At Exceptional Women Alliance (EWA), we encourage women to mentor one another. As the founder, president, and CEO of the nonprofit, I have the privilege of interviewing and sharing insights from some of the leaders who are part of our peer mentoring program.
Meet Sunny Webb, a seasoned product leader who combines technical innovation with business intelligence. Sunny has led and launched more than 50 AI-enabled products, as well as developed other complex technology products for some of the world’s largest and most respected companies, including Apple, Accenture, Disney, Philips Healthcare, and Microsoft, among others.
She is also a member of the OpenAI Forum, an initiative that brings together experts, academics and students to collaborate and debate the present and future of artificial intelligence.
Fast Company – How did you become interested in artificial intelligence?
Sunny Webb – I’ve worked with AI and machine learning my entire career. I’m really excited about the broader applications of AI that eliminate tedious tasks from life and work.
I started working with a stealth (confidential) company that is applying machine learning to compress cloud infrastructure observation data.
Last year, I researched how generative AI can be designed and built to be trustworthy. In my opinion, building trust-centric AI systems is probably the most important topic today in terms of technology and impact.
Fast Company – How can we create trustworthy generative AIs?
Sunny Webb – Generative AI is a powerful tool that accelerates creative use cases. Typically, these tools create multimodal data types, such as text, images, and videos. At a time when deep fakes are taking down social media channels, being able to trust that an image or video is real is important to the public, especially in an election year.
I recently led a research project focused on understanding user concerns about the trustworthiness of generative AI. The top solution users suggested to increase their trust in the results was to notify them when they shared sensitive information and to allow them to choose what information the generative AI could store.
This aligns with something I heard recently at an event attended by OpenAI CTO Mira Murati, when she was asked whether a watermark on an image generated by generative AI would be enough to guarantee the trustworthiness of that content.
Murati explained that OpenAI was considering a method for humans to identify components of hybrid images, that is, images that are part real and part AI-generated.
Imagine a generative AI-generated photo of Dolly Parton on Mars, with the Mars image identified as “synthetic” and the Dolly image as authentic. This would inspire trust because it would establish a governance for users to trust a video or image.
I think this kind of transparent information is one approach to gaining users' trust. Other alternatives would be to develop interface designs that prevent users from sharing sensitive information and allow them to select what data is stored.
Fast Company – Can we trust generative AI today?
Sunny Webb – We’re getting there. I encourage my friends and family to read the privacy policies of any generative AI and hallucination data tools they use. This information is easy to find on the technology vendor’s website.
A good example of a company that has successfully built trust in its products is Apple. This fall, they will launch Apple Intelligence, which will incorporate generative AI into iPhone, iPad, and Mac to help with writing. The announcement was notable because the processing is done on the device, rather than in the cloud.
Artificial intelligence will provide ways
for us to be more human, in ways we
cannot imagine today.
When I worked with Apple while at Accenture, I got a behind-the-scenes look at how they operate. Every technical and business decision at Apple starts with protecting user privacy – without compromise or second thoughts.
For example, Siri was originally designed to not have access to previous Apple Maps searches. While this makes Siri a bit more limited, users know that when they pick up an Apple product, their data and behaviors won’t be shared with another app for profit.
Fast Company – What are some of the challenges facing generative AI?
Sunny Webb – Beyond trust? Speeding up multimodal processing, while filtering out hallucinations or errors. And a bad reputation for replacing humans.
Fast Company – Why Are We Talking About AI Replacing Humans?
Sunny Webb – Tech leaders realize there are two ways to win in this new frontier: speed of development and power of data collection. Sometimes power is gained by spreading fear, and some top CEOs have played this power game by spreading myths about how devastating AI could be for humanity.
It reminds me of the arrival of the internet. No one could have predicted how powerful it would be, but everyone understood that it was a huge step forward for sharing information.
Similarly, before the dot-com bubble burst, there was talk about how chat rooms would destroy the fabric of society. It’s the same with generative AI: it won’t replace people, but it will provide ways for us to be more human in ways we can’t imagine in our current state.
Comments