Should We Train AI for Manipulation? Balancing Ethics and Innovation

Manipulation is never a good thing and introducing such capabilities creates a slippery slope to loss of control and free will. Influence, on the other hand, can be, especially if it is done without an underlying agenda which benefits someone other than the person being influenced.

Kiron

Unfortunately, the manipulation train has already left the station and has a 90% on-time arrival rate at the doorstep of your multimedia device.

Any ethics scale should/would collapse under the mass of “manipulation for a greater good.” Which is why those who have convinced themselves that they are the “arbiters of truth and societal good” implement internal policies preventing the use of “certified ethics scales” during regular business hours.

Although I stated the above with a bit of “tongue-in-cheek,” transparency in its proper form has no companions, it’s all-or-nothing. I know it sounds skewed to far in one direction, but unless it stands alone there are lifecycle consequences.

Are those consequences worth the cost? Ethics can/should guide a response to that question, but only time speaks the truth, and (back to tongue-in-cheek) by the way, “retrospectives” are also banned on the aforementioned internal policy -:)

George

Edward L. Bernays wrote in his seminal work ‘Propaganda’ (Bernays, 1928), p.9.: ‘The conscious and intelligent manipulation of the organized habits and opinions of the masses is an important element in democratic society.’

Bernays later was involved in building the CIA capability to interfere in other nations and heavily involved in building marketing practices.
He was born Austrian, a nephew of Sigmund Freud. His Jewish family migrated to the US in the 1890s.

So, yes AI will be/is already used for manipulation, agree with George on that.