Member-only story
Deepfakes Don’t Matter
Why the presently fashionable alarm about AI-generated images is completely pointless
As Artificial Intelligence (AI) continues to improve, many commentators are becoming more and more worried about its impact on Western society. Now that algorithms can generate text output at speeds and in volumes that can saturate social media with any content the creator desires, and now that algorithms can generate lifelike pictures that can present anyone anywhere doing anything, and now that algorithms can generate videos that can accomplish the very same thing in animated form with realistic voices, how can ordinary people know what’s real and what is a lie?
This problem, we are told, can undermine the basis of democratic society. What if ordinary people see images or videoclips that are misleading? They’d believe lies to be true! Therefore, the agitated commentators explain, we urgently need technological fixes to ensure ordinary people can distinguish between deepfakes and reality.
To which I say: nonsense.
I am not denying the ever-increasing realism of deepfakes. They are very accomplished artifacts today and no doubt within a very short period of time will indeed be utterly indistinguishable from reality. I have no doubt that unscrupulous individuals and organizations will use…
