The Illusion of Certainty
app.spectator.co.uk
app.spectator.co.uk
The author sees two obstacles. One is what he calls "the boredom threshold:" The creative individuals who could perform those feats quickly get bored with the mind-numbing minutia of bureaucratic processes at larger organizations and leave.
The second obstacle is "the illusion of certainty:" Conventional operational improvements, cost-cutting efforts, legal reorganizations, etc. can easily be 'proven' to work in advance, but more interesting lines of inquiry come with career-threatening unknowability. Many bureaucrats and businesspeople, the OP argues, are attracted to the illusion of certainty.
Quoting from the OP: "If you engage engineers, you don’t know what you are going to get. You may be unlucky and get nothing. Or their solution may be so outlandish that it is hard to compare with other competing solutions. On average, though, what you get will be more valuable than the gains produced by some tedious restructuring enshrined in a fat PowerPoint deck. But in business, let alone in government, it is only in crises that people find a budget for probabilistic interventions of this kind."
The essay concludes that "one reason why the world is in a mess is because, for a long time, the ratio between 'explore' and 'exploit' has been badly out of whack. Entities like procurement have been allowed to claim full credit for money-grabbing cost-savings without commensurate responsibility for delayed or hidden costs."
The essay is about much more than that one anecdote.
I'm not sure I buy this premise: Menlo Park, Bell Labs, Xerox PARC, IBM Research, major universities, the US military/government. Giant organizations like these provided enormous innovation and discovery in the last century. Maybe they don't anymore, but the question of why isn't the same as proclaiming "larger organizations [are] so bad at performing feats of brilliance most of the time".
Never seen that idea summarized before, but it resonates with a lot of thing of how the world works. The system wants some subset of reality and elects it as the reality because the system is stabler this way for a period.
I also think that past 2 decades went overboard with that.
As far as I can see they have done none of these things and are only helping in providing manufacturing capability [1]. The ventilator they are helping to manufacture is an iteration on a previous design made by a medical specialist Penlon [2].
Not that there is anything bad with that strategy: experts in a given field are the experts so best to use the medical expertise of one company and the rapid batch production capabilities of companies like Formula 1 teams. Racing teams often have to turn components around for an update at the next grand prix e.g. 1 - 2 weeks. This means having processes and suppliers that can operate at that fast pace.
[1] https://www.inceptivemind.com/penlon-prima-eso2-newly-adapte... [2] https://www.penlon.com/
"The [University of Southampton & NIHR] researchers worked alongside F1 team McLaren and industry experts including INDO and Kemp Sails to come up with the design, which is now available as open-source to be manufactured around the world.
https://www.itv.com/news/2020-04-23/hospital-first-to-widely...
We'd be better off if they just admitted the failure so we can move on to the next step, rather than be distracted by utter b.s.
The reality is even if some science-project ventilator was made, no hospital would approve its use. But we knew that in advance, before the b.s. even started.
And we've known since Jan. that ventilators kill their patiences 66% to 90% or more often.
It seems consistent with what you're saying. There's some broad consortium of companies, but there are already designs approved by regulators and they're more or less providing manufacturing capacity to make more of them as fast as possible.
I see the same in software. We sometimes have interns whip up some system that on the surface looks good and innovative and makes the old timers look bad and unimaginative. But once you try to deploy them they usually fall apart because they ignored some inconvenient real world details.
<QUOTE> ... 22. And then I’ve got development and organizational confusion from say-something syndrome.
And here my favorite thing is the bee, a honeybee. And a honeybee goes out and finds the nectar and he comes back, he does a dance that communicates to the other bees where the nectar is, and they go out and get it. Well some scientist who is clever, like B.F. Skinner, decided to do an experiment. He put the nectar straight up. Way up. Well, in a natural setting, there is no nectar where they’re all straight up, and the poor honeybee doesn’t have a genetic program that is adequate to handle what he now has to communicate. And you’d think the honeybee would come back to the hive and slink into a corner, but he doesn’t. He comes into the hive and does this incoherent dance, and all my life I’ve been dealing with the human equivalent of that honeybee. [Laughter] And it’s a very important part of human organization so the noise and the reciprocation and so forth of all these people who have what I call say-something syndrome don’t really affect the decisions. <END QUOTE>
So this "Say-Something-Syndrome" is a negative input that introduces development and organizational confusion in to an otherwise well tested method. It doesn't scale. It is to be ignored.
This troubles me as an absolute position. Yes there is danger funneling good resources down a rabbit hole to nowhere and yur core strategies should not flap around like a faulty nic card chasing shiny new things. But, you cannot close a system off completely or it will get stuck. Contuious improvement, iteration, hot wash lessons learned, red teams, squad checks, all these are dependent on effective positive & negative feedback processing. Not writing off all dissenting opinion as noise.
It's a balance. It's form of art. Dynamic systems are messy in real time both for managers and engineers and not much about that will change. Best we can do is learn what we can from some old bee and monkey [2] stories.
[1] https://buffettmungerwisdom.wordpress.com/2013/01/14/full-te...
You’ve criticized the article but ended up reaching essentially the same conclusion. Having interns spin up a new feature which can later be tidied up and put into production is a great use of time. Interns are cheap, less busy, see things that more jaded people may not, and are liable to be less risk averse too. If you take what the interns build and put it directly into production, that’s on you, but the interns may be able to get you 80% of the way there in half the time.
From my experience they don't. Either it gets thrown away or the cleanup requires as much or more even more work than if experienced people had done it from scratch.
Edit: I am not saying there is no value in work from newcomers but I think it's often overrated. They usually do the easy part but finishing it takes much more work than people think.
(Game, chat app, anything.)
Well, large organizations have overhead for communication, management, etc (see v). But also there is a statistical effect - if we get many, many shots - sometimes we get lucky (in most cases - not; think of a typical hackathon project - usually it never gets properly started). So, in short, we see a huge survival bias.
...and it takes time to polish rough edges. For example, game Superhot started as a hackathon demo (with a single scene). It took Kickstarted crowdfunding, and a year of work to turn in into a fully-fledged commercially-viable game. (Personally, I think it is one of the most creative game ideas in the last years, especially for VR. I am happy that I crowdfunded it. :))
Otherwise you over-adapt (overfitting) and lose the reservoir of variability required to cope with extraordinary circumstances.
Gregory Bateson describes it in one of his essays, I haven't had enough coffee yet but if you're interested bug me and I'll go look up which one. (It's in "Steps to an Ecology of Mind" or "Mind and Nature: A Necessary Unity" IIRC.)
We do exactly the same. We exploit in the times of plenty and explore when the going get tough. The same evolutionary forces act on us as a society as on the bees, so it is unfair to say that bees know better than us how to get that ratio right.
https://en.wikipedia.org/wiki/Multi-armed_bandit#Optimal_sol...
I think the article stops short of saying how we would know enough resources are being put into innovation. And for that, I have no idea either.
Mostly they seem to be acquiring smaller companies that have come up with interesting ideas.
Apple, especially, appears to be churning out annual updates on a product line that is not changing at all (I'm not blaming them, they're making more money than god, but still).
And pivoting? I guess Alphabet's "it didn't go viral in the first two years so we're dropping it" strategy could be thought of as pivoting, but not really.
I'd say a company as large as Apple would need to innovate hundreds or thousands of times per year, to be comparable to a small startup, in innovativeness per person.
I would dispute that paragraph. The popularity of "Primani" is proof that there are plenty of people like to be able to buy a larger range of cheap stuff. The number of times I've heard female friends, on being complimented on a garment, say "and it was only a fiver", or "10 quid at Primark".
I would say there's plenty of both types in this world. Those who prefer quality over quantity, and those who don't.
There is a relevant theory in economics called "the market for lemons" that speaks to what you are saying though. The idea is this: if someone is trying to buy a washing machine as cheaply as possible, they would rationally pick an $1000 machine that lasts 15 years over a $500 machine that lasts 5. The problem, however, is that the consumer often can't tell if the expensive machine is really better quality, or just going for a higher markup. With that uncertainty, it is safer to buy the cheap machine. Manufacturers know this, so they compete on price and features more than longevity. Without a time machine, consumers can't see if the extra money they spend is worth it, so cheap crap is the equilibrium for the market.
If you're working for a large, bureaucratic organization, producing a radically different solution can be a career-limiting move, even if it's a success and the problem is quite important. (Or perhaps especially if it's quite important.)
Why / how can that happen?
In the real world, getting things done is not your job. Your job is to make your boss look good to his boss. In reasonably functional organizations, the former is aligned to the latter. But they are hardly the rule.
Two things:
1. From my own personal experience, I think the observations here are _accurate_
2. I really appreciate that is article offers an implicit good-side to tough time. That is, success in tough times is made by ditching safe behavior