If we think now that we know about fake news disinformation efforts around the US elections that we are somehow inoculated against it, we’re probably wrong. The technology that’s ahead of us might make fakes in ten years from now appear even more real than the real news.
First, there’s speech synthesis that’s going to sound indistinguishable from a news casters. Then, there’s the ability to morph recordings or to take a voice and create a full speech synthesis engine out of it. The sped up Acosta video from a few months back will look like child’s play compared to what’s ahead.
After that? Natural language generation of articles that might not involve human writers. These might be so tailored and convincing to us that they’ll be difficult to thwart. They might involve things like GANs developing and testing the best combination of words and imagery to make a piece of fake news spreadable.
Then, there’s the creation of fake faces that will seem lifelike. This could be a boon for fake twitter or Instagram accounts.
We might end up needing a certification authority, something akin to the checkmark for verified twitter accounts. Maybe this is where Blockchain technology might create a foothold… not only a way of tracking where and when a post was made and by who, but ensuring it wasn’t manipulated along the way and if so, who could be the culprit.
Maybe we’ll start seeing the first seeds of this in 2019?