He created several subsequent tests and wrote a book about it [1]. I make a version of a few of the tests for iPhone if you're so inclined [2].
0: https://www.sciencedirect.com/science/article/abs/pii/B97800... 1: https://amzn.to/3PVOne9 2: https://apps.apple.com/us/app/test-your-chess/id362448420
This reminds me of the curse of working with really good senior engineers. They already know the answers, they've already solved the puzzles. It can be very easy to just defer to them all the time.
If you are a senior engineer who really understands a system, you need to be conscious of this effect if you ever want someone else to start learning your system.
There's also the pressure from above to fix things quickly, meaning some people don't have time to really explore and learn and need to be given answers...
- during long game, chess grand masters have physiology comparable to marathon runner, while he runs. Deep thinking for several hours, takes huge load on body. All the logic and critical thinking, is not going to save you, if you are not fit, and your brain does not work correctly.
- real life is not about solving puzzles. Real life is a rigged game where rules are not enforced. Instead of finding problems to solve, you need to find oportunities (and loopholes) and exploit them!
- game is rigged, and oportunities close fast. What worked a couple of years ago, probably does not work anymore.
That is obviously not true https://old.reddit.com/r/chess/comments/s0tqcd/chess_grandma...
Many top chess players have considered themselves athletes, and if you've ever tried to calculate under pressure at a board for a few hours, I'm sure you'll agree it is an exhausting activity. Fischer engaged in athletic training when preparing for tournaments, for example, to aid in maintaining mental focus (he wasn't unique).
It would be silly to say that 2400 Elo indicates you can run a four-minute mile, or that calculating 6-ply in a closed position burns the same calories as running a block. If the claim is 2400 Elo tend to have similar vascular flow in the brain to people who engage in aerobic exercise or something of that nature -- maybe?
But I still maintain my claim, health really matters for proper deep thinking.
He says, before laying out the outline of the puzzle and giving suggestions on how to go about solving it
Everything can be about solving puzzles if you let it. Given enough time and patience you can understand anything - the only interesting question is, how to you decide what to focus on?
Navigating life is absolutely an exercise in puzzle solving, at every step you know where you’re at, you know where you want to be, and you know what resources are available to you - given all that, how do you plan your next step? If your first solution doesn’t work, you do a retro, learn your lessons, and move on to your second solution, and your third. It’s all engineering.
> It’s hard in real life, too: vanishingly few people are meta-rational enough to try really hard to falsify their own ideas. Your brain really wants to find reasons to support what you believe.
I don't think he goes with "meta" deep enough. It is great for engineering problem solving mindset But it is also a good way to end up like underpaid post doc, who needs second job just to pay rent.
And this type of advices are usually coming from someone who "made it", has its own house and is practically retired. Very impractical and harmful (to some extend) for young minds.
Practical implementation for young person is not "falsifying" and trying again again. But coming with solutions that takes minimal time, is good enough and comparable to coworkers who work on the same salary. Time you save can be invested into education, family, hustle and so on.
"Real life puzzles" are too open-ended and have too many levels to really be called puzzles. A puzzle has a closed set of rules that usually gives you only one level on which to solve the problem. Many interview questions could be described as puzzles. A "real life" programming task has a bunch of different levels: what's the real problem the customer wants solved, is this the problem this customer wants solved first, do we have a bigger customer with a bigger problem you should be working on instead, can the problem be solved without programming, should the problem be solved in a different system, are there other people on the team who solve problems like this in their sleep and they'll give you the answer in five minutes if you describe it on Slack? If it does seem like you need to solve the problem, what's your level of confidence that with investment of X time you can solve the problem, for different values of X, and given this information, does it still make sense to try to solve it?
What makes puzzles relaxing and reassuring is knowing that there is a solution, and that you know all the rules. Also, you know that you'll recognize the solution when you get it. Real life rarely gives you that reassurance. With a real-life problem, you don't know if there's a solution, and even when you have one, you can't know that there wasn't another solution that would have been much better, because of possibilities you failed to consider. The only way to turn a real-life problem into a "puzzle" is to strip away the open-ended real-world context and present a subset of it that can be described in a closed form.
Life is turtles all the way down and drawbridges all the way up. Anyone seeking opportunities is looking for the openings between those moats and drawbridges.
Starting family today is very difficult, there is no easy and direct route.
I thought maybe I could find some primary sources, but the [1] notation is just footnotes.
You’re right that GMs are much faster ideating, I point this out later in the essay. But they also spend longer on falsification, even in absolute terms.
Are there legitimately multiple good examples of "Criticized on HN pre-launch, yet became surprisingly successful"? I'm curious if the lesson to learn from Dropbox & iPod is more of "believe in a product, despite the criticism" or "sometimes, even accurate predictions are wrong"
For developers or managers on HN, one outcome would be that it's best to start one's career in testing, or to respect the resumes of those who started in QA. If/since there are hundreds of ways things can break, it's a harder problem to show how it will, or prove it won't; and building a mental library of fault models helps in vetting designs and implementations.
Or, we could teach fault models directly, instead of accumulating by experience. See e.g., Robert Binder, "Testing Object-oriented Systems" (and ignore the model-driven-development gloss from later editors).
But the most important note is the aside: the author avoids chess as addictive. Should we ask ourselves: how can this be? Should that change how I think about my own work?
Everything that shows up to the help desk is broken. QA people need to have a skill for breaking things or at least an awareness of how things break. They will learn this at the help desk.
Otherwise: I completely agree.
It's not much deeper either. There's having a skeptical yet positive mindset which is somewhere in-between. I once made a complex toolkit of runtime query optimizations that was a hodge-podge of kludgy things, which worked out well in practice. Someone asked me how I ever came up with the whole thing. I said I just started out and made one new improvement after another. They said that it was so fraught with obstacles (listing them off) that they'd be daunted to even start. I said I wished I'd talked to you sooner, you just gave me a roadmap that I didn't have while I was figuring out the pieces. Note that he was a high-level chess player (unlike no-rank me), but perhaps not as much of an optimist or risk-taker.
If I had to say what a good founder mentality is, I'd say that they have experience navigating uncharted territories and finding success, whether on small or large scales. That kind of practice makes them good at sizing up risks and rewards. Related to this might be of a kinesthetic learner type--learn by interacting.
Thank you for posting this.
I play chess (poorly for the time spent on it) and I'm also a reasonably successful founder of a couple software companies. I find my struggle with chess is that I want to act intuitively, something that has served me well all my life in other avenues. But the board doesn't lie and if you don't think thoroughly you will get punished.
I have the capacity for it, I can think thoroughly in puzzles and perform much better there than my on board play but I just struggle so much with the discipline during regular games to falsify my moves. So much so that I've mostly given up on trying to improve despite really loving the game, it just grates on me. I know I could be better but I lack the discipline and I guess I just don't want to exercise that discipline in a game.
Anyways, great article.
I believe strong players do act intuitively when playing chess (especially fast chess), it's just that they've developed their intuition through lots of practice and thorough thinking in the past. For some reason our intuition about life seems to be more developed, or perhaps the game of life is incredibly complex and most people are roughly at the same skill level.
Same. Truth be told, I don't want to have to explicitly think at all. I want my brain to recognize the situation and steer me towards victory through entirely unconscious processes.
Edit: Ah, okay, it's probably in the book being discussed where he says they recorded thought process while playing ( https://www.amazon.com/Think-Like-Super-GM-Michael-Adams/dp/... ).
Except it doesn't work, I needed to falsify the falsification of the falsification 4 move down the line to see why :)
That’s the point of the article. It contrasts the thinking styles of ‘founder-types’ and ‘scientist-types’.
As a (in the terms of the article) ‘scientist-type’ who regularly gets lost in the weeds of the details, I found it a pretty interesting commentary.
Startups are a beauty contest where each player focuses on maximizing the things about them that appeal most to a panel of judges (customers). Similar to the scramble competition that Benenson cites here, rather than an arm-wrestling contest.
PS: I'm an early stage founder, who has finally some traction with my current B2B data infra SaaS. I've had a failed company in the past which had 4 major pivots, where we decided to return most of the funding to learn few things again.
Nearly all engineering is balancing different concerns - durability vs price, weight vs features, and so it goes.
Besides engineering, you also have game theory, political science, etc.
