Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The Lesson to Unlearn

paulgraham.com

Post reply on HN
Print view View original

181–190 of 588 posts

Re: The Lesson to Unlearn

#181

Post by impendia »

hooande wrote: post #106

I know that learning how to hack systems is valuable because that's how I got into YCombinator. I set out specifically to hack the application process and succeeded. It wasn't even that difficult to do. If pg believes what he's saying in this essay, I have two questions: 1. why did he design a testing system that was easily hacked? 2. why does YCombinator itself focus so much on hacking the venture capital system, as…

> If you can't create a 200 person per year startup incubator that focuses on true learning instead of process hacking, then how could someone do it for a 20,000 person university?

As a university professor, I can at least try. Not at scale -- certainly not for my entire university -- but for my own classes as an individual.

There are several constraints, of course. I have to give grades, and students care about grades, a lot. These grades should measure (to some extent) how prepared students are to proceed to whatever is next. Students have a number of conflicting priorities, and it is natural to concentrate on whatever is most urgent. And students come with wildly differing backgrounds and interest levels (and amounts of available time), and I have to be consistent and fair to everyone.

But, when I grade, I have a lot of leeway to decide what I believe is worth measuring, and measure that. My aim -- which I hope I at least partially realize -- is to ensure that the best way to "hack" my classes is to learn what I hope the students will learn.

Re: The Lesson to Unlearn

#183

Post by kr4 »

tomjen3 wrote: post #45
volume wrote: post #39

Earlier quoted context omitted.

"The capacity to just build things without fear of being judged is invaluable" For me, I had to unlearn self judgement. This has helped so much over the last few years.

How did you do that? It seems to be a big stumbling block to many.

You can try a two-step exercise mentioned in this post: https://os.me/forgive-yourself/

Re: The Lesson to Unlearn

#184

Post by jtanderson »

impendia wrote: post #181
hooande wrote: post #106

I know that learning how to hack systems is valuable because that's how I got into YCombinator. I set out specifically to hack the application process and succeeded. It wasn't even that difficult to do. If pg believes what he's saying in this essay, I have two questions: 1. why did he design a testing system that was easily hacked? 2. why does YCombinator itself focus so much on hacking the venture capital system, as…

> If you can't create a 200 person per year startup incubator that focuses on true learning instead of process hacking, then how could someone do it for a 20,000 person university? As a university professor, I can at least try . Not at scale -- certainly not for my entire university -- but for my own classes as an individual. There are several constraints, of course. I have to give grades, and students care about gra…

Agree wholeheartedly. That's exactly the approach I take, but the problem of course is that 1) it's difficult to do successfully, and 2) students "freak out" because the exams become unpredictable and they end up getting more stressed out that they need to about succeeding at the exam. Of course, my regular high-performing students do well regardless, but I haven't found a way to convince them that they have the skills already to do well on the exam. Instead, they pull all-nighters and ultimately damage their results on exam day when, if I had given it as a surprise exam, they would have been just fine. So on some level, it's a need to teach self-confidence, self-diagnostics, and introspection. This (again) is a TON of work, since it's necessarily person-dependent, so only professors who have the time and energy can actually attempt to do it. The others will just say "screw it" and give an exam of sentences copy/pasted from the textbook, but with keywords, blanked out. Yet another extreme are the professors who endlessly allow exam re-submissions, so students just brute-force hack their way to a passing grade.

I want the obvious response to the above to be "okay, just raise the bar for teaching professors" but, unfortunately, those decisions are usually made by people who are completely ignorant to the process, and who rely on broken systems of teaching evaluation. Those who would be good at making those decisions are busy doing the teaching, and doing it well. Also, I don't personally think we want a system that so micro-manages our professors' pedagogical styles. I get fairly positive student feedback, but so do the professors who offer endless exam re-submissions and whose students have basically zero material retention after they pass the class. The optimal (but impossible, partially due to sheer laziness) way would be for students to self-select themselves into classes with professors that they believe will ultimately teach them more and better material, without having the extra pressure of the effect on their GPA.

(edited with some last-minute thoughts).

Re: The Lesson to Unlearn

#185

Post by Ozzie_osman »

hooande wrote: post #106

I know that learning how to hack systems is valuable because that's how I got into YCombinator. I set out specifically to hack the application process and succeeded. It wasn't even that difficult to do. If pg believes what he's saying in this essay, I have two questions: 1. why did he design a testing system that was easily hacked? 2. why does YCombinator itself focus so much on hacking the venture capital system, as…

Spot on. If you're interested in this sort of stuff, I'd recommend the book Seeing Like A State. The author basically walks through most of history as viewed through the lens of authorities trying to measure things (usually to extract taxes or exercise control). But basically traces everything from the layout of cities to how we name ourselves down to authority trying to standardize things so they can measure them.

Re: The Lesson to Unlearn

#186

Post by watwut »

hn_throwaway_99 wrote: post #122
ken wrote: post #90

I'm glad he made the leap from school tests to funding tests, as it seemed to be a thinly veiled analogy from the start. I'm disappointed, though, that he didn't take the final step and admit that money itself is also a poor, hackable test. Funding, growth, usage -- these are all still one level removed from something worthwhile or beneficial. Cigarettes have amazing usage numbers even in 2019. I'm waiting for the en…

I think this is an excellent point. I really agree with pg's thesis, but I think it would be worthwhile for him to go more in depth into the "edge cases" he talks about in passing, mainly because I think these edge cases actually bolster his thesis, not detract from it. Companies like Theranos and WeWork succeeded for a short while because they tried to "hack the test". E.g. Theranos hacked the metrics of "buzz", "in…

Theranos issue was not lack of product. Their problem was massive fraud going effectively demanded from top levels.

Framing it as lack of product is unfair to all startups who lacked good product, hacked buzz, got investment, got notable people and then did not committed fraud.

Plus it suggests that Theranos fraud is result of not having product. The same people would be engaging in unethical conduct with product too.

Re: The Lesson to Unlearn

#187

Post by derefr »

hooande wrote: post #106

I know that learning how to hack systems is valuable because that's how I got into YCombinator. I set out specifically to hack the application process and succeeded. It wasn't even that difficult to do. If pg believes what he's saying in this essay, I have two questions: 1. why did he design a testing system that was easily hacked? 2. why does YCombinator itself focus so much on hacking the venture capital system, as…

> There seemed to be no time given to "how to make a great product", likely because this is something that can't be taught

It can be taught, but the general principles are weak (it’s basically “how to be an effective futurist”), with most of the useful knowledge being domain-specific.

And the thing with the specific domains is, either they’re not understood yet—in which case everyone is flailing around—or they are well-understood—in which case the market has already consolidated around the companies who can best operationalize the set of techniques required to create great products for that market, and the best you’ll hope to do is to get bought out by one of them, not to compete or replace them.

Because, in the end, having a great product is a multiplier, but so is execution; and execution provides compound interest on its gains, such that corporations inevitably execute exponentially better as they age.

In other words, the only time where you can set out to win a market, and actually have a hope of doing so, is precisely when we don’t yet know how to operationalize the production of great products for that domain.

Re: The Lesson to Unlearn

#188

Post by blondin »

esotericn wrote: post #8

Not sure I agree with this. The real world is very much about optimizing for and beating tests. It might be producing a specific CV or preparing for an interview whiteboard session to land a job. It might be socializing and networking in a specific way in order to land funding. The real world does not commonly reward just being really good at arbitrary things. It's almost always focused on meeting a need that someone…

> The real world is very much about optimizing for and beating tests.

thank you.

it's awesome for someone like PG to bring the subject upfront. but what you are saying resonates with me.

genuine interest is missing in our field, and to some extent many others. "beating tests", as PG puts it, is at all time high. i am at a point where i don't know who or what is right.

the numbers are telling an important story! those who learned to hack the tests or the system are popular online or offline. and they are "succeeding" in life. they seem to have outnumbered those who put genuine interest first.

for instance, as many of my peers, i wanna learn machine learning and AI. but it is hard for me to find materials that resonate with me. materials that teach from first principles like those that got me hooked back in the days. they are lacking because we are so good at shortcuts and hacking. maybe we don't know how to do it anymore?

but who am i to blame anyone? there are so many, many, ways to hack and beat the tests and the system. and seemingly you can get ahead of many others in life by doing so. #AceYourFirstInterviewAfterBootcamp #TensorflowPyTorch #BeSureToLikeBelow #ThanksToMyPatreons #SubForChatAndEmojis #InstagramFacebookDown

i am very thankful for PG for starting the discussion. and for this comment.

Re: The Lesson to Unlearn

#189

Post by strictfp »

daly wrote: post #93

Are you in EECS? Most big data centers use FPGAs "on the wire" to do things like encryption on the fly or compression so that they get "wire speedups" without burning CPU cycles. So FPGA solutions that "live on the wire" and can be "plugged into the LAN port" are a really useful area you can hack in your dorm room. I worked in the security area. One of the key problems is "exfiltrations" where someone tries to copy v…

If you're one of those persons, where would you look for a job?

Re: The Lesson to Unlearn

#190

Post by derefr »

TeMPOraL wrote: post #145
adwn wrote: post #133

> The way you get lots of users is to make the product really great. That is wrong – or, at the very least, incomplete. "Build it and they will come" is the dream and misconception of every programmer who's talented at software development but not at sales. Hell, this is the number one advice of every business how-to book ever: You can't rely on people finding out about your great product by chance, you need to put i…

100% right, and the corollary of this is: you don't actually need to make a great product, if your sales&marketing game is good enough. Or, in other words, a lot of successful business - including startups - is exactly the "test hacking" PG urges founders to unlearn. I'm honestly surprised by this essay arguing that "hacking the test" is the wrong approach in startups/business. "Hacking the test" is essentially what…

> you just have to discover how the system really operates.

One might say that, by analogy, this is like saying you cheated on a test by studying. “Learning how the system really operates” isn’t hacking; it’s the thing you’re formally setting out to do as a business!

Post reply on HN