As industries mature, they distribute intelligence differently.
What do I mean by that? Here’s a diagram that I sketched out when I was at the CEEC Workshop last month:
It shows that we can build intelligence about the system in two places – in people or in technology (which can be any kind of tool). I use “Dumb” and “Smart” to indicate relatively high or low levels of embedded intelligence.
Here is how it works. As industries mature, they go through the stages outlined by the numbers. I’ll use mining as an example, because that’s the industry I was thinking about when I first drew this.
- Stage 1: Dumb People, Dumb Technology: this is where we always start. In mining, this was where we were for a long time, probably up until around the late 19th century. Prior to that, there wasn’t really a science of mining. No one knew too much where to dig for what, and the tools were primitive. If you’re in this stage, the industry is still very young, and the number one objective is to learn. You have to figure out what the system is good for, and how to build it.
- Stage 2: Smart People, Dumb Technology: As people learn, they get smarter. But the tech is still dumb. Mining hit this phase around 1890 or so. That’s when they started doing things like sorting rocks before crushing them to improve yields. They also started to invent new technology, like the flotation process. If you’re in this stage, the number one objective is to find the business model that will work best.
- Stage 3: Dumb People, Smart Technology: as the industry matures, it makes sense to start embedding a lot of the intelligence into the technology. We have been seeing this to an increasing extent in mining over the past 40 years or so. This is not to say that the people currently working in mining are dumb, far from it. It’s just that the systems are being designed as much as possible to not need much input from people – think of the driverless trucks, and so on. This is a tricky stage to be in, because it seems like a final state. But it’s dangerous, because when all of the intelligence is embedded in the technology, the system tends to collapse. Consequently, if you’re in this stage, the number one objective is to move on to Stage 4.
- Stage 4: Smart People, Smart Technology: this should be the target end point for a mature industry. We get the best results when smart technology is run by smart people – people with training, experience and (most importantly) judgement. There are pockets of this in mining, but not many yet. If you’re in this stage, the number one objective is to differentiate yourself.
Here’s another example – the development of the computer industry. In Stage 1, the tech was primitive, and people had trouble figuring out what to do with it. Once they learned, we hit Stage 2. This is where we had smart operators, but to do anything with a computer, you had to be a programmer. There was still no intelligence built into the machine. As we started to get more sophisticated hardware and software, we entered Stage 3. We started to embed more and more intelligence into the technology, to the point where you didn’t really have to know anything about computers to use one. The problem, though, is that this is dangerous. So we now we are at the point where we need to be moving more into Stage 4.
We’ve seen similar trajectories in medicine (keyhole surgery is a Stage 4 system), and the military – just compare the effectiveness of the Stage 3 smart bombs with the current high-skill military drones.
The really interesting part of this is the interaction between Stages 3 and 4. It is pretty useful to sink a fair bit of intelligence into your supporting technology. This frees up your mind to move on to other important topics. However, if you only let the intelligence sit in the technology, it leads to problems.
Here are some of the issues:
- Stage 3 systems tend to collapse. Why? Because when there isn’t any intelligence embedded in the people, there is no judgment. They tend to blow up as a result of unintended consequences. For example, automatic trading has led to a few stock market crashes. Automatic trading is a classic Stage 3 system – nearly all of the intelligence is sunk into the algorithms. On a less drastic level, this is also what you get when you give a highly sophisticated digital camera to someone with absolutely no idea how to take a good picture. This is because:
- Stage 3 systems lead to an over-reliance on tools. This is the whole issue that Terri Griffith is addressing in The Plugged-In Manager: Get in Tune with Your People, Technology, and Organization to Thrive. Her key point is that to succeed, you must understand how people, processes and tools interact. If you get all three of those working, then you have a Stage 4 system, which is much more stable and productive. Tools alone don’t solve your problems.
- If you’re in a Stage 3 system, you can change it by learning. If you find yourself operating in a Stage 3 system, you can learn your way out of it. This is the option that we face with computers, digital cameras, and even innovation systems. For discussions of why and how you should do this, check out In the Beginning… Was the Command Line by Neal Stephenson, and Program or Be Programmed by Douglas Rushkoff.
At the firm level, you can create value in all four stages – usually by figuring out how to move the system to the next stage. Again, the dangerous one is Stage 3, where it is tempting to treat your market like a black box. The assumption here is usually that if your users don’t understand the technology (because all of the intelligence sits in it, not the people), then they will continue to need you and your product.
This might work over the short term, but it’s not a good long-term strategy. People will always learn. To take advantage of this natural tendency, the best strategy is to figure out how to help them develop a Stage 4 system.
Smart people working smart technology is the winner.
Note: if you want to see an interesting theoretical economics perspective on this, check out this .pdf of a paper by my colleagues Jason Potts, Joe Clark and Kate Morrison.