Wonderwoman has recently been featured in a pornographic movie. It turns out that the producer could launch the movie without paying neither rights nor actually getting the famous star Gal Gadot (Wonderwoman) on the porno movie set.
Yet she is the main actress in this movie! Yet the movie photograms are very realistic.
Each Sunday morning when I wake up I get about 100 emails and weekly updates on what is happening in Artificial Intelligence and Machine Learning. I try to read as many as possible but this was really and easy one to spot!
Here is what I was thinking while reading the news:
Imagine you are a company that owns the image rights of a particular actor/actress, how are you going to go about this? How do you make sure you can spot something like this and stop it or ask for $$?
The technology used in the Deep Porn is known as “deep fake” and it is actually very interesting technically. But I want to focus on the business in this email.
As you can see, given a new industry a new set of needs arise, new problems need to be solved. Who would think of “Deep Porn” few years ago? Who would imagine of thee need for an “Anti Deep Porn” Sentinel service?
Although you may not see it immediately the connection, this problem is similar to the broader problem of understanding what machine learning is doing (the black box problem), how it is reaching a conclusion and whether or not there is a bias.
In essence we need the ability to reverse engineer the machine in one case to understand why it reaches a conclusion in the other case ( the pornographyc movie) to understand that the video is a fake.
The need to spot a bad use of images and audio will be prevalent in the future. You will be likely to face a situation where you either have to
- Put in place a solution
- Buy some solution from a vendor
- Combine an existing commercial solution with your home made process.
Providing a sounded business solution requires much more than just a data scientist. It requires an understanding of business objectives, legal landscape and technological trade offs necessary.
This is exactly what we explains in one of our course real use cases. Where we analyze a hybrid home made-partially purchased solution for monitoring E-commerce content and the presence of offensive content.
You are only one click away from our course, that features real life use cases that will set you apart from the crowd and help you develop a unique blend of business and technical mindset
Act now and be ahead of the curve, start thinking in thee way the A.I. economy will work and start rising to the new opportunities it will create. .