Killing hedgehogs, fast and slow
I’ll admit - I always have trouble remembering the role of the fox or the hedgehog in this story, but famously in an essay, Isaiah Berlin drew attention to an old snippet: "a fox knows many things, but a hedgehog knows one big thing". I think I must have different impressions of foxes and hedgehogs, which leads to my need to check each time…
Berlin’s essay broadened the idea: ‘hedgehogs, who view the world through the lens of a single defining idea, and foxes, who draw on a wide variety of experiences and for whom the world cannot be boiled down to a single idea’.
I’ve had several debates about which is a better model for pharma, with people genuinely arguing for the hedgehog approach (perhaps influenced by Jim Collins’ book Good to Great, which comes down on that side… Of course, appealing though Good To Great is, it fails its basic test: many of the companies he singled out as Great in 2001 are either defunct, on their way there, or didn’t turn out so great in the following years).
The hedgehog view certainly works when communicating a story. In fact, it is the basis for positioning - framing a view of things through an active single lens. Its appeal is to those in the audience who find uncertainty awkward, which is most of us. It also works really well in investor presentations, where uncertainty is rarely taken as a positive. Confidence in your hedgehoggy view of why your platform is the one that will transform a disease is unlikely to be displaced in that kind of setting.
But it works less well when it comes to Development and Asymmetric Learning. Fixing your idea as the only lens through which to see things (‘it’s amyloid, stupid’ or ‘it’s about reducing calories, obviously’) changes mindset to confirmatory versus exploratory. To be a fox, you don’t have to know all of the other things, but you do have to believe that there might be other approaches of value.
Pharma’s linear Development path is a hedgehog strategy: ‘JAK-27 is the rational intervention in the CHY cascade, and therefore…’ It is a strategy in which all else is ignored, and confirmation that this one thing is right is explored, for as long as it does confirm. If it might not, or if it might disprove the idea, it is excluded. Into this model, the idea of ‘fast kill’ makes sense - if you have 30 hedgehogs in your portfolio, one of them might be right, so it makes sense to entertain their ideas for as long as they survive.
The development approaches at places like Amazon (‘invent and wander’) and Google would be more fox-like - pursuing pivots and serendipity. It starts out by saying ‘whatever the appeal of the idea, we may not be right, so let’s look not just to kill it, but to see where it could go, how it might mutate…’ .
As Philip Tetlock would say, of ‘superforecasters’:
Those people stood out not for their credentials or ideology but for their style of thinking. They rejected the idea that any single force determines an outcome. They used multiple information sources and analytical tools and combined competing explanations for a given phenomenon. Above all, they were allergic to certainty. They gave weight to multiple perspectives and weren’t afraid to change their opinions. They were curious, humble, self-critical, and less likely than most other people to believe in fate. And although they seldom used math to make their predictions, all were highly numerate. “I have yet to find a superforecaster who isn’t comfortable with numbers,” Tetlock writes.
We’ve all got used to Daniel Kahneman’s Thinking Fast and Slow, but in a great example to all of us, he hasn’t quite ditched System 1 and System 2 as a lens through which to view, but he is acknowledging that some of the things which seemed to fit either were fallible, or don’t.
There are a couple of principles in strategy: survivability, and adaptability. They say, in the presence of new information, or possible outcomes, does the strategy survive or could it be killed? And/ or could it pivot, or is it inflexible? Some companies become foxes only after their Phase II kills their worldview, and they scramble for a way to keep the molecule alive.
Asymmetric Learning itself could be seen as a hedgehog worldview - it is certainly a single lens through which to see pharmaceutical development. But it has survivability and adaptability built into its principles, as well as accepting that, as Safi Bahcall’s Loonshots shows, some single-shot ideas (like Avastin, or statins) do work despite their multiple false fails. It is not the only way to do this, but a more fox-like pharma development approach would be an advantage for any one pharma company, and therefore better for all of us.
