Created: 2023-05-09 18:28

HUMANNESS IN THE AGE OF AI: A path to an open and permissionless identity protocol

The idea that a perfect system of identification might solve the problem of of fake news and miss information spreading is both wrong and naive.

Yes, of course it makes it harder to do so, since actual humans would be needed in certain situations (eg. when a platform expects the account to be ā€œunique human verifiedā€). However click farms are already widely used today and would always accessible to the powerful or wealthy enough to pay for them.

Iā€™m not sure whatā€™s worse, a world where virtually anybody has access to the spreading of misinformation or a world where only the rich enough do.

The fact that people would have the option to filter out human generated content doesnā€™t mean people would be interested in doing so. People might trust more AI generated content that confirms their biases/preconceptions than human verified content that contradicts them.

And finally, human verification might be useful for purposes like the example of resource allocation in India, but when it comes to content online thereā€™s no way to verify that a human verified account isnā€™t sharing AI generated content. The action of publishing might be carrier out by a human, but the source of content cannot be enforced.