Deepfake | Doctored | Technology | Buffalo | B Corp
Earlier this year the Florida Dali museum unveiled a lifelike simulation of Spanish painter Salvador Dali. See here (https://mashable.com/article/salvador-dali-deepfake/). The technology itself is enough to bring tourists out. When tourists enter the museum they’re greeted by Dali himself. That’s one of the big selling points for this technology. It isn’t an artist rendition of Dali and a voice actor speaking lines. Using an archive of images and soundbytes the museum was able to re-create the real Dali, introducing attendees to his museum and ending with a selfie he summons from his pant pocket.
Pretty impressive right?
Well, while deepfake technology has offered entertaining headlines, including a fabricated Joe Rogan podcast where he talks about investing in a Chimpanzee sports league[1] it has raised concerns about distinction between reality and fiction. Only three years removed from Russian interference in the US election, individuals are concerned what impact this kind of technology can have on our daily digest of digital information.
Worries about artificial intelligence media content has grown since a doctored video of Nancy Pelosi garnered 2.5 million views on Facebook, making her appear intoxicated and slow.[2] This attempt was poorly executed and most viewers were able to spot the fake.
Ben Sasse introduced a bill earlier this year that would criminalize the creation of deepfake videos.[3] Is the stress warranted? Yes and no. While this technology can be misused it’s also been accessible longer than the general public is aware. “Part of the issue is that they’re too easy to track. The existing deepfake architectures leave predictable artifacts on doctored video, which are easy for a machine learning algorithm to detect.”[4] Though it only recently started making headlines, defensive technologies have been researching detection software for years.[5]
While its capabilities can still cause trouble and confusion, these were issues before deepfake technology entered the political sphere. During the last presidential election, a clearly doctored video of Hilary Clinton shaking Osama Bin Laden’s hand made rounds on social media like Facebook and those who sought reason to dislike her had no issue believing, or at the very least sharing, the picture. The confusion deepfake videos could inspire is nothing new, it’s just repackaged.
https://www.theverge.com/2019/5/17/18629024/joe-rogan-ai-fake-voice-clone-deepfake-dessa
https://www.cbsnews.com/news/doctored-nancy-pelosi-video-highlights-threat-of-deepfake-tech-2019-05-25/
https://www.axios.com/deepfake-laws-fb5de200-1bfe-4aaf-9c93-19c0ba16d744.html
https://www.theverge.com/2019/3/5/18251736/deepfake-propaganda-misinformation-troll-video-hoax
https://gcn.com/articles/2019/05/24/deepfake-detection.aspx