This Tech Entrepreneur Founded One Of Boston's Largest AI Firms. However, He Believes That The West Coast Strategy Is The Way Of The Future

Talk to Jeremy Wertheimer if you want to understand the shifting landscape of artificial intelligence — and Boston’s role in it.

In 1996, the MIT graduate founded ITA Software, one of the region’s largest AI firms. ITA, which pioneered technology that allows customers to compare airline flights and rates online, was purchased by Google for $700 million in 2010 and utilized to anchor and develop the search giant’s presence in Kendall Square.

ITA’s initial AI program had approximately 2 million lines of code. Wertheimer, on the other hand, claims that if he were to establish the company today, the full program for sifting and comparing thousands of flights every day could be written in just 100 lines of code — or even 10.

“Back in the day, we called it good old-fashioned AI,” said Wertheimer, who is now a visiting scientist at the Broad Institute. “However, in the future, all that clever coding will be forgotten.” You want an extremely basic software with enough data and computational power.”

Cambridge, MA. 06/21/23 – Jeremy Wertheimer (cq) has been an essential figure on the Boston AI scene for decades, first at his flight shopping company ITA Software, then at Google in Cambridge, and now mentoring and funding new AI startups here. (Lane Turner/Globe Staff) Reporter: (Aaron Pressman) Topic: (XXWERTHEIMERAI )

This is a tribute not only to modern computing improvements, but also to a fundamental shift in how AI systems are created, a shift that lies at the heart of ChatGPT and other generative AI applications. And it explains why the Boston area, which has been highly involved in the older approach, has not been at the forefront of generative AI, a potentially trillion-dollar business.

How two Olin College students sparked the AI chatbot revolution
An MIT student requested that AI make her headshot look more ‘professional.’ It made her complexion lighter and her eyes blue.

13 current applications of artificial intelligence in the Boston area
Recently, the majority of the generative AI and machine learning entrepreneurial activity has occurred in other locations. According to market tracker CB Insights, only five of the 204 recent venture-capital-backed businesses leveraging generative AI technologies are based in Massachusetts. There are 112 in California, 37 in New York, and eight in Texas. (Another 12 were identified with Delaware headquarters, most likely for legal and tax purposes, but their actual firms could be located elsewhere.)

And, of the 13 firms that have achieved “unicorn” status, defined as being worth $1 billion or more, seven are in California, with none in Massachusetts.

Wertheimer, a native New Yorker who attended computer trade exhibitions in his teens and attended Cooper Union in Manhattan for college, joined MIT’s Artificial Intelligence Lab in the 1980s. For decades, the lab’s primary focus has been on developing complex “smart” systems to handle challenges such as comprehending human speech, building autonomous robots, and detecting individuals in photographs.

Marvin Minsky and John McCarthy, both MIT academics who eventually transferred to Stanford, were among the first researchers to create the term “artificial intelligence” in the 1950s. MIT then created LISP, a computer programming language designed exclusively for developing AI systems. (The vast majority of ITA’s software was written in LISP.)

“That’s the world we came from,” Wertheimer explained. Minsky was Minsky’s graduate adviser at MIT and a member of his PhD dissertation committee. Wertheimer’s 1996 thesis investigated an AI system for answering particular questions in a large body of published biological research.

Wertheimer interned at Cambridge software startup Interleaf in the early 1990s, when Paul English, who later cofounded travel-search behemoth Kayak, was vice president of engineering. “Crazy smart,” English remembered Wertheimer. “At one point, he was writing some ideas on a whiteboard in front of a few of us involving some math, and after he left, the rest of us looked at the board to try to figure it out.”

Meanwhile, machine learning, a competitive approach to developing AI programs, was gaining traction. Inspired by the human brain, these “neural network” algorithms taught themselves to execute tasks by evaluating vast volumes of data. However, Minsky and others found neural networks insufficient in the 1960s, and the concept fell out of favor for decades. It wasn’t until the last 10 to 15 years that academics in New York and Canada collaborated with Google and Meta to make significant advances in machine learning for AI.

As a result, until recently, scholars at MIT and other schools mostly continued to address AI difficulties by building intricate programs, from speech recognition to chess to predicting protein structures.

ITA’s software, which powered major airline websites and the travel site Orbitz, scanned flight schedules and associated ticket rules using a formula known as a junction tree algorithm, which basically produced a massive map of the data. The application was capable of significantly more complex and thorough searches than the industry’s previous, mainframe-based SABRE system.

Wertheimer was inspired to enhance aviation software by a summer roommate who showed him the limits of SABRE in 1992. Following the completion of his dissertation in 1996, he enlisted the help of fellow MIT AI Lab graduate students Carl de Marcken and Dave Baggett to create a prototype.

Peter Miller, a former top Lotus Development executive, was considering launching a travel software firm when he learned about Wertheimer’s endeavor.

“He was a brilliant guy with this prototype that could do everything,” Miller, an MIT graduate, recalled recently. “The moment he told me about it, I thought we had to acquire this.” When Miller’s investors refused to purchase the software, Wertheimer created ITA with his MIT friends and turned it into a powerhouse.

When Wertheimer joined Google in 2010, he was put in charge of travel. The addition of ITA doubled the size of Google’s Cambridge office and elevated its significance inside the firm as the worldwide headquarters of the travel endeavor. During his eight years at Google, Wertheimer witnessed the company grow from 20,000 to 200,000 personnel. His group eventually became a part of the advertising industry, and he left.

Unsurprisingly, Wertheimer has abandoned “old-fashioned AI.” “None of that is the future,” he stated emphatically. His most recent venture will employ machine learning to create new medications. It’s regarded as one of the most promising applications of artificial intelligence, and Wertheimer will face competition from dozens of startups, not to mention in-house research at huge pharmaceutical corporations.

“This is a fantastic opportunity, but there has already been a lot of work done in that area,” said Miller, who is now a startup CEO coach.

Wertheimer’s new firm intends to capitalize on the area’s strong life sciences environment, and startups that want to leverage apps like ChatGPT to help certain industries are beginning to get a foothold here as a result of the region’s strengths in health care, financial services, and climate tech.

Portrait Analytics, another local machine-learning business, originates from the financial services sector. David Plon, cofounder and CEO, began his career as an investment analyst at Boston hedge fund firm Baupost Group. His startup seeks to employ chatbot technology to assist analysts go through mountains of company research more efficiently.

Portrait Analytics raised $7 million last month from two Boston venture capital firms, Unusual Ventures and.406 Ventures, as well as a group of hedge fund experts.

“There’s a nice community of Boston-based investors that are excited about this product,” Plon stated. And, contrary to popular belief, “there’s a really strong AI community in Boston,” he insisted.

However, with ITA being acquired by Google, iRobot being acquired by Amazon, and Nuance Communications being acquired by Microsoft, the old guard of Massachusetts AI businesses will not be leading the next wave of innovation.

Indeed, Wertheimer’s time at Google exposed him to some of Silicon Valley’s tech mentalities’ characteristics, most notably a lesser fear of risk-taking. But, despite West Coast AI firms’ present dominance, he’s not going anywhere.

“I spent my whole career in Kendall Square within a couple-block radius,” he stated. “I thought when I came for grad school, I would go back to New York, of course, but I never did.”

This Tech Entrepreneur Founded One Of Boston’s Largest AI Firms. However, He Believes That The West Coast Strategy Is The Way Of The Future

Talk to Jeremy Wertheimer if you want to understand the shifting landscape of artificial intelligence — and Boston’s role in it.

Download The Radiant App To Start Watching!