Recently, I found in my hometown, general-circulation newspaper stories about new technologies and economic trends that banks will to have to deal with or that raise issues—and possibilities—bankers and payments organizations should probably be thinking about.
Dell is creating a new $1 billion division that will focus on the Internet of Things, forecasting that in the next seven years there will be a trillion connected IoT devices. With 40% of the world unbanked, that’s 200 IoT devices for every bank account, each pumping out messages every few seconds. Could the technology of a typical financial institution scan thousands of messages per account and flag those that affect one of their transactions even when it’s not addressed to the bank or its customer? And fraud-check each one? And make the correct decision about what to do?
The IMF forecasts global economic growth. Trade, which generates complex payments requirements, is predicted to grow 4.2%. How many banks (many of which are still climbing back from the financial crisis and recession) are budgeting for a 4% or greater increase in technology capacity next year?
Nvidia’s Pegasus minicomputer claims a processing speed of 320 trillion transactions per second. While initially programmed for driverless driving, not banker-less banking, this supercomputer isn’t (yet) being programmed for transaction processing. But recently BNP Paribas bought a Nvidia machine to run option-pricing models. Running the predicted flood of IoT-generated payments-related data and the jump in complex trade payments could be accomplished using a fraction of that Pegasus’s computing power.
Intel touts new quantum chip. Another article relates how Intel, a grand-daddy of the semiconductor business, has released to its testers a 17-qubit chip. How big is this in classical bytes? The question is probably irrelevant. The whole idea of quantum computing is doing an unlimited number of calculations on a given amount of silicon. Quantum computing is about understanding phenomena in the natural world, instantly. Is this something a bank doesn’t need to be concerned about? Only if the financial institution cares solely about counting things. But banks that think counting is all their customers need probably won’t be around too much longer.
The customers of financial institutions are going to be asking for help understanding such “natural phenomena” as how people make decisions when things are uncertain. Or what the long-range weather forecast means for the agricultural industry the bank lends to and processes payments for.
The reading of one day’s news with my morning coffee illustrates the kind of thinking that is going to be required if bankers are to exploit this coming explosion in computing power.
A colleague who works at U.C. Berkeley’s Lawrence Radiation Lab likes to pose a question that strikes me as an appropriate way to think about all this. He asks: What does a supercomputer dream? While there is probably not a definitive answer, chasing the question could be a good exercise for the next board meeting of your bank, credit union, or fintech. It could serve to identify the person in the room who is thinking like the Nvidias, the Dells, and the Intels. The future, some claim, is about fully computerized banks not needing human employees, a vision that many simply don’t believe. But then, 10 years ago, driverless cars seemed easy to scoff at too.
Conference Notes: Highlights at the Bank Administration Institute’s Beacon 2017 conference included Oliver Wyman’s forecast of more attention from the Office of the Comptroller of the Currency to fair lending, the Federal Deposit Insurance Corp. focusing on cybersecurity issues, and the Consumer Financial Protection Bureau taking up debt-collection practices next.
Pega demoed an application of artificial intelligence where, upon noticing photos of Corvette cars on a consumer’s social-media page, the AI software triggers targeted offers of three different auto-loan packages from financial institutions. It’s something Pega calls focusing on the customer’s journey rather than the bank’s.
—George Warfel • GWarfel@haddonhillgroup.com