Anyone worth their salt here already knows that predictive analytics obviously works on humans, since the the human brain has zero days that can obviously be discovered with enough behavior data (e.g. by having multi-armed bandit algorithms run A/B tests, you don’t even need AI).
The Cambridge Analytica people knew that too, and took advantage of it.
Nobody should be persuaded in either direction by the clusterfuck that emerged as a result, nor by any of the people who ended up persuaded by it. This should be ABSOLUTELY BASIC operational security for anything related to researching persuasion technology.
I don’t think this invalidates the point that microtargeting can be very effective.
Anyone worth their salt here already knows that predictive analytics obviously works on humans, since the the human brain has zero days that can obviously be discovered with enough behavior data (e.g. by having multi-armed bandit algorithms run A/B tests, you don’t even need AI).
The Cambridge Analytica people knew that too, and took advantage of it.
Nobody should be persuaded in either direction by the clusterfuck that emerged as a result, nor by any of the people who ended up persuaded by it. This should be ABSOLUTELY BASIC operational security for anything related to researching persuasion technology.