One Big Thing | Will Deep Fakes or Cheap Fakes Determine The Election?

Deep and cheap fakes are making one thing clear. In arguably the most important U.S. presidential race of our time, they might determine it.

We already knew fake news was going to be a problem in November but we didn’t foresee (at least I didn’t) the hypergrowth of deep and cheap fakes made possible by advances in generative AI.

First, an oversimplification of a deepfake versus a cheap fake.

Cheap fake example: Think “photoshopping.” Taking an existing video and replacing one word with another that would change the meaning of what Harris or Trump actually said. These are called “cheap” as they don’t require much programming knowledge or use sophisticated tools.

Deep fake example: Create a fake clip from scratch via the use of sophisticated generative AI tools.

Technology has always helped determine presidential winners:

• FDR’s mastery of radio
• JFK’s mastery of TV
• Obama’s mastery of social media
• Trump’s mastery of microtargeting at scale (Cambridge Analytica)

The situation is compounded by the fact we now have two candidates who wrestle with the truth. Essentially the perfect storm for cheap and deep fakes to determine the most powerful office in the world.

I only wish this whole affair was fake.