect that your parents will choose for you, and the majority will be ad supported.
- b_tterc_p 8y agoI would wager less ad based, more political / religious / nationalist indoctrinated AI personalities.
- mrfredward 8y agoThe ultimate bubble. Your "friend" won't ever disagree, nor share any viewpoint other than what parents want the child to hear. The only interesting point will be to see what the "friend" chooses to do when the parents and the child disagree. AI "friends" could be terrific in teaching children, but there is a danger in consumers taking the short-sighted approach and buying the A.I. which is most likable. Imagine how spoiled kids will be when most of their "friends" always agree and serve them like butlers. People love to pay for indulgence, but never want to pay for tough-love.
- jon_richards 8y ago>there is a danger in consumers taking the short-sighted approach and buying the A.I. which is most likable. There are far more dangers than just that, but what if you've been brought up by an AI that taught you the self-control to avoid them? The AIs will fight over their ability to influence you. The hope is that the ones that use truth and reasoning will have an advantage, but there will always be the groups that are stuck in some fantasy.