The Head of AI Strategy & Research in the Office of the CTO at Bloomberg reflects on the state of artificial intelligence.
Whenever Amanda Stent ’01 (PhD) visits one of Bloomberg’s 159 locations worldwide for the first time, they like to play a little game.
“I like to sit on the main floor of the office and watch people go by and try to guess their role at the company,” says Stent. “We have three product organizations at Bloomberg and then we have a big media organization. The product and media people are distinguished by the nature of their suit jackets.”
As someone whose work cuts across an organization of more than 25,000 employees, Stent knows the company and its people well and seemingly has a hand in every aspect of the business.
As the Head of Artificial Intelligence (AI) Strategy & Research in the Office of the Chief Technology Officer (CTO) at Bloomberg, Stent works across the company to help it break down barriers in key strategic areas related to AI. Stent’s team works with Bloomberg’s engineers, product managers, data scientists, data analysts, legal team, and more, as well as with outside partners, including clients, regulators, and academics.

If AI can help improve a product, ease a business challenge, or empower decision makers, Stent and their team are there to help. Stent keeps the company, which provides financial data, news, and analysis, on the cutting edge of AI research and enables Bloomberg to apply the latest techniques wherever they can make an impact.
With more than 100 research papers and two books published, Stent’s influence goes well beyond the company’s reaches. As one of the world’s leading authorities on natural language processing (NLP), Stent has helped drive the advancement of AI in both academia and industry for decades.
While AI technology has grown by leaps and bounds since Stent was a computer science PhD student at the University of Rochester, many of the problems that existed then still persist today. Although natural language generation is much more fluent, as demonstrated by products like ChatGPT, Stent says AI can still struggle with producing outputs that are factually correct, as well as with reasoning and planning—two longstanding areas of research at the Department of Computer Science. Stent is excited by the prospect of chipping away at those problems.
“Over the next couple of years, I think we’re going to see a lot of use of these new AI techniques, not just for language, but also for images, music, sequences of genomes, any kind of data, really,” says Stent. “But we’re also going to see an increasing focus on how to make these things work in embedded contexts, where the system is in the real world, where the language is grounded in the real world, and the interaction achieves an actual task.”
Rochester roots
Stent has been fascinated by AI since learning a programming language called Prolog in high school, but their parents pushed them to study mathematics and music as an undergraduate at Houghton University. Still, Stent continued to dream of working in AI and got closer to that dream while spending a year abroad at the University of Edinburgh, which had the world’s largest AI library.
When exploring graduate schools and trying to decide between programs in AI and cybersecurity, Stent visited Rochester’s Department of Computer Science. The department provided a demo of TRAINS, one of the earliest fully functional spoken dialogue systems in the world, and Stent was hooked.
“The department took a big bet on someone from a little-known liberal arts college with majors in math and music, who was a ball of anxiety, determined to pursue this field, but pretty sure I would fail,” says Stent. “They enabled me to achieve my intellectual dream and have a satisfying career full of interesting problems, wonderful colleagues, and the chance to think broadly and deeply about computer science.”
While at Rochester, Stent would study with researchers who produced seminal works in the field of AI, including professors James Allen and Chris Brown. In true Rochester fashion, Stent was also exposed to a wide array of interdisciplinary collaborators and learned from experts in psychology and cognitive sciences. Through invited talks hosted by the department, they met future mentors like Susan Brennan, a cognitive scientist and distinguished professor at Stony Brook University, experiences they called “invaluable.”
“The small things that happen in a graduate program—meeting a person at a conference or the conversation you have with someone who comes to give a talk—can be career-forming events, even though they seem minor at the time,” says Stent.
A leader in industry and academia
Stent went on to a decorated career in both academia and industry. They spent several years as a faculty member in the computer science department at Stony Brook, then moved to industry to work as a principal member of technical staff at AT&T Labs–Research, as director of research and principal research scientist at Yahoo, and as the NLP architect in the Office of the CTO at Bloomberg.
In 2021, Stent returned to academia as the inaugural director of the Davis Institute for Artificial Intelligence at Colby College, the country’s first institute for AI at a primarily undergraduate institution. Earlier this year, Stent began their second stint with Bloomberg, this time as the company’s Head of AI Strategy & Research.
While Stent is often asked about jumping back and forth between industry and academia, they do not pay much attention such distinctions.

“I encourage people not to think about whether a job is in academia or industry, a government lab or at a startup,” says Stent. “As long as you keep publishing and remain active, you will have options. I encourage people to think broadly, especially for the first job after completing their PhD. That way, they’re not deciding about academia versus industry and instead they’re deciding about the job, the organization, and the people they will work with.”
Stent says that part of what propelled their career is avoiding the scrum and following their own path. Stent tells students who are deciding on a direction for their PhD research not to frustrate themselves by going after the hottest topic and trying to solve it all at once. Instead, Stent urges them to take small but meaningful bites off important problems. And Stent urges students to think wisely about their destination after completing their PhD.
“The organization that is the biggest and the fanciest today is probably not the one that will be the biggest and fanciest tomorrow,” says Stent. “One can have incredibly successful careers at many organizations, and you don’t have to make the decision based on what everyone else thinks is popular. It should be about what interests you—not some status seeking thing—because you might be the one who makes a critical difference at a small organization.”
Witnessing the birth of a new field
Stent sees AI emerging as a distinct discipline from computer science, just as computer science did about 60 years ago from mathematics, electrical engineering, and physics.
“I argue that the core concept in computer science is the algorithm—a step-by-step procedure for solving tasks that is guaranteed to complete in a finite amount of time and give you a correct solution when it does,” says Stent. “So, if the core concept in computer science is the algorithm, then maybe AI is the field of solving problems that do not have algorithmic solutions, such as protein folding, natural language processing, or computer vision.”
Stent hopes AI will be a broadly interdisciplinary field encompassing engineering, science, ethics, and applications. However, for the burgeoning field to reach its potential, Stent says changes need to happen.
“We have to grow up,” says Stent. “We call ourselves engineers, but we reject licensure, certification, and formal on-the-job training. We call ourselves scientists, but in AI, we say, ‘oh, we can’t be bothered to wait for peer review. Reproducibility isn’t important.’”
Stent served as a member on the 2020–22 National Academies Committee studying Responsible Computing Research. Two of the committee’s key recommendations were to reshape computing research to adequately include diverse perspectives—such as those from sociologists, anthropologists, and historians—in research projects from the outset, and to provide researchers with access to knowledge and expertise needed to identify and assess the ethical and social implications of their work.
Stent urges fellow researchers to reckon with the impact of their work, both on society more broadly, and the planet.
“We’re going to see an increasing focus on efficiency because this current trajectory is wildly unsustainable,” says Stent. “There’s an idea that we should just reopen all of our nuclear power plants in order to fuel bigger and bigger models that don’t really have more explanatory power. I find this concept incredibly wasteful and not very scientific.”