The Language of Pattern Recognition
Ask a businessperson, “What exactly do you do?” and the answer will probably be, “I make decisions!” Ask a scientist, and the response will be, “I discover truth!” Actually, both of them are on the same quest-the search for patterns.
Every day scientists come to my office with their business plans-and lie to me. They tell me that their patents are solid, that they have no competition, that the technical discoveries needed are within the state of the art, and that their market is going to grow ten thousandfold in the next year. That’s okay. I spend part of my day lying back to them. I tell them that my money is different, that there is no need for me to sign a confidentiality agreement, and that I am their new best friend. On the surface, the scientists and I couldn’t be more different.
Like every other hard-nosed venture capitalist, I am, of course, trying to guess just when technology will break open and markets will emerge. If I pick the right sector, and pick it early, I will make 10 times my investment. The people over at MIT’s Whitehead Institute for Biomedical Research are also looking for patterns: Where should they take their research? What amount of computing power can unlock secrets to the human genome sooner?
The Great Ones of science connect the dots faster than their peers. Often it is because they can extract more insight from a set of facts or generate data that reveal more about how the world works. If the conventional wisdom was that poisons kill, then the contrarian’s view that low doses can be beneficial was at first scary. The field of hormesis exists because someone noticed that some plants thrived in situations of chemical adversity. The possibilities were out there for anyone to see, but few did. Hormesis basically says that a little poison can help you. The German philosopher Friedrich Nietzsche wrote, “That which does not kill us makes us stronger,” but it took some clever pattern recognition to find this also works in plants.
A friend of mine, Gordon Mathews, came back from lunch one day and was handed a fistful of pink while-you-were-out slips. He went into his office and invented voice mail. Seeing that the pattern of messages and the advance of computer technology could be linked, he came up with the first significant voice application since the telephone. Mathews recognized not one but three patterns. First, 80 percent of business telephone calls were not completed in real time because Party A did not actually talk to Party B. Second, with secretaries becoming a vanishing breed, more and more managers were placing and answering their own calls. Third, e-mail had indicated that a store-and-forward technology could provide immense productivity. Combined, these patterns led to a multibillion-dollar business.
Scientific methodology and business practices do essentially the same thing: make a hypothesis about the way things are, then gather data to confirm or disprove the hypothesis. The trick is in reading the data points better or faster-or gathering more data-than anyone else. In a scenario akin to the old television game show Name That Tune, the winner is the one who needs the fewest clues to make out the big picture.
Breakthroughs often occur when scientists or businesspeople can cascade developments on top of each other. A businessperson may find that neural computer technology and data mining can yield a way to determine who is most likely not to repay a loan; a scientist may use the same methodology to discover a better way to fabricate semiconductors. Neural networks deduce patterns and relationships that are not immediately obvious. Who would guess that people born between 1944 and 1964, and whose last names end in vowels, would be three times likelier than the general population to repay bank loans on time?
Businesspeople, however, view themselves not as pattern recognizers but as insightful, tough managers who are competing in a masterful game of cat and mouse. What they really do is sample data to determine whether there are changes- ever so slight-that signify a shift in relative position. If so, they act. Act too precipitously, they die. (Example: Early pen-based computing products failed. But just three years later, the Palm Pilot was a winner.) Act too slowly, however, and they die. This Darwinian view restrains scientific advancement: rush to judgment with data that cannot be replicated, and you will never get tenure; you will wind up doing administrative dog work.
Businesspeople and scientists abhor risks. Find someone of either genus who went against the grain, and you find someone who read the tea leaves a little differently-and saw a pattern no one else did.
This article originally appeared in the MIT Technology Insider, a monthly newsletter covering MIT research and commercial spinoff activity.
Geoffrey Hinton tells us why he’s now scared of the tech he helped build
“I have suddenly switched my views on whether these things are going to be more intelligent than us.”
Meet the people who use Notion to plan their whole lives
The workplace tool’s appeal extends far beyond organizing work projects. Many users find it’s just as useful for managing their free time.
Learning to code isn’t enough
Historically, learn-to-code efforts have provided opportunities for the few, but new efforts are aiming to be inclusive.
Deep learning pioneer Geoffrey Hinton has quit Google
Hinton will be speaking at EmTech Digital on Wednesday.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.