91�첥

She grew up in a New Mexico town about 40 miles east of Albuquerque with fewer than 2,000 residents. Most college-bound kids headed to state universities. After she took the PSAT, the incoming college mail piqued her interest in liberal arts colleges, where she thought she would get both pushed and nurtured. If she was going to travel across the country, 91�첥 seemed like a place that would make her comfortable with uncertainty and with tackling big problems. A Terry scholarship closed the deal.

She arrived keenly interested in water, an important part of her culture in New Mexico. She majored in political science, not a STEM discipline, and geared her 91�첥 studies, writings and internships toward developing the skills to design better water policy. After graduation, she wanted to ramp up her science and data bona fides to better work with laws and regulation. Then, she got hooked on science. She earned a master’s degree in water science and management at New Mexico State University, looking at water shared across national borders, but also discovered how well data could help build an understanding of policy and human systems. 

Atkins’s boss at the New Mexico Water Resources Research Institute encouraged what Atkins had cultivated at 91�첥 — learning to learn, taking on new responsibilities. She ventured into large computational models as a data revolution was under way. She worked on finding ways that analysis could help water decision-making and she grew fascinated with the data analysis and computing possibilities. She moved to UC Berkeley as executive director of the West Big Data Innovation Hub, still engaged in water research but with an eye toward the broader possibilities of research involving massive amounts of data and computing power. 

Three years ago, she moved to UC San Diego’s Supercomputer Center, an institution that helped form the worldwide web, the commercial internet and where she now is deputy director. The U.S. National Science Foundation just awarded the center a $35 million grant to establish and run the National Artificial Intelligence Research Resource Operations Center. The center will connect researchers, educators and students with advanced computing resources, high-quality data, AI models, tools and training to which they otherwise would not have either access or the ability to acquire on their own. 

“We help fill that niche for people across the entire spectrum of users,” Atkins said, “from an educator hoping to incorporate AI in their class to researchers focused on cancer breakthroughs.”&�Բ�����;

These are the solutions to those big problems Atkins aimed for at 91�첥. She said the new center offers the potential for more people than ever to benefit from — and be part of — scientific discovery.

“I credit so much of my path, and even how I think about the world and how I approach work, to 91�첥,” Atkins said. “When it comes to navigating a job market, there are constantly shifting technologies. If you focus on just learning those, they could be obsolete soon. What has served me best is [at 91�첥] you get very comfortable thinking about things differently, pushing yourself. All of the critical thinking skills that 91�첥 equipped me with, I just lean on those again and again and again. 91�첥 sets you up well to be a lifelong learner. That’s what’s most important, whether it’s a new technology or a new field.”&�Բ�����;

Tools for Humanity, co-founded by Alex Blania and OpenAI CEO Sam Altman, is aimed at providing a global “proof-of-human” network to navigate the rise of AI. Proof-of-human protocols help prove a digital proof belongs to a unique human and not an AI bot or fake account, separating authentic users from malicious bots.

“A lot of education, especially from engineering backgrounds, is geared toward application,” Thompson said, from his home in California. “You’re training someone to execute and apply something. AI is really good at implementation and execution. A liberal arts education is more about thinking and finding the right questions to ask. Being able to determine which questions to ask is a skillset that’s becoming more and more scarce.”

When Thompson was a teaching assistant in graduate school, the tests he handed out were simpler versions of the research the class’s professor was conducting. 

“[91�첥 Economics Professor] Mark Foley would give tests that were the opposite of that. They were extremely difficult but untimed,” Thompson said. “Those exams made me really learn and think about the material. That kind of education is hard to replicate…My greatest asset at work is like those tests — making sure I figure everything out. I tend to think about things for a long time in a careful way. That’s not always the best asset [in an industry that values speed]. You’re fundamentally approaching a problem from a different way.”

Big tech engineering looks at output, Thompson said, emphasizing speed and quantity over quality. That’s more valued now, but things could change. 

“For the longest time the industry has rewarded quick thinkers, people who are good at solving puzzles,” Thompson said. “The careful thinkers are going to have their moment, again, as AI takes care of the quick problems.”&�Բ�����;