"Deepfakes" are a category of videos that use AI to create or recreate a video using someone's face and voice in a place or scenario that looks and sounds real. Think about the old Photoshop tricks on steroids demonstrated in the Einstein image above.
Many of these are harmless fun, but they are also increasingly become a way to mislead viewers into believing something that isn't true. On the fun side, if you search on YouTube for "deepfake", you are sure to see quite a few videos with Nicholas Cage in movie scenes where he doesn't belong. On the dark side, these can be used to create and share misleading stories in political campaigns or celebrity interviews.
At one point or another most of us have drawn a conclusion about something based on limited or even misleading information. I know I have been guilty making a judgement about a video or story online only to find out that I didn't have all the information.
Our students have or will be mislead by one of the deepfake videos, which is why I wanted to share a great new resource created by MIT and and a couple of AI companies. I think this site is a great way to engage students in thinking critically about digital media and could be incorporated into digital citizenship lessons and programs.
The site brings to light a planned speech by President Nixon in the event the something went wrong with the first moon landing.
Clicking on continue brings up the question, "Can you spot a deepfake?" and a video about the moon landing starts. The challenge is to figure out what is real and what is fake. After the video the viewer has a chance to guess what was real and what was a deepfake.
The site also includes some great resources to dive deeper into how they created the video and more about the future of deepfake videos. There really are some incredible resources on the site.
What do you think? Should this be something we all learn from?
No comments:
Post a Comment