Deepfakes Are Easy to Detect – Business POV Is Where Problems Start | Podcast 40

Deepfakes Are Easy to Detect – Business POV Is Where Problems Start | Podcast 40
Interview by Nikola Djuric
Published: March 29, 2024

Though many people don't realize it, 77% of us interact with AI daily, from organizing our schedules to keeping us healthy and entertained.

In business, AI is a big deal, too.

This technology is expected to change U.S. employment figures dramatically by 2025, creating 12 million more jobs and boosting the global economy.

But AI isn't all good news.

It brings challenges, like fake news and privacy concerns.

Dr. Jeff Lu, a leader in AI and the founder of the AI face swap tool Akool, talks to DesignRush about the good and bad sides of AI.

Tune in to the show to learn:

  • What are marketing applications for cloning influencer images digitally
  • How is AI used to uncover hidden data patterns and enhance customer experience
  • Which three Ts are crucial in creating AI-led user experiences

Who Is Jeff Lu?

Jeff is one of the global leaders in AI and computer vision. With over ten years of experience, Jeff has managed several teams and startups to deliver outstanding AI products to the market. His book "Enhanced Human" has won Jeff the award of Top 10 Pioneering CEOs and Top 50 CEOs of AI companies. Since 2022, he's building Akool to become a leading company for personalized visual marketing with AI-powered videos, images, and avatars.

There are a bunch of ethical considerations that companies must address when developing AI-powered customer experiences.

This is especially true in the case of face swapping due to the potential of users finding malicious ways to use deepfakes for profit or just harming other people.

However, Jeff argues face swapping isn't anything new, but just an improved way of using technology due to the rapid development of generative AI.

Akool's Face Swap Tool Example
Akool's Face Swap Tool Example | Source: Akool

"There is no good or bad from the technological point of view. There's a huge amount of ongoing research around the topic of deepfakes as people care about their faces a lot all around the world," he explains.

"It's a security threat that all face recognition systems on the market consider – think of the banking security system, for example."

We Need an Industry-Wide Consensus

OpenAI, as an industry leader, is accountable for introducing additional security rules around how people use AI-powered apps, including the upcoming text-to-video software Sora.

Some ways to increase the security of AI-generated content are to:

  1. Include watermarks
  2. Enforce stricter copyright protections

Jeff tells DesignRush that the latter will be very hard to achieve, as companies can always look for places around the globe where copyrights aren't a topic of focus in government policies.

"We have an in-depth plan to tighten regulations surrounding AI. It just depends on reaching an industry consensus so all tech companies can implement the rules simultaneously."

The third option is already in motion and includes introducing consent to the use of face recognition software.

"It's not just a face swap, it could also be voice clone or digital clone," Jeff adds.

"From the technical point of view, we can solve all problems easily. It's just from the business point of view when we're going to launch new security features."

Conquering Space & Time and Elon Musk's Legacy

Asked about his role model, Jeff tells us that he loves Elon Musk's work.

"Elon Musk wants to conquer space by going to Mars, traveling to other satellites, and expanding on the use of electric cars."

"There are a lot of people who are working on conquering the space. Our work, on the other hand, is all about conquering the time. Space and time are the internal topics for the human race."

"Digital clones our industry is developing help people preserve themselves in digital formats and, in the future, preserve their intelligence.

"This is how conquering the time looks in real life," Jeff concludes.

Elon Musk was a significant contributor to OpenAI, envisioning it as a counterbalance to the likes of other tech giants. Musk and OpenAI explored various avenues, including a possible merger with Tesla or transitioning to a for-profit entity.

However, differences in vision for governance led Musk to part ways and pursue AGI development within Tesla. The relationship between OpenAI and Musk eventually became litigious, with OpenAI seeking dismissal of Musk's claims and the case being pushed for dedicated case management.

Listen to the full podcast episode on Spotify and other streaming platforms:

Subscribe to Spotlight Newsletter
Subscribe to our newsletter to get the latest industry news