I recently attended the annual meeting of a private equity firm and as the newest company in their financial services portfolio, we were the center of attention as our investment partners wanted to learn more about "Fintech" and the much-talked-about small business alternative finance space. In particular, they wanted to know all about the buzz surrounding machine intelligence, automated underwriting, predictive analytics and the magic bullet of our times: "algorithms."
Technology: A Lens to Amplify Thought
As the evening's networking began, I stood in anticipation among a group of incredibly sophisticated financiers as they excitedly approached to learn all about the miracles of technology in changing the world of small business finance. Everyone was well-versed in the subject as there has been a great deal of media interest in this industry, with online alternative funding for small businesses growing about 175 percent a year. Questions flew on about how algorithms are able to underwrite the hundreds of applicants seeking small business financing without the touch of a single human. I've been at this for a long time and have seen tens of thousands of deals, but every single one of them, regardless of how much technology was used, still involved a human decision. The truth is that success in profitability, growth, scalability and low default rates is because of the "bio-computers"- the underwriters - who call upon their human experience and insight, and who are supported by sophisticated technology.
Some seemed surprised by my answer. They cite that surely in this age of "Big Data," analytics and the likes of Watson supercomputers, algorithms could outperform humans in making speedy credit decisions.
But this isn't the case. One of our board members and a veteran of Silicon Valley, Tom Scoville, did extensive work in developing predictive analytics and neural networks. He explained it most succinctly saying:
Technology is a lens. As the telescope was to the eye, the computer is to the mind; a lens to amplify thought. The silicon computer extends and expands the capability of the human brain, and can automate much of the grunt-work associated with complex problem solving. But the mind has to remain at the center of the enterprise, even if it is surrounded by silicon computers.
A Test of Knowledge
The point is further driven home in a recent National Geographic article written by Roff Smith on the London cab drivers who not only demonstrate the true capacity and capabilities of the human brain, but have also overridden technology consistently. Because of these drivers' ability to hold vast amounts of data in their heads and their possession of instincts, they hold insights and experience that no machine can duplicate. Before being granted the coveted green badge awarded to a London cabbie, one must pass "The Knowledge," a grueling test of endurance, memorization and endless testing. Not only do candidates need to memorize over 25,000 streets, roads, avenues, boulevards, back alleys and over 20,000 landmarks, pubs, restaurants and clubs, but they must also know exactly where they are, the direction of each street and nuances including which side of the road the destination is located. It takes years to achieve and thousands of hours of practice runs. Less than 20 percent of those who try eventually succeed.
Smith questioned why in an age of GPS-equipped cars, smartphones, Google Maps and every other imaginable device would one have to possess "The Knowledge"? Couldn't cab drivers just type an address into the GPS and get to their destination? This was a true challenge of man versus machine.
In the article, Smith reported:
In May, London's Guardian newspaper pitted a cabbie against a Sat-Nav-equipped driver from Uber, the new 'taxi' company that allows passengers to book cars via their smartphones. The Uber driver did the run from the newspaper's office in King's Cross to Big Ben, in Westminster, in 22 minutes; the cabbie did it in 18, by taking a slightly longer route he knew to be quicker.
Why? Humans possess knowledge that algorithms and cognitive computing will never have. Instincts, experience, imagination, emotions, insight and an unexplained second sense that some call "gut feelings." Our neural net can spot patterns, but it cannot possess the traits that make the human experience "human."
The Importance of Human Insight
As someone who's been involved in business finance for over 30 years, first as a business owner seeking capital and later as someone financing other enterprises, the core credit decisions usually lie within the financial condition of each company, the business model and certain performance metrics. However, the most important element is the management of the company. In every single transaction that I have been involved in, people made the decisions for the company and it was their judgment driving the health of that business.
Fintech or human insight? The simple logic remains that people make the best decisions concerning people and the small businesses they run. We supercharge our data collection and availability, but in the end, it is still a human looking through that telescope of technology. Business quality comes down to the business owner's judgment, character, integrity and quality standards. Hard data can be quantified and fed to an algorithm, but qualitative characteristics cannot. Patterns of behavior can be set in a neural net, but in quest of a credit decision for a small business, it's all about the people.