(By Andrew McAfee)
“So what do our brains do in such cases? How do they come up with better answers? As far as I can tell, we aren’t sure. But we’re clearly doing something that our best digital technologists have not yet been able to master. The same seems to be true, at least for now, in many domains that require taste, creativity, or an aesthetic or emotional response. Computers still can’t write a good short story, or design a beautiful computer“.
People and computers are coming together in all kinds of interesting ways these days. The right combination of human and digital smarts in chess will beat the top grandmaster, the best chess supercomputer, and the top grandmaster with the best supercomputer. At least one VC firm is giving an algorithm a formal vote on its investments. And robots (which I consider to be computers with a physical presence) are increasingly working side by side with people on factory and warehouse floors.
In some cases it’s clear what each party brings to the collaboration. Because humans still have greater manual dexterity they’re the ones picking parts out of bins in the newest Amazon warehouses, while Kiva robots bring the shelves full of bins to the people quickly and reliably. The VC algorithm, if properly constructed, will systematically and objectively take into account“prospective companies’ financing… intellectual property and previous funding rounds” in a way that might be hard for biased, pressed-for-time humans to replicate. And chess computers keep human players from some kinds of dumb moves — the ones whose negative future consequences should have been foreseen, but weren’t.
But alchemy between people and computers — combinations that are way better than either party could do on its own — remains mysterious. In particular, it’s not clear to me (and many others) how people continue to add value as technology races ahead. Computers are clearly better at brute force computation and search, and their pattern matching abilities are improving by leaps and bounds these days. So what are we better at?
That’s a surprisingly hard question to nail down. It appears that when the task is so wide open that searching through history or enumerating all the possibilities won’t work, our abilities are superior. In domains as diverse as playing the Asian board game Go and predicting how proteins will fold, the human brain is still the best tool available. In both of these cases, there are just too many possibilities for even a network of supercomputers to go through all of them.
So what do our brains do in such cases? How do they come up with better answers? As far as I can tell, we aren’t sure. But we’re clearly doing something that our best digital technologists have not yet been able to master. The same seems to be true, at least for now, in many domains that require taste, creativity, or an aesthetic or emotional response. Computers still can’t write a good short story, or design a beautiful computer.
Will they learn to? As Erik Brynjolfsson and I wrote in The Second Machine Age, the mantra we learned from studying many examples of digital progress is “never say never.” But I haven’t seen these things yet, which gives me hope that people will have important roles to play in our societies and economies for some time to come.
“Opinion pieces of this sort published on RISE Networks are those of the original authors and do not in anyway represent the thoughts, beliefs and ideas of RISE Networks.”