Comment by adwn | Hacker News Reader

Comment by adwn | Hacker News Reader

Parent Full thread

> Every time I read one of PG's posts, it seems like he's working from a narrative that he's trying to conform facts to.

Indeed. Compare this gem from [1]:

You could probably work twice as many hours as a corporate employee, and if you focus you can probably get three times as much done in an hour. [1] You should get another multiple of two, at least, by eliminating the drag of the pointy-haired middle manager who would be your boss in a big company. Then there is one more multiple: how much smarter are you than your job description expects you to be? Suppose another multiple of three. Combine all these multipliers, and I'm claiming you could be 36 times more productive than you're expected to be in a random corporate job. [2] If a fairly good hacker is worth $80,000 a year at a big company, then a smart hacker working very hard without any corporate bullshit to slow him down should be able to do work worth about $3 million a year.

The process is: Start with a hidden conclusion, selectively pick convenient facts (or make up bullshit numbers), ignore that they don't make sense in context, and finally present your a-priori conclusion.

[1] http://www.paulgraham.com/wealth.html

I'm not seeing the issue with the except you posted. In context, it seems clear to me that those are all very back of the envelope estimations. Not everything needs to be backed by multiple citations to double blind trials; approximations can be quite useful.

The point (IMO) is to take statements you strongly suspect to be true from personal experience (ex that a pointy-haired middle manager reduces your productivity) and attempt to assign ballpark numbers to them. The numbers don't need to be super accurate, merely within the correct order of magnitude. The hope is that such an exercise might allow you to estimate an expected return (to within an order of magnitude) for decision making purposes, or perhaps just to better understand the world around you.

Put another way, even if you dispute the exact outcome ($3 million a year), presumably you agree with the broader premise of being significantly more productive? (Note that increased productivity won't necessarily translate to increased income. It should give you something to work with though.)

Software Engineers sometimes act as if administration and coordination of effort are unwelcome distractions to productivity, but those are necessary components to any viable commercial effort. The super-productive, "$3M/year engineer" who goes and starts a successful new startup always metamorphosizes into the pointy-haired boss. Why is that? Because that is the steady state of companies, anything else is transient.

However, VC's make money from a steady stream of disaffected engineers who are convinced that there is a better way - for a while, there is, until the same VCs ask the founders to step-up their administration-game, or they have to raise capital, or manage paperwork.

I mostly agree, but I think it's important to note that the necessity of administrative overhead (and associated friction) scales with team size and project scope. I expect a lone developer to be the most productive by far in the narrow sense of producing working code. However, at some point a single person just can't do everything that needs doing themselves.

So I'll agree that as things grow administrative overhead will necessarily creep in. But I expect that if you can find a way to turn a profit, a smaller team will probably involve significantly less overhead. So if (as in PG's hypothetical) the developer is worth $80k at a big company, how much would you estimate them to be worth at a company of only 4 people that has a good idea and funding?

> So if (as in PG's hypothetical) the developer is worth $80k at a big company, how much would you estimate them to be worth at a company of only 4 people that has a good idea and funding?

My point is that a 4-person company with a good idea and funding doesn't usually remain a 4-person company. Sure, they could be churning out good code at an amazing rate, but is transient state - in order to make money at the levels that the VC-class demands, pointy-haired bosses will be required, and they come with structure, meetings, emails, templates, and other administrata rituals.

edit: I think this is proof enough that VCs don't optimize for individual productivity - nobody who likes having more money does.

Hm? That paragraph is accurate. When you're free to work flat-out – as hard as you can, as much as you want – you can make a lot more progress than if you were paid to do the same thing for someone else.

36x may not be the exact multiple, but it's correct within a power of two.

I think you're making the point. Correct according to... what... lol? The preconceived notion that it should be?

I don't even necessarily disagree, but, "No, it totally is right" isn't exactly an argument.

Personal experience. But, my experience is merely anecdata. Does it hold true in general?

It seems to. Suppose your company consists of ten people, and its valuation rises to $10M. The value must have come from those ten people. If they weren't at the company, the company would not be worth $10M.

It would be easy to dismiss that example too. After all, I made it up. But that kind of example happens all the time in startups. It's such a striking correlation that you'd now have to explain away many hundreds of examples as coincidences, or claim they're due to other factors like the founders' connections, upbringing, being in the right place at the right time, etc. Those things do matter -- in fact, you'd be hard-pressed to start a successful startup without at least one of those -- but you can't bullshit your way through a startup. Either you've made something people want, or you haven't. And if you haven't, your upbringing and personal connections won't change the fact that your startup is doomed.

So it seems the simplest explanation is probably true: a small group of smart, dedicated people can do enormously valuable work, given the right conditions. And pg's paragraph seems to explain those conditions pretty well -- at least in my experience.

Valuations are bait that VCs use to motivate founders and employees to take below market wages. But after all the rounds are done, the employees are almost always massively diluted back to below market wages. You could just have diluted them in advance, but then they wouldn’t have worked as hard.

Opportunity, luck, timing, capital, risk, decisions, specific knowledge. Converting hours into value is way more than productivity.

> Hm? That paragraph is accurate.

No, the assumptions in it are ridiculous in context:

1) "You could probably work twice as many hours as a corporate employee": Okay, maybe you can "work" 80 hours a week, as in "be in the office 80 hours per week".

2) "if you focus you can probably get three times as much done in an hour": No. You can't stay focused for 11.5 hours a day, especially not 7 days in a row, especially especially not week after week after week.

3) "You should get another multiple of two, at least, by eliminating the drag of the pointy-haired middle manager who would be your boss in a big company": No. First, that's already covered by "being focused" in point 2. Second, there will be other aspects that will drag down your raw coding output, like clients, employees, legal matters, or organizational issues. They can be worse than your boss.

4) "Then there is one more multiple: how much smarter are you than your job description expects you to be? Suppose another multiple of three": No. At 80 hours per week, you're not three times smarter/more productive per unit of time than an average employee at 40 hours per week.

Look, I'm not saying that an equivalent compensation of $3M per year for a founder is wrong. I'm saying that those numbers are bullshit. Everyone can make up figures, multiply them, and arrive at any conclusion they want. That doesn't mean there's any truth in them.

No. You can't stay focused for 11.5 hours a day, especially not 7 days in a row, especially especially not week after week after week.

I'm a former software engineer. The best engineers absolutely, genuinely can work very long hours and remain productive. 80 hrs/week might be a bit high, but 65 hours a week at 3x average-programmer-productivity is definitely something that exists in the real world, and although it is rare, it is not vanishingly rare.

Hacker News is in denial about this extremely simple fact, even though you can easily verify it by e.g. working at FAANG for a couple years. Probably because lots of Hacker News readers are kids who have not gotten to work with a broad cross-section of programmers; and because the rest of us are literally slacking as we post here, and want to believe our slacking is some kind of biological necessity.

I agree with your broader point that PG's line of reasoning is, at best, a gross simplification of the real world (although I still enjoy it as an intellectual exercise).

> and because the rest of us are literally slacking as we post here, and want to believe our slacking is some kind of biological necessity.

Ugh, this one hit deep... why must you do this to me?

Lots of HN readers are familiar with things like burnout from personal experience at FAANG.

> [...] and because the rest of us are literally slacking as we post here, and want to believe our slacking is some kind of biological necessity

Maybe that's because we've observed one too many "I work 80-100 hours per week" CEOs, who still find time to write blog posts, keep up with the latest memes, and shitpost on Twitter.

hey now. that blog is part of the job, memes are cultural research, and shitposting is how you stay in touch with retail investors. that's at least thirty hours a week together

> and because the rest of us are literally slacking as we post here, and want to believe our slacking is some kind of biological necessity.

When people are shamed for taking a break, no wonder there is a lot of burnout and mental problems in tech.

You don’t have to work 80h’s though. Many employees waffle around so much that you can get the same work done in an hour or two of distraction-free, meeting-free, daydream-free grafting.

I agree with point 2.

Point 3 may be valid, but have you ever worked for a real idiot?

Point 4: remember that an IQ of 100 makes you smarter than roughly half the population. If you can string two formulae together in Excel, you’re probably ahead of the game. That may be a little cynical but my point stands.

I agree with you. I run a corporate data science department. Maybe I'm an anomaly, but I have a lot of freedom to decide how things get done. And when our website went down the other day, I got to hand the problem to our Tech team, go to sleep, and wake up to find a functioning website because I work in a company that pays a skilled team of people to handle problems like that. One of my friends was moving from a salaried job to freelance and balking at charging really high rates. I explained how much BS other people handle for her in her job, like buying computers, finding clients, etc.

Reply on news.ycombinator.com