I enjoyed the article, but I think the author learned the wrong lesson from his friend's success story. Csongor says:
> Then learning the answer revealed that it's actually something that can be fixed --- as is always the case with these things if you think about them enough.
The author focuses on the consequent, thinking long and hard about something, but misses the antecedent, acquiring a deep understanding of a problem area.
I'd say the antecedent is as important as the consequent, if not more important. If you went to grad school you likely knew tons of brilliant people who worked long and hard on important and difficult problems and got nowhere. What they were missing was not some secret effortsauce, but that they picked a problem on which they didn't have an opening or insight into, so all their hard work was them spinning their wheels while standing still.
One thing I tell my students now is that they should always pick problems where they have something everybody else working in the same area doesn't. Like if you want to do formal verification but your background is years spent in industry as a kernel developer, you should avoid the temptation to chase the latest fad (e.g., adversarial machine learning or Spectre/Meltdown), and instead pick a poorly understood and validated module in the kernel and try to find out what you can do to make that module more secure.
So true. This is also the 'Dilbert Theory of Value Creation' that I have used to great profit. Scott Adams likes to say, he is not the funniest guy around but funnier than most people. At best, he'd be a mediocre comedian. He can also draw better than most but hardly a Picasso. And yet, by combining both skills he has something "rare and valuable" that has led to enormous financial success.
He also adds that it easier to do this (combining two or more skills) than it is to be the very best at one thing e.g. an NBA player--it is actually easier, statistically, to become a billionaire than it is to become an NBA player.
There are more than 2,000 billionaires worldwide but only about 480 NBA players (32 teams with roughly 15 people on the roster). And if you think about it, becoming a billionaire usually involves multiple skills (good at programming + business, bill gates. Good at research/analysis + cultivating a rare temperament that allows you to act on the insights produced by the research even when everyone around you is losing their shirt, Warren Buffett etc, etc).
Being an NBA player requires one skill at a minimum. Being tall!
> There are more than 2,000 billionaires worldwide but only about 480 NBA players (32 teams with roughly 15 people on the roster). And if you think about it, becoming a billionaire usually involves multiple skills (good at programming + business, bill gates. Good at research/analysis + cultivating a rare temperament that allows you to act on the insights produced by the research even when everyone around you is losing their shirt, Warren Buffett etc, etc).
That's not really true. Once you are billionaire you can stay a billionaire for decades. Meanwhile most NBA players don't have long careers, many stay in the league only for a few years. I'd expect that over 10-20 years there have been more NBA players than billionaires.
You have a point but there is also quite a bit of churn in the billionaire ranks. So if you include all that have been billionaires in the last 10 - 20 years you'll have many more than just 2,000. And for practical purposes, having $500 million vs. $1 billion is really the same league of wealth. So if we make the criterion $500 million my point is absolutely true! 5,000 at any given time vs 480.
The path to billionare is unique and can't be copied and involves luck and usually some business skills and almost always communication skills or skills to get others to do work for you.
Being an nba player involves luck and hard work and a high spacial iq.
Anyone could become a billiarie with luck (powerball lottery in one extreme). Only 100,000 people in the world would even have a chance to compete to be an nba player.
You do realize there are 6 foot NBA players right. Hardly 100000 in the world. So if you're gonna go to the power ball extreme on one end then you must extend the analogy just as far on the other side of the argument.
Most billionaires were born in top 0.01 percentile of their respective population pool. And had to be lucky and smart amongst other traits.
A friend recounted a conversation where he was talking to one of his friends, saying he couldn't understand how these people could be so dumb. (I don't recall what "these" refers to...) His friend said, "I don't think it's that they're dumb. They just aren't curious." That was an enlightened comment. I think that is one of the elements that is needed to do brilliant things, just as Csonger says he was interested in understanding why this was "not possible".
Not even “not curious”, but often just “not curious about that thing, right now”.
Heck, most people understand very little about the myriad of everyday systems that keep them alive. Forget engines, computers or say, type theory. Most literally don’t know where their water comes from (to any level deeper than “it comes in a pipe”) or where their poo goes - and collectively that’s ok because the benefit of society is that this stuff gets taken care of for us.
I think looking up now and then, scratching the surface of our own ignorance, can help specialists build empathy when communicating with others outside their field.
There is a cost to curiosity. Lack of focus. I had a real challenge in my career- I continuously ranged far afield in learning well beyond the direct subject areas of my network engineering career.
That often took away time I 'should' have spent directly researching problem areas applicable to my day job. It finally led me to separating with a company I had been with for 8 years and had a lot of close relationships with.
But what it did was put me on a path to actually use that store of knowledge. Now I'm a Network/Enterprise Architect and a key cog for my current CTO. I wasn't being used optimally previously but I also didn't make sure I was in the correct position that those skills/tendencies could be recognized and not be a hindrance. So much in life is in the framing of things- control that story, and your success is much more likely.
That's a good point about specialists building empathy. One of my college physics professors complained about how dumb the ROTC people were - but a lot of them are humanities majors forced to take physics, which they couldn't care less about. Meanwhile, my political science professor was berating the science and engineering majors because all they would do is memorize and not learn concepts. That was on the day I scored higher on my physics test than my government test, and the physics test was out of 50 and the government test was out of 100.
Although high general intelligence is common among hackers, it is not the sine qua non one might expect. Another trait is probably even more important: the ability to mentally absorb, retain, and reference large amounts of ‘meaningless’ detail, trusting to later experience to give it context and meaning. A person of merely average analytical intelligence who has this trait can become an effective hacker, but a creative genius who lacks it will swiftly find himself outdistanced by people who routinely upload the contents of thick reference manuals into their brains. [During the production of the first book version of this document, for example, I learned most of the rather complex typesetting language TeX over about four working days, mainly by inhaling Knuth's 477-page manual. My editor's flabbergasted reaction to this genuinely surprised me, because years of associating with hackers have conditioned me to consider such performances routine and to be expected. —ESR]
For the last few years whenever I need to accumulate a large and comprehensive declarative knowledge of some subject I use spaced repetition and Anki. My system is simply find a good several hundred page book on the subject then spend four or five days of intense card making. At the end I have a working knowledge and memory of the subject matter at least as good as what most people accumulate in months of ad hoc reading and studying. Also the mere act of distilling the information to cards helps clarify what's important. To maintain the knowledge, I just keep up with the repetition schedule. If I want to go a bit deeper, I'll find the most tagged questions on Stack Overflow and Anki the good stuff from those.
I've had a few situations where people thought I had years of experience with some technology, framework, or programming language when actually I just picked it up last week. One situation in particular, I had a buddy who was using WordPress as the basis of a pretty profitable affiliate site business he was running until he ended up running into some issues that involved needing to make some pretty significant modifications to WordPress itself. He offered to get me in on his affiliate business if I could solve his problem for him. To make a long story short, though I had never even looked at PHP or WordPress before this, within 4 days of using spaced repetition to learn PHP and the intricacies of the WP framework, I was fluently writing plugins that handily solved his problem and a whole lot more and never had to open the reference books again as the whole thing was committed to memory. Not tooting my horn just giving an example of how what ESR was talking about above can be actualized.
I would be very interested in reading about your process (e.g., what kinds of cards you make, what sort of questions you ask yourself, how this varies depending on the source or the subject, etc.).
Have you written about this in any detail somewhere? Or do you have links to resources you used to help learn how to do this?
I too would be interested in seeing how some example cards are structured - to see how much information is on each side of the card, what the card is comprised of, what sort of things are amenable to cards. As I remember cards have a stimulus and expected response. So there’s a bit of an art in distilling things down so the stimulus triggers the correct mental associations and the response can be related back in a way that matches the expected response.
Could you give a concrete example of what you mean by storing pointers to data? Do you mean having trigger words or ideas that you use to rebuild complex ideas?
I always tend to get interested in some technology or library and go straight to the docs. They usually have intro or tutorial/gallery type sections that help you get started. It helps to have a use case in mind and imagine applying it as you read through the API details. This worked for me equally well for ML libraries as it did for learning kernel stuff (except for the latter i went to a reference book).
It's real similar to just doing your normal software job, but in this case it's just for fun!
I think brilliant, not-brilliant, idiot or any other judgemental terms spent on tagging someone or ourselves is just a waste of time.
The greater the effort spent on something by someone, greater is the probability of a favourable outcome to them. But it is still a probability and not a certainty because of variables involved.
As long as people work to live, there will be a pressing demand for such labels, flawed though they may be. Top organizations will always want to hire the smartest/most brilliant/non-idiotic people they can afford, for various practical definitions of these terms.
If two people seem similar in every way, but Person 1 has demonstrated that they achieve greater probability of success compared to Person 2 given comparable time investments, then you'd probably want to hire Person 1. Real life is never this cut and dried, but the principle holds.
I largely agree and work is a test environment, like any other test e.g. SAT; one who performs better is rewarded.
But there's plenty of occasions where judgements is a waste of time, like at the personal level expressed in the OP article. OP does agree that the activity is result from the ego, I don't think he/she did enough to explore the ego resulting in injustice to the colleague as whole in the article.
This goes along with a study I read a while back that you can do practically any job with an IQ of 120. Higher than average, but not anywhere close to genius level. With an average IQ of 100 you can still preform well in something like 80% of jobs.
Software engineers have an average IQ of 110ish so most HN readers are capable of doing almost any job. It's really about effort and things that are harder to measure like creativity
Depends on what you count as "brilliant work". All of the Neuroscience PhDs at Stanford, UCSF and MIT (non-exhaustive list, just examples) I've met are at least 2 standard deviations away from average intelligence and many are 3 SDs away. Makes me think that some brilliant work requires being brilliant.
I guess this highlights the problem, that “brilliance” is in fact subjective and relative. To some “brilliance” is a PhD at MIT, to others it is being able to do something much more mundane.
They need to get more brilliant, because frankly I don't see that many useful results coming from the field of neuroscience lately. E.g. the standard drug for depression being SSRIs for so long makes it seem like we are stuck in the dark ages.
Brains and behavior are absolutely, absurdly, ridiculously complicated, at every level from individual ion channels on up to neurons, circuits, brain regions, and even between people.
It's true that there's a lot of hype, both from well-meaning folks who are excited about their results and cynically manipulative people trying to boost up their careers. It's true that the incentive and career structures are falling apart. But it's certainly not true that the lack of progress is because the researchers are dummies...
> But it's certainly not true that the lack of progress is because the researchers are dummies
True. I have no doubt that it takes extreme-outlier-level intellect to fully grasp the mechanics of neuroscience.
But do you really believe that's it's due simply to misaligned incentives that there has not been a major breakthrough in psychiatry for so long?
Any researcher who made a huge breakthrough would be hailed as hero and should be able to enjoy vast career benefits.
Here's an alternate hypothesis: the experts in the field are so fixated on the mechanics of neurological function, that they can't think laterally and/or holistically about the problems, which is why those problems have continued to go unsolved after decades of research.
The idiom "can't see the forest for the trees" comes to mind.
My own experience is that I spent several years experiencing a combination of symptoms that psychiatry would diagnose with terms like depression, anxiety/panic disorder, bipolar disorder, OCD and personality disorder (narcissistic etc).
The psychiatric profession considers most of these conditions to be incurable, and really only treatable with long-term medication.
I wasn't willing to accept this, so I did my own research and ended up undertaking a combination of subconscious trauma-healing practices, along with nutritional work, exercise (yoga etc) and detoxification of heavy metals and hormone-disruptors.
After about 7 years undertaking these treatments, my symptoms are nearly all gone. Some still linger, but my treatment continues, as does the steady improvement.
Where are the researchers in the field looking into this kind of stuff? It's not as if it's not talked about by prominent figures within the biomedical field. Bruce Lipton, Gabor Maté, Rupert Sheldrake and Stan Grof have been talking about this stuff for a long time, and the hypotheses and anecdotes are right there waiting to be researched.
Sure, one might look at the incentives and career paths that keep most researchers focused on drug discovery and nothing else.
But a key part of being a ”brilliant” person involves a willingness to break ranks with the mainstream and find new explanations and solutions for persistent old problems.
I'd have to agree with your parent commenter that there doesn't seem to have been much of that in recent memory, at least from the psychiatric side of neuroscience.
While this may be true I've become increasingly skeptical of people who make claims like this, if only because I've noticed a strong correlation that they think less of me for my average IQ.
It's not hard to get a perfect score in the quantitative section. You just need to be decent at math, and simply careful and tedious not to make mistakes. It's not hard to get in the 99 percentile of verbal either - just memorize a lot of words, practice, and learn some of the tricks (e.g. for analogy questions, you can eliminate 3 of the 5 choices even if you don't know the meaning of both words presented to you).
Now when they had the analytical section (different from the current one - they retired it in about 2004) - that was a bit more challenging. And that was the one that set people in top schools apart. Getting a near perfect score (770-800) in quantitative was the norm for people applying to any of the top 10 grad schools in technical fields. But the analytical section was tougher. I think Caltech had the highest median (about 780). The next highest was Wisconsin with 750. The next highest after that was 720, and the rest of the top 10 were in that vicinity.
But they scrapped that section - the only one that was a useful differentiator.
Anyway - bottom line is getting good grades in GRE in the old days (and I suspect today as well) is mostly an exercise in memorizing words, and learning test taking strategies - including simple things like if you have to solve an equation and have 5 choices for x - don't bother solving it - just plug each x in to the equation to see which one works.
Most people don't do this well because:
1. They cram - spend only 2 months preparing for it.
2. They don't spend time in figuring out efficient test taking strategies. Actually, they don't need time - they just need to read a few books on the topic!
I spent over a year preparing for it (only a few minutes a day, and with gaps of weeks at a time). It wasn't hard to get a perfect score in two of the categories and a 99 percentile in the third. And then when I went to grad school, I didn't perform better than those whose score was nowhere near as good.
Is there a reference or study for that? I thought GRE scores were mostly a proxy for cramming English vocabulary into your brain (verbal) and for knowledge of basic geometry facts and inferences from them (quantitative).
Only to some extent. I've tutored for the SAT and seen some improvement, but you can't get people to perfect scores. I got a perfect score with a bit of practice. Khan academy and the college board say your score can improve on average 115 pts with study, which is not nothing, but is still a long way from 1600 when the average score is ~1000.
So when someone says that people treat SAT scores as a proxy for an IQ test, they mean someone with a 1550 likely has a well-above average IQ, and a higher one than someone who got a 1250, who's still probably somewhat above average.
This was a good article, but I'm skeptical if someone from UWaterloo can attest to being "Normal". People that came from Waterloo are pretty exceptional on average unlike normal people like me.
Really good read. So true that most 'innovation' is really a lot of tedious work most people don't bother doing. That is in sharp contrast to the 'Eurika!' moment that most people think of. There may never be a eurika moment, just a bunch of small insights put together into a body of work.
Not sure about that. Brilliant work is by definition brilliant. :D Although you can do pretty great things and even be very famous or very rich even if you are not smart.
> Then learning the answer revealed that it's actually something that can be fixed --- as is always the case with these things if you think about them enough.
The author focuses on the consequent, thinking long and hard about something, but misses the antecedent, acquiring a deep understanding of a problem area.
I'd say the antecedent is as important as the consequent, if not more important. If you went to grad school you likely knew tons of brilliant people who worked long and hard on important and difficult problems and got nowhere. What they were missing was not some secret effortsauce, but that they picked a problem on which they didn't have an opening or insight into, so all their hard work was them spinning their wheels while standing still.
One thing I tell my students now is that they should always pick problems where they have something everybody else working in the same area doesn't. Like if you want to do formal verification but your background is years spent in industry as a kernel developer, you should avoid the temptation to chase the latest fad (e.g., adversarial machine learning or Spectre/Meltdown), and instead pick a poorly understood and validated module in the kernel and try to find out what you can do to make that module more secure.