“If the machines are doing the routine work, then humans must do the non-routine work. To be an innovator, you need to talk about unsolved, important problems.” – University of Waterloo Professor, Larry Smith
Last night I joined two former colleagues, Don Bowman and Dan Deeth, in the University of Waterloo’s Velocity Start centre for What’s Your Problem?, a talk by the renowned Economics professor Larry Smith.
As always, I took some notes so you didn’t have to, and I’ve since reflected upon them. Enjoy!
OK, first up, a bit of housekeeping:
- Velocity Start, touted as “a place for anyone to learn about entrepreneurship”, is part of Velocity, which is “a leading entrepreneurship program at the University of Waterloo, and the largest free startup incubator in the world”
- Larry Smith is an econ prof at UW, and in that capacity is somewhat legendary on campus. Due to the popularity of his classes, which in no small part is because of his engaging speaking manner and overall lecture style, it’s estimated that he’s taught 10% of UW’s students (counting in all faculties) over the years. Lately, he’s gained wider acclaim for a couple of books and a TEDx talk (see the bottom of this post). Innovation is kind’ve his thing.
- The Problem Lab helps “University of Waterloo students, faculty and staff identify and understand important problems”
Alright, you’re all caught up.
As you know, I’m in-between things right now, so I figured this talk might get the old creative juices flowing.
In general, Larry (may I call him Larry?) had a few points…
Tumultuous Times Are Ahead
Larry recapped some of the amazing technological changes that the world has seen in the last 50 or so years, and stated his thesis that the pace of change is actually accelerating.
His primary point here is that the world of tomorrow will be substantially different from the world of today, with sub-points being that:
- People are underestimating the pace and degree of these changes
- Tumultuous times bring both danger and opportunity
- While specific changes are often unpredictable, there remains an unmistakable statistical trend of the automation of routine work
Automation, and the Rise of the Machines
Post-secondary education institutions are pretty adept at producing people who are well-trained in advanced, sophisticated, elaborate – but nevertheless routine – work. Why is the work routine? Well, let me ask you this question: did you learn your stuff from a textbook or some other manual that listed rules, formulae, concepts, etc.? Cool, well if you can learn from a manual, then so can a machine. Shit, right?
The definition of “routine work” changes over time, and – again – many people forget this reality: self-driving vehicles are coming for cabbies, truckers, transit drivers, and chauffeurs; robotic attendants are coming for health-care workers, machine learning is coming for analysts; AlphaGo’s already come for the Go players (really good documentary, btw); DJ-3000s are coming for radio DJs, etc.
Lucky for me, there’s no possible way marketing can be automated. [checks Internet, discovers that ‘marketing automation’ is a multi-billion dollar industry]
Well shit. Now what?
Basically, to stay relevant and employed, we humanfolk need to stay ahead of the machines.
Larry summarized his points by saying, “If the machines are doing the routine work, then humans must do the non-routine work. To be an innovator, you need to talk about unsolved, important problems.”
“If the machines are doing the routine work, then humans must do the non-routine work. To be an innovator, you need to talk about unsolved, important problems.”
I Got 99 Important Problems
Attempts at innovation frequently fail because either the innovation is not important enough to be widely adopted, or the problem is poorly understood.
The good news is that if you’re solving an important problem, then you’ll often find it easier to get funding, to attract talent, and to sell your solution.
The bad news is that important problems are typically poorly understood (so solutions miss the mark), which is kind’ve where the Problem Lab and Velocity’s Problem Pitch competitions come into play. Quoting from the lab’s home page:
Attempts at innovation frequently fail because either the innovation is not important enough to be widely adopted, or the problem is poorly understood. The Problem Lab helps you address both issues.
We ask questions about the problem, not the solution. Is it really important? Why is it really important? Do we understand the problem thoroughly? Have we identified the underlying problem, instead of a superficial one? How has the problem arisen? Why have others failed to solve this problem?
We should do all this before we ask ourselves how we might solve it.
Don’t Underestimate Yourself
In Larry’s view, shaped by his experience, people and the organizations to which they belong often overlook important aspects of problems – aspects that are staring them right in the face. He shared a few examples to support this assertion:
- Steve Jobs recognized that the aesthetics of computing matter, when most others didn’t
- Walt Disney recognized that children have long attention spans when they’re engaged and interested, but that the real problem was that they got bored with most child-focused offerings
- Similarly, JK Rowling recognized that kids would be happy to read longer and somewhat complex stories, provided those stories were compelling
Larry recommends testing and challenging well-held assumptions: if all the competitors in a market place take X for granted, and then you test and find that X isn’t really the case, then you can gain an innovative advantage.
Larry implored the audience not to conclude that we can’t compete against big companies, “because big companies are often stupid”.
Larry implored the audience not to conclude that we can’t compete against big companies, “because big companies are often stupid”. As a supporting anecdote, he explained how Sears had pioneered catalogue ordering by mail and by telephone, had built out a massive distribution architecture, and had tremendous brand trust, and then completely missed the mark on what was a very logical incremental evolutionary step: Internet commerce. Instead, likely fearful of cannibalizing their own business (modern discussions are now talking about creative destruction), they fell into obsolescence as companies like Amazon started from nothing.
This part reminded me a bit of Tim Ferriss, who’s pointed out that we often give our competitors far too much credit – much more, in fact, than they deserve. By doing so, we fool ourselves into thinking we can’t possibly solve a problem, if the competitor hasn’t been able to do so.
Circling back to his opening remarks about the pace of change, Larry was demonstrably insistent (as he always is) that one critical factor is that companies need to move quickly; to do that, companies need the best and brightest people, because these problems and their associated solutions are not yet routine.
So there you have it:
- The pace of change is accelerating, which is a mixed bag
- The machines are coming for routine work, and what’s ‘routine’ changes every day
- We need to tackle the non-routine, by finding and understanding important problems
- Many important problems are recognized, but poorly understood, often due to false assumptions
- Don’t underestimate ourselves, and don’t give existing companies too much credit
- Be quick about it!
If you’ve got any questions, or if you were there and want to add comments or refutations, then by all means please do so in the comments section!
[…] I need to write reports” ever since. I finally got around to writing the report because of the talk I attended earlier in the week – I realized there were many, many similarities between what Professor Smith was telling the […]
I think there has to be more to the future of work than simply trying to stay ahead of the robots. The distinction between routine and non-routine work is important, and one of the most important human skills today is to distill the routine and codify it so that machines can do it.
But there are things that humans can do that machines can never do, things that have nothing to do with intellectual capacity, and which don’t get us into the fraught question of whether imagination and creativity are simply manifestations of mechanical intelligence or whether they are something else that may not be programmable.
First and most important, machines cannot care. Only humans (and maybe dogs) can care. Robots may be able to do all the material things that a human caregiver can do, but they cannot actually care, and that makes all the difference. One could die of loneliness being immaculately cared for by a robot. (A million versions of the Pinocchio story explore this territory and ask if a machine can, in fact, care, and if we would, in fact, take comfort in its caring as we take comfort in human caring. Personally, I doubt that the uncanny valley is bridgeable. I think a sympathetic robot will always seem goulish, at least to many.)
Secondly, robots cannot have experiences. They can, of course, gather data, but the don’t receive the fullness of experience the way humans do, nor are they marked by it the way human are. This limits the ability of the robot to interact meaningfully with human being, to laugh with them, to weep with them, to share stories with them. Whether it is possible to build a machine intelligence with the same physical and emotional vulnerability, the same love of companionship, the same consciousness of limitations and fear of death that a human being has seems to be one we are a long way from answering, but what would be the point, at least from an economic point of view? Robots are economically useful precisely because they don’t get bored, make mistakes, get tired, or fear death. They rival us not by duplicating us but by being strong were we are weak. But by this same token, they cannot have human experience, and therefore cannot participate in human life and community as full members.
Rather that running to keep ahead of the robots, therefore, we should probably be thanking them for freeing us up to occupy the caring professions. We have enormous deficits in the caring professions today. Loneliness is epidemic. Care for the aging will become an increasing problem for decades to come. Let the robots free us up to tend our gardens, to beautify our neighbourhoods, to play with our children and our grandchildren, and to care for our parents, our grandparents, and our neighbours.
These activities, these professions, have not been attractive in recent years because they have not been as well compensated as professions that make and distribute material goods. But when the production of material things becomes the domain or robots, there will be resource to compensate the caring professions, and an increasing demand for them as lose the camaraderie of the workplace.
Robots may eclipse our strength, but they can never share our weakness. The may eclipse our intellect but they will never share our empathy. Let them work their side of the uncanny valley and let us work ours.
To be fair, Larry did strongly imply – if not outright state – that the traditional ‘arts’ are vitally important to innovation.
Also, while you didn’t explicitly state it in your thoughtful reply, there’s certainly an element of Moravec’s Paradox in there. Simplifying a bit: there’s a whole host of tasks that are trivial for a human and hard for a robot/machine, and vice versa.
Now, as a slight counter-argument, citing the uncanny valley presumes that future generations care what their robotic caregivers (overlords?) look like; however, there’s a whole generation growing up now who are perfectly content to interact with their phones and rounded household assistant speakerbox things – who said the assistant has to look human?
I also worry a bit about who’s going to pay for the presumably relatively more-expensive human caregivers. As automation puts many blue- and white-collar folks out of work, disposable income will likely fall. The risk, then, is that human caregivers will be a luxury affordable only by the elites.
Thanks again for the thoughtful comment!