For Tuomas Sandholm, this is a clear case. The co-director of Artificial Intelligence (CMU AI) at Carnegie Mellon University is firmly convinced that the advantages of artificial intelligence far outweigh the disadvantages.

“AI can make better decisions than humans, which will make the world a better place,” he says.

To better understand his perspective, take a look at his track record.

In 2010, Angel Jordan, a professor of computer science at Carnegie Mellon University—one of many titles he holds—developed the algorithm for the national kidney exchange pro bono. His algorithmic techniques are also behind the global kidney exchange, which has saved 10,000 lives to date.

The medical field is one of the biggest beneficiaries of AI today and in the future, he says. “Many decisions will be AI-supported. For one thing, it’s hard for humans to absorb all the medical knowledge today. There’s too much! AI can sift through a lot of medical knowledge and treatments.”

In addition, AI can process huge amounts of data to arrive at holistic decisions that mere human intelligence cannot. “This is where organ allocation comes into play,” says Sandholm. “Individual doctors only care for one or a few patients in their transplant centers. They cannot make the best decisions from a national perspective. AI is able to take the entire national perspective into account.”

Another AI milestone that has been described as “superhuman” is leading the team that developed the AI ​​program Libratus, which defeated the top four poker players in the world during a 20-day no-limit Texas Hold’em marathon in 2017 at the Carnegie Science Center in Pittsburgh.

“It was a benchmark,” he says of the legendary poker match, “representative of a much larger class of strategic thinking. It proved that AI is better than humans at strategic thinking. Since then, we’ve seen it in other applications as well, mainly in multiplayer games like various defense strategies.”

Sandholm was, of course, already busy working on these projects. He founded five companies, including Strategy Robot, a Pittsburgh-based company with 10 employees that developed the software now used by the U.S. Department of Defense. He describes his company as “a company that brings the game theory paradigm to planning against adversaries. So far we have only worked with U.S. customers, but we are open to working with other NATO countries as well.”

The company, Sandholm said, is profitable without ever having invested venture capital. A related company, Strategic Machine, which he also founded and runs, focuses on computational game theory for business applications. And another, Optimized Markets, optimizes ad sales, campaigns, and inventory and pricing allocation for television, internet and more.

In 2010, the serial entrepreneur sold a company, the highly successful CombineNet, for an amount he kept confidential. Three years later, the company was sold again for $43 million.

The innovative technology behind CombineNet enables customers to model complex scenarios with different objectives and constraints to determine optimal procurement strategies. Simply put, it’s about finding the best places to buy goods at the best prices and in the best quality. The result? Significantly higher efficiency with significant cost savings.

One example is Proctor and Gamble, which optimized its truck transport in North America by using so-called combinatorial auctions, thereby achieving maximum efficiency in route planning, loading and scheduling of transport.

With 29 patents and more than 500 scientific papers, the personable professor is clearly at the top of his field, especially in the academic and research world of AI. When you point this out to him, he quickly adds: “With all the people who work in my startup companies.” For the sake of completeness, it should be mentioned that over the years he has hired at least 400 of these people in his companies and for his research.

Sandholm came to the United States from Finland in 1992 as a doctoral student and, after stints in St. Louis and Los Angeles, landed at Carnegie Mellon in 2000. He has many specialties in subfields of AI, especially game theory (think of processing data to find the best strategy from many options), strategic thinking, and automated negotiations.

There is more to the man. In 1986 he won the title of best windsurfer in Finland. The following year he won the Finnish championship in this sport that is extremely popular in Europe. He continues to nurture his love of racing on the water as a member of the Moraine Sailing Club, with whom he travels to races with his team.

He also served as a pilot in the Finnish Navy. “There was conscription in Finland,” he says when the subject is brought up. He waits until he is asked about it before revealing: “I flew a Vinka, a Finnish acrobatic robot training aircraft” – which he admits was great fun.

When asked if he has ever failed, he laughs and says, “Many, many times.” When asked for an example, he thinks for a moment and then says, “When we played Texas Hold’em in 2015, we couldn’t beat them.”

That’s true. But he turned that defeat into a resounding success by winning in 2017 when they “developed better algorithms.” The victory was hailed as a “superhuman milestone.”

Although he is a big supporter of AI and its proven ability to change our lives for the better, Sandholm warns of the potentially negative consequences of AI. In his opinion, the greatest danger comes from large language models.

“Like ChatGPT,” he adds.

If you ask ChatGPT to define itself, the answer is:

“Large language models like mine are advanced artificial intelligence systems designed to understand and generate human-like text. They are built using massive amounts of text data and sophisticated algorithms. These models can understand context, generate answers, and even perform tasks like translation or summarization.”

Sandholm doesn’t think so. “People think that large language models are intelligent because they sound so convincing, even though they are hallucinating.”

Hallucination is pretty much what it sounds like – the term for blurting out things that aren’t true. For example, ChatGPT once claimed that Sandholm was an expert in computer vision – an anecdote someone brought up at a meeting he was attending – but he’s not, he says.

“It’s often wrong,” he says. “They argue in well-crafted language that they’re right even when they’re wrong. We have to be careful with this kind of AI, and it’s becoming increasingly popular.”

What is being done to combat disinformation?

“There’s a lot of work going on, and we’re doing that too, to figure out how to stop large language models from saying things you don’t want them to say – how to build a bomb, a classic example, and how to prevent hallucinations.”

Although Sandholm has witnessed profound changes brought about by AI throughout his career, he has also witnessed major changes in the Pittsburgh area since arriving more than two decades ago.

You only have to look at his university, he emphasizes: “At CMU, we are number one in computer science, together with MIT, Stanford and Berkeley. And in the subfield of computer science, AI, we are the only number one.”

He notes the presence of larger technology companies in Pittsburgh, such as Meta/Facebook, as well as his many colleagues and students working on various projects. “And of course my own company,” he adds with a mischievous smile. “We’re hiring!”

For the talented Sandholm, it was “very, very easy to get a visa” and move to the United States as a graduate student. He came to study multi-agent systems with Victor Lesser at the University of Massachusetts Amherst after, not surprisingly, researching the best people in the field in the sector.

Would it be the same for others in similar situations to obtain a visa today?

“It would be easy for someone like me to get the visa because there are all kinds of programs that give preference to people with specialized skills. If you’re a general programmer, especially from a country that the U.S. doesn’t like, it would be very difficult.”

In the battle for talent, any city in the country – or rather the world – would consider Tuomas Sandholm a top-quality catch. His contributions to Carnegie Mellon and Pittsburgh were invaluable.

He remains a staunch Pittsburgh fan, but would like to see the city do more to clean the air, which he believes hinders local businesses’ recruitment efforts.

Sandholm lives happily in Shadyside with his wife, Christina Fong Sandholm, a PhD in economics from Carnegie Mellon University, and their two teenage daughters.

His older daughter, he proudly tells us, has just been accepted to her dream university for the upcoming fall semester: Carnegie Mellon.

Your major? Computer science.