YouTube Execs Weigh In On AI Safeguards, Living Room App Tweaks & How Pat McAfee’s ESPN Move Could Be Emmys Precursor

deadline.com:

After YouTube announced new creator tools and product updates at its Made on YouTube event Wednesday, CEO Neal Mohan emphasized that security, especially surrounding updated AI tools, remains “top of mind” for the tech giant. “YouTube occupies a truly unique space,” he explained during an onstage Q&A session at the end of the event. “We get to work closely with cutting-edge technologies that are invented at YouTube, invented at sister organizations like Google DeepMind.” The company has settled into a place where it is “very gradual about the rollout of these technologies instead of just putting it out there and seeing what happens.” As far as potential misuse of AI, Mohan continued, “This is really what the name says.

It’s a tool. It’s about streamlining” the creative process. In a separate interview with reporters after Mohan’s Q&A, senior execs were asked how the AI tools would respond if a creator were to type in “Kamala Harris,” for example.

With just weeks remaining before the election, concern remains elevated about the potential for deepfakes or misleading images and videos skewing the vote. (Politics, not surprisingly, was not mentioned during the hour-and-a-half presentation.) Chief Business Officer Mary Ellen Coe affirmed that “all things created by creators with the tools are subject to our community guidelines.

So, they will go through our same trust and safety systems.” She added that if material is generated synthetically, “it is automatically labeled as such.” Amjad Hanif, VP of Product Management, Fan Funding at YouTube said plugging “Kamala Harris” into the tool would prompt it to ask, “Does this output actually match something that already exists?” He explained, “We’ve built the technology to be