As part of TechCrunch's ongoing “Women in AI” series, which aims to give AI-focused academics and other women the time they deserve (and are overdue) in the spotlight, TechCrunch interviewed Sofia Verastegui. Mr. Velastegui is a member of the National Science Foundation (NSF) National AI Advisory Board and former Chief AI Officer of Microsoft's Business Software Division.
Verastegui had no intention of pursuing a career in AI. She studied mechanical engineering as an undergraduate at Georgia Tech. But after joining Apple in 2009, she became fascinated with apps, especially those powered by AI.
“We started realizing that products with AI resonated with customers because of the sense of personalization,” Velastegui told TechCrunch. “The potential for developing AI to improve our lives on both small and large scales seemed endless, and I wanted to be a part of that revolution. So I decided to focus on AI. I started looking for projects that fit the bill and took every opportunity to expand from there.”
Careers driving AI
Verastegui worked on the original MacBook Air and the original iPad, and was soon named product manager for all Apple laptops and accessories. A few years later, Velastegui moved to Apple's Special Projects Group, where he helped develop CarPlay, iCloud, Apple Maps, and Apple's data pipelines and AI systems.
In 2015, Velastegui joined Google as head of silicon architecture and director of the company's Nest branded product line. After a short stint at audio technology company Doppler Labs, she accepted an offer as general manager of AI products and search at Microsoft.
At Microsoft, Velastegui ended up leading all business app-related AI initiatives, guiding teams to bring AI to products such as LinkedIn, Bing, PowerPoint, Outlook, and Azure. She also spearheaded internal research and projects built using GPT-3, OpenAI's text generation model, which Microsoft recently acquired an exclusive license for.
“My time at Microsoft was truly amazing,” Verastegui said. “I joined the company at a time when the company was in the midst of major changes under CEO Satya Nadella's leadership. Leaders and colleagues saw Microsoft as a laggard in the industry. But in a short period of time, Microsoft started to make real strides in AI, and I wanted to join.”
Verastegui left Microsoft in 2022 to start a consulting firm and lead product development at automotive technology company Aptiv. In 2023, she joined NSF's AI Committee, which brings together industry, academia, and government to support basic research in AI.
navigate the industry
When asked how she overcomes the challenges of a male-dominated technology industry, Verastegui cited the women she considers to be her strongest mentors. Verastegui says it's important for women to support each other. And perhaps more importantly, men stand up for their female colleagues.
“If you're a woman in tech, if you've ever participated in transformation, hiring, change management, don't be afraid to take a seat because you deserve a seat at the table,” Verastegui said. “Raise your hand to take on more AI responsibilities, whether it's part of your current job or a larger project. You'll have the best managers to support you and encourage you to keep moving forward. But if that's not possible in your 9-to-5 life, look for community or university programs where you can be part of an AI team.
Verastegui points out that a lack of diverse perspectives in the workplace (i.e., AI teams are mostly made up of men) can lead to groupthink. That's why we encourage women to share their feedback as often as possible.
“We strongly encourage more women to join AI so that our voices, experiences, and perspectives are represented at this important starting point where the foundational AI technologies of today and tomorrow are defined. ” she said. “It's important that women across all industries really lean into AI. If we join the conversation, we can shape the industry and change the power imbalance.”
Verastegui says her current work with NSF is focused on addressing fundamental unresolved problems in AI, such as the lack of what she calls “digital representation.” She asserts that today's AI is rife with bias and bias, and one of the reasons for this is the homogeneous mix of companies developing AI.
“AI is trained on data from developers, but most of the developers are men with a particular point of view, and they represent a tiny fraction of the world's 8 billion people,” she said. . “If we don’t include women as developers, and if women don’t provide feedback as users, then AI doesn’t represent women at all.”
Balancing innovation and safety
Verastegui sees the breakneck pace of the AI industry as a “big problem”: the lack of a common ethical and safety framework. She believes that if such frameworks become widely accepted, developers will be able to build systems quickly without stifling innovation.
But she's not counting on it.
“We have never seen such a revolutionary technology evolve at such a relentless pace,” Verastegui said. “People, regulations, legacy systems…never before have we had to keep up with the speed of AI today. The challenge is how to stay informed, stay informed, and think forward. However, it is important to be aware of the dangers of moving too quickly.”
How can companies (or developers) responsibly create AI products today? Verastegui says they can learn from past mistakes and adopt a “human-centered” approach that puts the user's well-being front and center. I support.
“Companies should empower diverse, cross-functional AI councils that consider issues and provide recommendations that reflect the current environment, and for regular feedback and oversight that adapts to the evolution of AI systems. You need to create a channel for it,” said Velastegui. And there needs to be channels for regular feedback and monitoring that adapt as the AI system evolves. ”