Bill Buxton

S&T policy and the Long Nose

By Bill Buxton

For a company to go from zero to $1 billion in revenue takes 20 years — what I call the Long Nose. Most people are not aware of how long this nose is. The first part where an invention happens is basic research, in the middle is refinement and augmentation which is applied research and the third is the productization. If you believe in R&D you first have to understand the nature of this Long Nose, what the timeframes are and set expectations accordingly. Who should do each of these phases, where should my investment be and how do I triage my portfolio?

Any invention that's going to become a $1-billion industry in the next 10 years is already 10 years old. It is not about alchemy, it's about prospecting and having the technologies. You need the tools, the people and the processes to prospect. You have to dig down with your pick below the radar, find the stuff, know that it's valuable, know how to mine it and how to refine it.

If you want hit that $1 billion, you need to be a goldsmith that turns it into something that's worth more than its weight in gold because of the value-added. Or else you remain a hewer of wood, drawer of water. What part do you want to play at? Investment strategies and how you perform in each one depends on whether you're a prospector, a miner, a refiner or a goldsmith. I would argue that if you want to have a balanced economy, you want to participate in all of those that you possibly can.

So let's talk about our federal government. If we understand that this is at least a 20-year proposition, our biggest challenge is reconciling the fact that our governments circulate at four-year intervals and our research and investments circulate in 20-year cycles. We have to get the politics out of policy and hold the politicians and policy makers to task and say ‘Damn it, this is your children's future as much as mine … Stop playing games with the country's future'.

With the policies we have, governments can't produce anything in four years. Garbage in, garbage out. It's true of computers and it's also true in terms of policy. Any definition has to distinguish between advanced research, applied research and production. And if you're not at that granularity — which is still too coarse — your numbers don't make sense. You're mixing apples and bananas.

If you do a study of American universities, you will find, with very few exceptions, none of their technology transfer programs actually brought a net gain to the university. None of them brought about a re-creation in their jurisdiction of Silicon Valley, which is what they were all trying to do.

Silicon Valley and Silicon Fen at the Univ of Cambridge did not have such a policy. Neither does the University of Waterloo. With all the technologies going around Waterloo, they handle intellectual property very differently from the University of Toronto (U of T) and most of the universities in Canada.

the Lazaridis example

Then there's Mike Lazaridis. Don't you find it curious that most of the federal policies are biased towards industry relevant research and perhaps Canada's greatest entrepreneur is putting his money into the most flaky, far-out curiosity driven research — the Perimeter Institute. Why? Who knows more about investments in R&D and innovation — the federal and provincial civil servants setting policy or Mike? Might we learn something from this? If there's another Nobel Prize coming from Canada, it's probably coming from the Perimeter Institute.

We have bought into the notion that the days of the big corporate research labs and the days when the private sector can do basic research are gone. So we need to transfer this research to the university which means taking attention away from basic research in universities and having them do applied research. But in doing this, you destroy the community of scholars which makes it impossible for scholars to try and work with each other.

This is why I left academia and I'm now at Microsoft Research. Here's their policy. If you want to publish your research, you don't have to tell your manager, legal, anybody. If you don't have the maturity to make that judgment on your own, you shouldn't be there. I have far more academic freedom to talk about my work working for the private sector than I do at the university.

The reason this investment in strategic industry-relevant research is a bad investment is because the large wins that cause great companies to happen and are sustainable are almost always unintended consequences of people doing curiosity-driven research. And if you cannot have the freedom to play and go off, you will not have the opportunities for these things to happen.

Canada has a couple of poster children in the digital technologies sector. One is mobile technology in the Canada space arm and the other is in computer graphics and digital media. And in the latter, the most important catalyst making it happen was work done at the National Research Council (in the 1970s). From them I learned the basics and got the catalyst as did a whole bunch of other people. There are about a dozen Academy Awards by Canadians that are a direct consequence of this work plus a multi-billion-dollar industry. And nobody took (the NRC researchers) seriously. And these projects were shut down to do things which were ‘real' and ‘serious'.

You need people like this to shape people and you need an educational system and culture that informs these values. Serious problems require serious people to play well and are deep in what they do. We don't want jacks of all trades and masters of nothing. For the pillars of any successful innovation, you need business, technology and design. Each is an essential pillar none on their own is sufficient. If you do that and bring them together, you have common ground and you have the depth for the wide base to bring these things together.

role of universities

Academic industry-relevant research is neither — it's not academic and it's not research. It's a diversion that hurts academia and prevents industry from doing what it needs to do. The shift in basic to applied research results in a reduction in productivity not an increase. In my opinion, demonstrating industry relevance in your academic research is a reason to stop funding at the academy not to quadruple it. And if the private sector doesn't pick it up, it's their own damn fault. We deserve to fail.

I shut down the best research group on telepresence at the U of T when we went for the review of the second term of funding and we were told, ‘I need you to tell me how many jobs you've created, what the companies are going to be and what those inventions are going to be'. The 20-year horizon is already too short-sighted and science and research cannot thrive in a cultural vacuum. If you do that, if you starve the arts, social sciences and focus just on the high tech stuff you are going to lose. Risk aversion is the most dangerous course of action. The final question is, can you change your culture and the answer is absolutely. If we want to change our culture we need to collectively set up with a strategy and borrow the tactics used by these other examples to help make a shift in our values.

I may be completely wrong in the direction I'm pointing but I am not wrong in the need for a shift — 30 years is too long to be making these mistakes. Look at the data, the data say it's not working. I believe because I'm a ‘sceptimist' that we can get it right. One thing we do know is, we are among the brightest jurisdictions in the world. So let's use that for our benefit.

Bill Buxton is a a pioneer in the field of human–computer interaction, principal researcher at Microsoft Research, an adjunct professor at the Univ of Toronto and formerly chief scientist at Alias Wavefront and SGI. This column is an abridged version of a keynote presentation Buxton made at the recent RE$EARCH MONEY conference in Ottawa.