Women in AI: Sophia Velastegui expresses concern that AI development is progressing too quickly.

TechCrunch has been interviewing Sophia Velastegui as part of its series on women in AI, with the goal of giving those women academics and others much-needed-and well-deserved-time in the spotlight.
Women in AI: Sophia Velastegui expresses concern that AI development is progressing too quickly.

TechCrunch has been interviewing Sophia Velastegui as part of its series on women in AI, with the goal of giving those women academics and others much-needed-and well-deserved-time in the spotlight. Velastegui serves on the National Science Foundation's national AI advisory committee and is former chief AI officer for Microsoft's business software division.

Velastegui never set out to work in AI. She graduated in mechanical engineering from Georgia Tech as an undergrad. After a stint at Apple in 2009, though, she became enthralled with apps, particularly those that used AI.

"I began to realize that AI-infused products were resonating with customers because of the feel of personalization," Velastegui told TechCrunch. "The possibilities seemed endless for developing AI that could make our lives better at small and large scale, and I wanted to be a part of that revolution. So I started seeking out AI-focused projects and took every opportunity to expand from there."

AI-first career
She was part of the team that created the first MacBook Air and the first iPad, and a few years later was promoted to product manager for all of Apple's laptops and accessories. A few years later, Velastegui moved into Apple's special projects group, where she helped develop CarPlay, iCloud, Apple Maps, and Apple's data pipeline and AI systems.

She went to Google in 2015 as head of silicon architecture and director of a line of products branded after Nest. She then followed another brief stint at the audio tech company Doppler Labs with a job in 2017 at Microsoft, where she is an AI product and search general manager.

At Microsoft, where Velastegui eventually oversaw all business app-related AI initiatives, Velastegui coached teams to add AI to products such as LinkedIn, Bing, PowerPoint, Outlook, and Azure. She also led internal explorations and projects built with GPT-3, OpenAI's text-generating model to which Microsoft had recently acquired the exclusive license.

My time at Microsoft truly stands out," Velastegui said. "I came in when the company was in the midst of huge changes under CEO Satya Nadella's leadership. Mentors and peers counselled me against making that jump in 2017 because they viewed Microsoft as lagging in the industry. But in a very short window, Microsoft began making real headway in AI and I wanted in.".

In 2022 she left Microsoft to start a consulting firm and head up product development for Aptiv, the company revolutionizing automotive technology. In 2023, she joined NSF's AI committee. "The committee works with the industry, academia, and government to advance basic research in AI."

Interaction with Industry
How does she navigate these challenges, in a still largely male-dominated tech industry? Velastegui credited the women she considers her strongest mentors. It is important, Velastegui says, that women support one another — but perhaps even more important is that men stand up for their female co-workers.

If you have ever been involved in a transformation, an adoption, or a change management process for women in tech, then you should be at the table-and don't be afraid to take your seat there. Raise your hand to own more AI responsibilities, in your current job or as part of a stretch project: the best managers will make sure you keep pushing ahead forward. But if that's not possible in your 9-5, seek out communities or university programs where you can be part of the AI team.

A lack of diverse viewpoints in the workplace—AI teams made up mostly of men, for example—can lead to groupthink, Velastegui notes, which is why she advocates that women share feedback as often as they can.

"I would hope more women get involved in AI so our voices, experiences, and points of view are included at this critical inception point where foundational AI technologies are being defined for now and the future," she said. "It's critical that women in every industry really lean into AI. When we join the conversation, we can help shape the industry and change that power imbalance."

Velastegui says that her work in the NSF now centers on those outstanding fundamental issues in AI, such as a lack of what she calls "digital representation." Biases and prejudices pervade today's AI, she avers, in part due to the homogenous makeup of the companies developing it.

"AI is training on developer data, yet the developer population is mostly composed of men with particular opinions and a very small piece of the 8 billion global human population," said Balsamo. "If we're not contributing women to the developer and not women as users as a means of feedback, AI is never going to represent them either."
Achieving innovation within safety
For Velastegui, the breakneck pace of the AI industry is a "huge issue" — that is, absent a common ethical safety framework. She believes such a framework could allow developers to build systems with speed without stifling innovation, were it ever to be widely embraced.

She's not counting on it, however.

"We've never seen technology this transformative evolve at such a relentless pace," said Velastegui. "People, regulation, legacy systems … nothing has ever had to keep up at the current speed of AI. The challenge becomes how to stay informed, up-to-date, and forward-thinking, while also aware of the dangers if we move too fast."

How does a firm — or a developer, for that matter — do it right now? The author espouses what he labels as "human-centered," meaning drawing on mistakes of the past, starting with making users' welfare central to any such strategy.

"Companies should empower a diverse, cross-functional AI council that reviews issues and provides recommendations that reflect the current environment," Velastegui said, "and create channels for regular feedback and oversight that will adapt as the AI system evolves. And there should be channels for regular feedback and oversight that will adapt as AI systems evolves."

Blog
|
2024-11-04 18:43:14