Philosophy of Innovation
Thoughts from Finland, February 2018
Facilitation Theory
This semester we have a course called “Usability Evaluation”, there’s something interesting about the course structure: it has an exam after the second week, and the main part of the course starts afterwards.
Here the exam is used as an instrument to motivate students to cover the reading materials, so the course can start with a common vocabulary. Compare this with the role of exam in the framework of “attainment theory”, where the goal is to install knowledge, and exam is deployed at the end of the course as quality control.
Props and Artefacts
Another example from the course is the “learning diary”, a meta-cognitive device to facilitate reflective introspection — by thinking about the experience and being mindful and deliberate about learning, it helps the learning set in.
In Oulasvirta’s Computational UI course where we learn about optimisation and basic machine learning techniques, for each topic there are several tasks (increasing in depth and difficulty). These are optional homework exercises, each intended to help you learn more and go further on the subject.
There’re many examples of how tools facilitate behaviour that wouldn’t easily occur otherwise, the ice-breaker game, or coffee table books as conversation starters, “36 questions of closeness” from A. Aron, or in behavioural economics, the impact of “visualising an older self” on health behaviour or savings, UK tax office’s “norm framing” by sending a gentle reminder citing that “most people pay on time” which has real bottom-line effects, or EF the accelerator programme with their team forming and re-forming methodologies.
Rituals and Exercises
At the Dash hackathon, when we were first assigned to a team, we spontaneously carried out an exercise of writing down our expectations on post-its, then comparing and aggregating them. We arrived at some consensus, e.g. “learn about the design process”, “meeting new people”, “gain new perspectives”, “have fun”, and most notably, “stay fresh”, not to burn out. The process helped with basic tone-setting for becoming a team.
In various startup circles, there have been several projects of making the paperwork less of a barrier. The efforts span from making the documents open-source (StarterDocs, SeriesSeed) to converting it to Plain English — my further imagination was having point-by-point Khan Academy style walk-throughs on each topic of the shareholder agreement, as a course of exercises to work through for partners considering a venture together — this kind of guided critical thinking helps build common understanding, beyond just having a dense document as insurance in case disputes arise.
Non-Obvious Improvements
Why does design matter? It makes possible what is otherwise astronomically rare. Design sets a system of incentives, e.g. market exchange, separation of powers. Design functions as prosthetics, e.g. eye-wear enabling clear vision.
How do you come up with innovative designs? You inevitably face some boxes of when it comes to design — it’s hard to break out of the box and arrive at truly innovative designs with disruptive results. If you ask students how to improve the course structure, they may say “make the tasks easier” or “make the exam multiple-choice” — it would not be as revolutionary as moving the exam to the beginning, unless they’ve seen this example elsewhere. How do you arrive at non-obvious designs — that’s a good question.
The Chasm of Insight
What can “User-Centered Design” solve? What is the role of “getting out of the building”? User interviews may be good for discovering glaring low-hanging fruit, but could not lead to optimisation techniques such as the Gale-Shapley algorithm. Deep tech solutions often require a certain level of domain-specific sophistication. In entrepreneurship education and even design courses, students are told to go out and chase down a problem — this approach is dogged by “user myopia”: if you asked people in Soviet times what the problem with the social system was, they would say “Uh, the queue in front of me is longer than yesterday”.
Aalto’s venture education with its problem-finding model and emphasis on pitching is more or less replicating bad algorithms already used elsewhere. For example, the Founder Institute follows a similar algorithm, and what have they got? Very few success cases, and graduate end up narrowing down the business to something like Holly Cardew’s picture cropping service for e-commerce. It’s unsurprising, it’s a likely outcome when you rush to create a venture “from the shallow end”, without an edge or deep tech, and follow an algorithm of run-away cornering towards local optima, iterating your pitch-oriented idea into a dead end.
Now it’s also unsurprising how students go to a “Future of Mobility” and “Smart City” summer school working on these grand business concepts, only to end up developing an idea that’s basically a Google Maps clone just not as good. Granted, unicorn creation is not simulated annealing. We just have to notice that the molehills people end up at are glaringly and absurdly bad.
Computational Perspectives and Techniques
There may be a lot to explore here, from the perspective of computational problem solving, e.g. finding co-founders as a search problem, reading profiles can be informed by information processing, and so on…
Another example, if you listen to Sam Harris’ podcast, say, an interview with a guest—often intelligent people discussing non-trivial topics — there are many sections, or points of interest where you can branch off and have a discussion (similar to how you can insert comments at different locations along a sound track on SoundCloud). Then, there’s an interesting technical problem to solve, “How do you surface good ideas?”
Imagine two hosts are having a conversation in a lecture hall, if anyone could just branch off a conversation or start interjecting, you might have 200-people-versus-200-people chaos. If you have Q&A, it offers a channel, but it’s constrained and still faces a filter and quality problem. But how do you design a system, where you can insert anchors for branching comments and discussions, and allow quality content that is of interest to you to surface, so that you can further engage with it?
Machine learning may help a lot. In the aisles of a library there maybe some useful information, useful to you, that is, buried somewhere in an ocean of camouflage — if only it can be discovered and brought to your attention, as you need it, and when it makes sense to you. It’s a discovery problem.
Bad designs create YouTube, with run-away Godwin’s law, where the first comment says “Hey man, that’s a really nice song!” then the 200th comment jumps to “But communism killed a lot of people!”. The descent into chaos seems inevitable, even though YouTube has a half-ass reputation system with accounts and voting. Then there’s a separate phenomenon of the quality of its unadjusted local feed, e.g. idiots bathing in chili sauce.
Reputation system and quality of feed must be something Quora thinks about a lot. “How do you bring up high quality answers” in an ocean of inane noise? And it’s probably constantly experimenting with “credentials” and whatnot… Even then, Quora descends into a stream of “Do I look fat in this selfie?”
Can design bring out the better or worse versions of ourselves? For example, the dumb bots in simple behavioural interventions? Really, we could use more of this kind of exploration with computational techniques, rather than just urging everyone to have a go at a solution from the shallow end. What’s being dished out at Aalto now resembles a cargo cult, fetishisation of pitching with clownish coaches, derivative rituals and frenzies regurgitated by retrenched ex-Nokia execs now acting as startup shamans. There’s enough materials to write a feature-length Finnish knockoff of Silicon Valley. The BMC even permeates through design classes as a recalcitrant haunting spectre. Meanwhile, hard-working people in surrounding buildings are developing satellites and robots — we are here reinventing calculus.
What Kind of Education
Speaking of which, two students just walked by with a miniature satellite in hand, very possibly to be actually launched into actual Space with an actual rocket at some later time — this was totally unimaginable earlier. “The Jeff Bezos of 1918 was hawking bananas at a train station”— we live in tremendous times, and we should really pay attention to tools available to us at this time and fitting to this time.
This is the area I’d like to investigate more, with an interest in the “philosophy of innovation”, leveraging design to achieve social goals, grounded in love for reasoning and understanding the larger interconnected picture, informed by the vast wealth of scientific knowledge, and enabled by engineering techniques to model and explore novel solutions. This combination of meta-skills is what I believe to be an edge humans have over machines.
