How do we detect deepfakes without creating them?

[plowsai]

Recently our team realized that the rate of progress on deepfake generation and the rate of progress on deepfake detection are dramatically misaligned. The only way to align these two would be to create the highest quality deepfake generation in both audio and video and continue to add layers of safety into them that industry doesn't have right now.

Additionally since we are a capital intensive business, it's of the belief that we can monetize the creation of high quality clones and that would drive more business towards our detection products.

We would love to hear your thoughts