Can We No Longer Believe Anything We See?

Like many, i've been thinking a lot about this. I feel like we are in the "hdr photo" phase of AI image creation. If you recall when HDR photography as a technique first gained popularity it was heavy handed and obvious, then it became more subtle and harder to identify. Now it's just par for the course. Any well exposed image might be HDR, or might not.

For AI media, given the current pace of improvement, very soon enough it will be truly impossible to tell "real" media from generated media. Images and voice clones and deepfake video will become impossible to determine authentic from fake. If we take it that the quality of generated images "is as bad as it will ever be", then the people born today will never know a world where and recorded image, sound, or video will ever be "true" or trustworthy. When that is true, what does it mean for news, relationships, trust, etc? Sure, we have had photoshop and CGI video, but now virtually anyone can make media of anyone alive today doing or saying anything within minutes, and in a way that will be indistinguishable from an "actual recording".

So, will the people born today just become unbelievably cynical? Will they ever trust...anything they see? What will their relationships be like? How will they get news? How will they elect leaders?

This has really been weighing on me, and I don't know what the future will hold.

It may be for the better.

For all great things that the Age of Internet brought to the humanity, detrimental effects of social media (and other non-traditional information sources) on our democracies is evident.

Counter-intuitively, the democratisation of the mechanisms for dissemination of information ultimately led to weakening of institutions that consist the pillars of our democratic societies. My take is that this was largely through the dilution of authority brought about by the cambrian explosion of sources of “information” (more accurately: sources of content).

Symptoms of this include various manifestations of populism, which are typically fueled or at least helped by a mix of establishment-challenging conspiracy theories (including the harmful challenges to well-established scientific conclusions based on objective and impartial process of reasoning).

If the generative AI will drive us to assuming everything we see online is fake, this would take us back to a state where track record in informational integrity would grow authority of individual information sources, and reverse the effects of authority dilution.

If that happens, it will be a full circle, driven by successive technological advancements.