Potentially in quantum computing there are some answers to why classical computational algorithms might not be able to get certain human reasoning tasks done. Nonetheless, recognizing the depth and subtlety of the qualitative domain, and recognizing it is joint and separate from the quantitative domain, this lets you start seeing problems differently. There’s lots of interest in Salesforce, other CRM products, web analytics. These are all essentially domains where we’re building a quantitative universe — the space of big data is where we build a quantitative universe, and then we study it. One approach to this is to say, let’s get as much data as possible, and then lets develop the most sophisticated algorithms possible for finding patterns in the sea of quantitative data. But I ultimately don’t believe that will be terribly successful.
I think the much more important questions are, let’s study the human aspects — the qualitative aspects — of the problem. What are we trying to get done? What’s actually happening? What are the subtle aspects of this process that, when we actually clarify, then we can learn that is where we want to collect the data. That’s what’s what we want to analyze and figure out. Through the study of the qualitative phenomenon that dovetails right next to this quantitative phenomenon in the universe, through this we can optimize how we use our computers to do what what we want them to do. This is where I think we are going to see a lot of technology companies in the 21st century. I think this is much closer to the 21st century big data analysis problem, at least much more so than getting more and more fancy algorithms to do the same problems we’ve already seen.
That was from an excellent 2013 podcast, “The Path to Palantir” with Stephen Cohen on Entrepreneurial Thought Leaders. I think he’s exactly right. Too often people speak as though the qualitative domain can be subsumed by the quantitative once we get more data and improved algorithms. This belief has reoccurred throughout the years. Indeed, the Department of Defense was founded by ‘whiz kids’ who thought there would no longer be a need for military experience in optimizing weapons choices. The structure of the DoD today reflects that belief, although no one has actually tried to operate that way since the 1961-1968 period. So real people in the DoD are left to stumble their way through a decision process that presumes the world is one big quantitative problem.
I highly doubt that AI will be a one-stop shop for answers except in narrow and well-defined applications. Yet in those areas, it will be highly productive. And a tremendous task will be ever-increasing the quantitative domain by correctly framing and curating the qualitative. We need people who know specific problems deeply and can be generative enough to redefine what is needed from data.
Here’s another good part:
Question: So you said there’s a line dividing the qualitative and quantitative domains that an algorithm can approach. Does that mean that the line itself — where you draw it — is a qualitative line and if so, can you make an algorithm for redefining qualitative problems to make them quantitative?
Cohen: This is a great question… I think the line is constantly moving. The logical issue you highlight is that I am defining qualitative negatively. If we could define a qualitative phenomenon in a very tight, precise description, then of course we can just turn it into a quantitative phenomenon and use a computer to solve it… In essence, there are these ambiguous aspects to these human phenomena that we experience all the time including subjective phenomena that because they defy precise logical descriptions we cannot turn them into algorithms. I don’t see that phenomenon being fundamentally tackle-able by these quantitative approaches but we can definitely chip away at it.
… Ultimately, there’s a lot of deep, subtle stuff that the users ultimately don’t know what the product is that they want. How could they? That’s why you’re the entrepreneur. If they knew, they’d just make it themselves. They wouldn’t really need you. The only way you seek satisfactory answers to what’s the right product, what’s the right features, is you seek these subtle aspects, study in their own right, and appreciate the depth of them as a phenomenon unto themselves — not just crassly reify them into a logical conception that resembles a much more friendly quantitative truism, if that makes sense.
Leave a Reply