Hacker News Reader

Comment by yummyfajitas | Hacker News Reader

I see a lot of articles claiming that programming talent is a myth. Yet none cite a source or back up this claim.

The closest they get is arguing by consequence against it - if it were true, people might be discouraged. That's a logical fallacy since some things can be both true and discouraging.

This article is even worse. It drops this paragraph in: The tech industry is rife with sexism, racism, homophobia, and discrimination...This is the "10x programmer" who is so good at his job that people have to work with him even though his behavior is toxic.

Arguing by juxtuposition, it's suggested that a belief in talent will somehow harm diversity. But that makes no sense - if talent is so important that we are forced to work with "toxic" 10x programmers, why aren't we also forced to work with black female 10x programmers no matter how much we might dislike them?

View on HN

> I see a lot of articles claiming that programming talent is a myth. Yet none cite a source or back up this claim

That's because the article writers have rarely worked with a 10x programmer. They get shit done, fast, with massively below average bugs. It comes from a deep understanding of the code, the language, and the metal. They're also more likely to innovate and/or find novel solutions.

In the past they were spread around or sometimes in research (Bell), now they're clustered in the big 4 (Google, Apple, Microsoft, FB) and SF.

So unless you are where they are, you won't find them or even know they exist. That guy who can wrote a cool Jquery plugin at your startup probably isn't a 10x programmer. John Carmack would be an example of a 10x.

The guy who is at your startup may actually be a better programmer than some of those who work for the big 4 you named. Many super-hero programmers are not motivated by working for huge companies with an established name, nor by the slightly better money they give compared to startups. I personally know some programmers at Google / FB / Apple / Microsoft and I also know a few working for startups or doing their own startups and, while all of them are probably one of the best I know, there is no obvious skill level difference between them and definitely not "10x".

GP didn't say all programmers at the big 4 were like that, just that most of such programmers were at the big 4. These are all huge companies so anecdotal evidence involving "some programmers" is unlikely to overlap.

Indeed those are huge companies and they probably have more skilled people than the average, but I seriously doubt they have most of the best programmers. Simply the amount of people working there is too small to cover even 1% of the best programmers in the world. And there are plenty of skilled programmers working in places where those huge companies do not have even an office, and there are plenty of programmers not considering working there for various reasons.

Agree but they likely have a disproportionate amount of the top 0.01% or whatever extreme number you want to use

since you can define the cutoff for "best" as you want, you can say 1% while someone else would say that that cover 10000%...

Fabrice Bellard (LZEXE, QEMU, ffmpeg, tcc, ...) is another example, although not many know about him - which agrees with your last point.

If you read hacker news you would notice that an article about him comes up every week or so. The same article.

In the past they were spread around or sometimes in research (Bell), now they're clustered in the big 4 (Google, Apple, Microsoft, FB) and SF.

The issue is that people are mostly context. Some programmers do well at Google and wouldn't do well in another corporate environment, and vice versa. "10x" is real-- hell, sometimes it's 100x or 1000x-- but it ignores the fact that it's not always the same people who are 10x. Drop an R&D innovator into a subordinate role where he's just expected to "get shit done, fast" and churn tickets, and you'll likely have a bored 0.3x or toxic -5x programmer.

10x is shorthand for "as important to the business as an executive", and good engineers often are more critical than many highly paid executives. The problem is that the terminology is a bit negative on "average" programmers and the idea that "10x" is a permanent identity is a bit silly.

>>10x is shorthand for "as important to the business as an executive"

This is very true. Another thing is lack of a clear purpose. To wake up every morning and to fix bugs day after day, without any higher or a coherent goal to chase demotivates people badly.

The only other thing I which causes burn out I know of is lack of meaningful rewards. To know that there so much politics around and you are being treated unfairly. And you are failing repeatedly for reasons beyond your control.

I don't know of one single person worth I know of who can't handle a lot of work or can't take long working hours. But at least give us a cause worth fighting for.

Except 10x programmers, if they exist don't just emerge fully formed into the world, they engage with some community somewhere to learn part of their skills. That community might be an online forum, a company, a tech meetup, university course etc.

If a hypothetical black female potential 10x programmer finds the first community is unwelcoming then she is less likely to stick around and thus become a 10x programmer.

There are plenty of previous articles on here from individuals who are "tech minorities" about the environment that we unconsciously create which isn't that welcoming to them. From assumptions about where and how we want to socialise, to dress codes, working hours, employee benefits etc.

Is your point, that if 10x programmers exists, then it's bad, because the community that formed them was possibly unwelcoming?

Lets distinguish two separate concepts:

Unwelcoming: the community dislikes a person and behaves badly towards them. As an example, a recruiter told me not to bother with Warby Parker or Etsy since I'm not a hipster and they'll just reject me. (No idea if this is true, might just be #shitrecruiterssay.)

Not my thing: a person dislikes the community, it's norms and culture, and avoids it. E.g., I don't like banking culture.

You seem to be suggesting that "tech minorities" feel that tech culture is "not my thing" - we socialize and dress in a manner they dislike, work hours incompatible with their preferred lifestyle, and don't give out the right employee benefit package. As a result, "tech minorities" don't put in the effort and don't build the relationships that would help them become 10x programmers. I get the impression you consider this a problem, but it's unclear to me why that is.

How you get from this to the idea that 10x programmers or programming talent are a myth is not something I could figure out from your post.

I think what he's suggesting is that minorities find our community unwelcoming, and by unwelcoming they're probably referring to the racism and sexism that's rampant in the tech community. Unlike you, they're not worried about being turned away from a job for not being a hipster, they're worried about being turned away from a job for being black or a woman. Or, worse, they're worried about the abuse and disrespect they will undoubtedly get while working in tech.

Cmdkeen explicitly refers to dress codes, benefits packages and socialization patterns, not abuse, disrespect and racism.

People constantly talk about racism and sexism being "rampant", but evidence of this would be helpful. Citing statistical disparities is insufficient since, as cmdkeen notes, this could also be caused by "tech minorities" disliking the culture of tech rather than techies disliking them.

In some (very loose) sense, those two things are the same thing. It's not that these people say to themselves "man, I hate how I can't wear X shirt to work, better quit" but rather "the culture here (dress code, socialization patterns, etc) seems to default to white, and people view me as an outsider." I think that that's probably much more common than explicit disrespect.

They aren't the same thing. One is something that is done to you, the other is something you choose to do to yourself (or not, if you find it more useful).

My company is about 95% one race (not mine) and there is precisely 2 people here of my race (since last week) and 4 of my nationality. I simply choose not to care, and define my work tribe as "folks who care about uptime and good stats" rather than "folks who are genetically similar".

Isn't this diametrically opposed to the prevailing opinion that 10x programmers aren't grown, they're born that way with "the knack" and burned out before age 35 at which point due to ageism they need to be in management or onto their second career?

If, theoretically, 10x programmers were grown, then obviously most of the 10x programmers would be older and more experienced, but that age group is actively strongly discriminated against. Perhaps there are no 10x programmers of 50 yrs, by that age they are all 100x programmers after all that accumulated experience and domain specific knowledge, at least compared to average recent grad noobs. With the obvious exception of people with 25 yrs of experience who've only had 25 first years over and over again, although that failure mode is hardly limited to programming...

With a side dish of the widely held belief that 10x programmers can only come from selection organizations like ivy unis which primarily select (don't get all wound up, I very intentionally selected the word primarily, not solely) based on kids winning the genetic lottery, not by working hard? If the only way to be a 10x is to win the genetic lotto, and if the lotto winners happen to be mostly rich white males, the problem is at the academic uni acceptance dept level and has nothing to do with programming itself. If 10x programmers can by some definition only be recent Stanford grads and Stanford refuses to admit "enough" minorities, the problem is very specifically Stanford, not programmers in general or education in general or culture in general. Ready Fire Aim isn't going to hit the right thing. Find the correct target.

If an individual is "born that way" should he be made to feel unwanted or guilty about who they are? Apparently if they're gay the answer is no, and if they're a 10x programmer the answer is yes. This seems inconsistent, and inhumane.

Also it assumes that programming is only learned at uni in a purely frat boy anti-minority environment yet mythology and anecdote indicates that really good programmers are all self taught before they're 18, and only go to uni to get the paper and polish off the highest end stuff.

As a completely isolated theory, it is internally self consistent, but rubbed up against pretty much every other "common knowledge" of tech or geek or nerd or 10x culture, it no longer is consistent. Either everything else is wrong, or the proposed theory is wrong. Its unlikely everything is wrong (or is it?). Therefore the proposed theory is wrong.

This actually provides a nice concrete way to test whether non-Asian monorities/women/etc are really harmed by outside factors: compare the percentage of autodidacts among non-Asian minorities to Asians/whites. If social factors push non-Asian minorities out, then the percentage of autodidacts should be higher (since social factors are reduced by a lack of social interaction).

>There are plenty of previous articles on here from individuals who are "tech minorities" about the environment that we unconsciously create which isn't that welcoming to them. From assumptions about where and how we want to socialise, to dress codes, working hours, employee benefits etc.

We are welcoming if our culture suits you, like any other culture since the dawn of time. Why should we change the way we dress, the way we want to work, that we socialise, if we, that identify with the culture are perfectly happy with it.

They are feeling unwelcome because they are unwilling to adapt, but it's taboo to say so.

What you are advocating is purely and simply destroying a culture because some people do not like it.

It's a job. Why should culture matter so much?

In this article/talk, the "talent myth" in question is not that individual 10x programmers exist. It's that the distribution of talent is bimodal, that you're either a rockstar or "not really a programmer" with nothing in between.

Note the example of Ann Trason, who would surely count as a 10x runner. The normal distribution doesn't preclude this; it merely makes it improbable that any given person is, in fact, a 10x runner.

Of course, it's much easier to identify 10x runners, since there's a clear metric to rank by; how do you do that in programming? I believe I've worked with 10x programmers, but how do I know this isn't just another "I know it when I see it" fallacy on my part?

Probably because you say terrible things like, "why aren't we also forced to work with black female 10x programmers no matter how much we might dislike them?", and all the people with money, voice, and hiring authority look to people like you for validation.

Let me introduce you to the argument technique of "proof by contradiction". Such a technique requires assuming the author's premises to be true (in this case, the article's premise that techies are racist and sexist).

https://en.wikipedia.org/wiki/Proof_by_contradiction

Also useful reading: https://en.wikipedia.org/wiki/Principle_of_charity

Condescension combined with a narrow view on diversity, that's a good look.

It's a question posed because of the statements in the article that the tech industry is racist and sexist, and 10x programmers are toxic, not his actual view on black female programmers...

Conversely, can you cite a source backing up that there is such a thing as programming talent? Not saying either side is right or wrong, but I'd be interested in seeing how you'd even go about researching such a thing.

I don't know of much good data on it, all I have are anecdotes. I personally believe in 100x programmers, because I've worked with quite a few who were 10x worse than me and I've worked with a small number who were 10x better.

I don't make the claim that 10x programmers definitely exist. My only point is that a) argument by consequence is a fallacy and b) the article is internally contradictory (the more talent matters, the more incentive there is to identify and exploit it regardless of what it looks like).

The truth is that programming isn't a passion or a talent, it is just a bunch of skills that can be learned.

I have to disagree with this part. I always considered passion to be a huge part of talent. And you can be passionate about anything in life.

Anyone can learn any skills (from painting to solving differential equations). However, if you're passionate about using those skills, you learn faster, easier, and you're more inclined to spend your time honing them and developing a deeper understanding.

Programming is, for the most part, a mental exercise.

How good someone is at programming is a combination of their problem solving skills (natural talent), their existing knowledge, and their ability to seek out and learn answers to problems they don't already know the answer to.

The talent part of the equation is a big factor, and ultimately determines where the individual tops out their skill set and "plateaus".

For an example of this, look at chess players and the ELO rating system. Chess is purely mental and has many similar parallels to programming.

I've met hundreds of chess players who have been playing chess for years, logging thousands of games. Many of them study the game on a regular basis and attend tournaments, and they all have ratings.

A beginner chess player starts with a provisional rating of 1200, which goes up or down depending on wins or losses logged over time. Grand masters are players that have earned a rating of at least 2600 or higher.

For every chess grand master out there (think of them as a 10x or even 100x producer), there are thousands of chess players that have been slogging away for years without their ratings improving past even 1500. They are simply blocked by their lack of natural talent, no matter how much they study or practice.

This is a fact that is objectively measurable and is backed up by millions of data points from officially rated chess games.

There's no reason why the same laws of nature related to raw mental talent (specifically problem solving and memorization) wouldn't also apply to programmers.

10x programmers do not exist, because programming ability is not a scalar. Give a hello-world task to write to an average programmer and a super-hero programmer. You won't probably see a big difference in their productivity. Actually if the task is too easy, your super-hero might get bored and become even less-productive thn the average programmer. Now give them something really hard to solve. If the task is hard enough, there are chances your average programmer would give up, regardless the amount of time you give him to solve the problem.

'10x' isn't meant literally. No one thinks there is one number we can assign to people (a lot of what we talk about isn't exactly measurable), and then one people is 10x some other person. Well, probably a few use it literally, but most don't.

Oh, yeah, I agree, but many people are taking this too-literally. That's why I don't like this label and call them "super-hero" programmers or "rock-star" programmers. As far as I remember the original term "10x" was created after some studies finding the differences between the worst and the best programmers. But this didn't mean there was nothing in the middle, only that the variance was high. And I completely agree there are plenty of really terrible programmers, so being 10x better is actually an easy task if you compare to them ;)

One of my beefs with the 10x term is that being concrete it leads to people taking it literally. These silly buzzwords have a way of ending up literal reality

I can use your logic to prove that no one is good at anything:

"Super star athletes do not exist, because athletic ability is not a scalar. Ask a little league player to hit a tee-ball 15 feet and an MLB player. The ball will get to 15 feet at roughly the same time. Actually since it's too easy, your super-hero might get bored and become even slower then the little league player. Now ask them to hit a home run at Wrigley Field. The little league player would give up, regardless the amount of time you give him."

I never said super-star programmers do not exist. I just don't like the term "10x" because it suggests there is some scalar measure of skill.

There are many scalar measures of skill. They are well-correlated with one another.

I ROFL when I read your comment, in all of its awesomeness.

I have a feeling that, if more people used analogies like this, the world would be a better place, because people would actually understand things they read.

Reply on news.ycombinator.com