If Jason Mars Has His Way, no one will have money problems ever again. We’ll all be diligent savers, with robust retirement portfolios, never losing track of what goes in and out of our bank accounts.
It’s an ambitious goal: According to Bankrate.com only 37 percent of people have enough savings to pay for a $500 emergency, while the American Psychological Association says 90 percent of Americans are stressed about money. Mars, though, is convinced that he holds the key to changing a person’s relationship with money.
How? Through artificial intelligence (A.I.).
You may think that A.I. is part of some sci-fi future, but the technology is becoming increasingly common. Mark Stevenson, author of “An Optimist’s Tour of the Future,” thinks we’re already deep into a world where computers are ubiquitous and smart enough to handle some of our most important tasks. “We’re completely surrounded by A.I. and machines that do things for us,” he says.
Still, it wasn’t long ago that managing critical things such as personal finances meant sitting down with a pen and paper or an online spreadsheet and manually inputting one’s income and expenses. While that may have worked for the few people who didn’t mind sorting through all those documents, it wasn’t a practical solution for anyone in a hurry.
This barrier to budgeting has caused problems. “There’s a lack of awareness as to what’s going on with people’s financial story,” Mars says. “But we think A.I. can actually do something transformative in that domain.”
In 2015, Mars started his own A.I. company, Clinc, which builds digital financial applications that use natural language processing — technology that allows people to have a conversation with a computer. A year later, Clinc launched a financial program called Finie, which lets people interact with their money via voice commands.
It’s like having a financial adviser, a bank teller and that financially-savvy friend all rolled into one — and right inside your home. You can tell Finie to transfer money from one account to another or instruct it to send money to a family member. Finie can tell you how much you’re spending and when you might want to consider dialing back. Quickly, financial institutions have taken note, and they have asked Clinc to incorporate Finie into their digital banking platforms.
The point is not to be lectured by a machine, says Mars, but rather to change bad habits and help motivate people to improve their lives. “If you know you’re spending 20 percent more this week on something, that knowledge gives you the chance to change that behavior,” he says. “It’s about empowering people.”
So, How Does A.I. Work?
So, how does A.I. work? In short, large amounts of information are fed into computer systems that identify patterns that humans would take impossibly long to detect. Clinc is just one example of how savvy businesses are using this technology.
In the financial world, A.I. is already analyzing decades’ worth of market data, Twitter sentiment and trading information to give fund managers a better grasp of when to buy and sell securities or why one particular stock might be better than another.
“A.I. makes prediction easier,” says Joshua Gans, a professor at Toronto’s Rotman School of Management and author of “Prediction Machines: The Simple Economics of Artificial Intelligence.” “Wherever companies face uncertainty in their decisions, which is pretty much everywhere, A.I. has the potential to support better decision-making,” he explains.
That’s why it’s transforming other sectors too. Manufacturing operations use A.I.-enabled, self-learning programs on their production lines. And the health care sector relies on A.I. to analyze millions of data points, including genome information and research to help identify patterns in a variety of diseases.
Many companies have been gathering data on their customers for years, and A.I. is changing how truly innovative companies can responsibly use that data, says Amir Khosrowshahi, chief technology officer and vice-president of Intel’s artificial intelligence products group.
“There’s data about the customer, their past purchases, when they click, and if they have called customer service before,” he says. “And that data is actionable. We can extract knowledge from it, and we can deliver a much better experience.”
MOVING A.I. AHEAD
How do we get A.I. from today to tomorrow?
As for A.I. reaching its full potential, we still have a long way to go. If progress could be viewed along a one to 1,000 scale, Gans says we’ve only moved from a one to a three. A.I. could potentially be used in everything from environmental research — understanding climate change patterns so cities can better plan for natural disasters — to health care, where people may be able to detect illnesses before they cause complications.
In recent years, Khosrowshahi and his colleagues at Intel have been developing technologies that other industries can use to power A.I.-related programs. “We have a long history of building general-purpose computing, and a lot of other hardware and software technologies that are very relevant to A.I.,” he says.
Intel recently announced the Nervana Neural Network Processor (NNP), a chip modeled after the human brain. The company thinks it will help computers and A.I. systems make decisions based on observed patterns and associations. “Everything that gets these systems to work — neural networks, algorithms, computer science — is a core technology that we’re building,” Amir says.
WE HAVE A LONG HISTORY OF BUILDING GENERAL-PURPOSE COMPUTING, AND A LOT OF OTHER TECHNOLOGIES THAT ARE VERY RELEVANT TO A.I.
Amir Khosrowshahi, vice president of Intel’s artificial intelligence products group.
Intel’s chips, and the company’s deep understanding of advanced computing, have been key to Mars’s work at Clinc. The chip maker’s processors have helped him create the kind of sophisticated natural language processing needed for the A.I. in Finie, while Intel’s understanding of advanced computing networks “is a powerhouse that has helped us realize our vision,” Mars says. Intel has also collaborated with Mars in his role as a computer sciences professor at the University of Michigan in Ann Arbor. Most recently that included funding a $4.6 million grant to study computer vision, which will drive understanding of how A.I. might better catalogue and identify objects in the physical world.
There’s plenty of work to do before A.I.’s full potential is realized. For Mars, that means using it to improve his product and thus people’s financial health. “For A.I. to be able to tell you exactly how much you spent on something this time versus last time. That is a transformative kind of experience,” he says.