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

Why is it hard to buy things that work well?

danluu.com

Post reply on HN
Print view View original

191–200 of 542 posts

Re: Why is it hard to buy things that work well?

#191

Post by scaramanga »

PragmaticPulp wrote: post #169
Strilanc wrote: post #88

It's pretty depressing how many of the comments are purely answering the title and not discussing the content of the article. In particular, the comments seem mostly focused on consumer product build quality whereas the article is more focused on e.g. businesses trying to outsource something (like same day delivery) and getting a worse result even when they pay more than doing it internally. And the cultural obstacle…

> I will note that I often wish Dan Luu was a bit less apparently-uniformly-confident in some of the statements he writes. He does back up what he says, but he'll use the same tone for something he's seen anecdotally and something he's spent a month personally investigating. Dan Luu is one of my favorite currently-active bloggers. He's obviously a smart person with a good amount of experience in specific areas. Howev…

I think that's kind of the point. Dan Luu can change his priorities to do a performative display of what interviewers expect and then Dan Luu can pass a job interview. In which case you are not hiring Dan Luu, but a different guy who has most likely internalised a different world-view and set of values. Even if you assume he can compartmentalize his interview performance skills, he will have at least devoted time to that, instead of to something "more useful".

I don't think Dan Luu is struggling for work.

So the question is do you want to hire someone like Dan Luu or not? If the answer is yes, then you might want to consider how your interview process might interact with such a person. If you are thinking from the perspective of someone doing the hiring (as those articles seem to be), it is nonsensical to simply respond by saying "well, the candidates just need to be more submissive and compliant to whatever our process is."

Perhaps Dan Luu is a unique snowflake and/or nobody needs to hire someone like that. Or it could be that there are quite a large number of developers with attitudes and experience that lead to similarly unproductive or inefficient interactions with tech interview processes because those processes may be fixating on having the candidate do a specific performance rather than trying to understand what individuals can offer and whether that would be useful when added to the existing team. It could be that such developers would be as good or better fit, in a lot of cases, than people who can do oscar-winning interview performances.

Re: Why is it hard to buy things that work well?

#192

Post by hyperpallium2 »

The bit about being unable assess expertise reminds me of Dunning-Kruger, and also The Prince:

> A prince who is not wise can never get good counsel, unless he puts himself completely in the hands of a wise man; but such a man will soon take over his state.

What do you do? Become expert at everything? Form a coalition of experts, who share their assessments? Or... somehow... acquire the ability to assess "expertise" in general, regardless of specific field?

Re: Why is it hard to buy things that work well?

#193

Post by emtel »

DannyBee wrote: post #148
mathieutd wrote: post #125

Earlier quoted context omitted.

Do you a reference for these statements? I'd be curious to read more about the topic.

Weak form efficiency - https://arxiv.org/abs/1002.2284 is one such proof. For strong form efficiency (which requires markets perfectly reflect all available information) - You would have trouble finding people who still believe in strong form efficiency. The weak form paper above cites several papers that go into why strong form efficiency is impossible. There are formal proof versions around. In practice, even empir…

I skimmed this paper and it seems to be exceptionally bad. The proof sketch offered is that a winning strategy can be verified quickly, but finding a winning strategy requires searching over 3^n possible strategies.

But this assumes that brute force is the only possible way to find a winning strategy, and thus proves far too much. A similar argument would prove that sorting is NP hard, if you start by the assumption that the only way to sort data is by trying every possible permutation.

I may be wrong, but I'm pretty sure any time you claim to have a proof that a problem is NP complete, but your proof doesn't include a reduction, you're doing it wrong.

(The paper does offer what it calls a reduction to 3-sat , but it's completely hand-wavy, and I can't even understand the intuition behind it at all.)

Re: Why is it hard to buy things that work well?

#194

Post by TOMDM »

klodolph wrote: post #118
Strilanc wrote: post #88

It's pretty depressing how many of the comments are purely answering the title and not discussing the content of the article. In particular, the comments seem mostly focused on consumer product build quality whereas the article is more focused on e.g. businesses trying to outsource something (like same day delivery) and getting a worse result even when they pay more than doing it internally. And the cultural obstacle…

> …he'll use the same tone for something he's seen anecdotally and something he's spent a month personally investigating. A.k.a. “avoiding weasel words”. I don’t mean to be clever here, it’s just that it’s far more efficient to assume that reader can make their own informed decision, and give them citations & instructions for reproducibility to help them along the way, rather than shovel a bunch of weasel words into…

I agree some people hide behind uncertainty and "weasel" in order to avoid accountability for anything they say.

What that requires though is they caveat everything.

I really respect people who say they are 99% sure and 60% sure when appropriate.

It requires a level of honesty not just to the audience but to the self to admit openly what you are and aren't sure about.

Re: Why is it hard to buy things that work well?

#195

Post by klodolph »

alecbz wrote: post #170
klodolph wrote: post #162

Earlier quoted context omitted.

> IMO It serves a useful purpose! I have to disagree! I simply do not understand what useful purpose it serves—IF we are allowed to assume that our readers have a proficient level of prose literacy. The default assumption, when someone is talking or writing, is that they are communicating their own opinions and thoughts, experiences, anecdotes. I don’t know about you, but when I think back to taking writing classes i…

People hold beliefs of different levels of confidence. I have some opinions I’m very confident in and others that are merely vague hunches. My level of confidence in something is the signal I’m trying to convey, not merely the fact that it is my thought at all. It can also be useful to communicate brief info about why I have a particular confidence level. “I had a friend who…” vs “I’ve seen several times…” vs “Severa…

I hedge a lot in HN comments. Uncharitably, because I’m thinking defensively about how people respond to my comments. Charitably, because it’s more accurate that way.

Long-form content suffers from hedging more that HN comments do. As a writer, you may think highly of your confidence level, and you may want to communicate the difference between a claim that you’re confident about and a claim that you are unsure about, but the right way to do that is to provide the information necessary for other people to come to the same conclusion. If that’s not possible, or it’s not germane, or you’re just busy doing something else, then you don’t do it, and that’s okay. It’s often just irrelevant for people to understand how much you believe something.

The tradeoff I see here is between clarity and precision. If you focus too much on precision, the clarity of whatever you’re trying to say suffers.

Re: Why is it hard to buy things that work well?

#196

Post by NickNameNick »

xeromal wrote: post #173
chii wrote: post #44

Earlier quoted context omitted.

an aeron chair costs about $2000 dollars (give or take). The construction is sturdy, so i expect it to last at least 10 years, if not 20 (you take care of it etc). So the amortized cost is some 100-200 dollars a year for an excellent chair. If you buy a crappy chair, i bet that the cushion starts failing after a year or so of constant sitting. So you'd probably replace it yearly, or suffer a bad chair for a few years…

As a random anecdote, I own a steelcase gesture that runs about $1,100. I also own a 70$ chair from costco and I can "barely" feel the difference. I have all my settings configured correctly. I really don't buy into the the expensive officechair hype. I think it's a level of diminishing returns. I think a $200 office chairs is about what anyone needs. That'll get you 90%+ of the way.

Where would you say you sit on the physical-dimensions bell curve?

I'd hope that the more expensive chair has better adjustability to suit a wider range of critical dimensions than the cheaper chair. But that may not matter for a lot of people.

Re: Why is it hard to buy things that work well?

#197

Post by tuatoru »

Strilanc wrote: post #116
AtlasBarfed wrote: post #105

Earlier quoted context omitted.

The article is yet another one about supply and demand that ignores first day of microeconomics price theory: If supply is low, demand is high, the price goes up.

In what way is the article about that? How is Apple getting a better chip by learning to do it themselves, because cpu manufacturers are trapped by misaligned benchmarks, just a trivial statement about supply and demand?

It isn't about supply and demand at all, but about another, even more basic, economic concept: incentives, and a second relatively modern economic concept, information asymmetries.[1] Over-summarising and over-simplifying:-

"Third-party service providers are incentivised to minimise their costs. If their clients cannot assess the quality of their service and/or end-users cannot reward or penalise them based on the quality of the services, then quality is poor, because it is cheaper to not provide services and later argue about it than to provide the service. Third party deficiencies/misbehaviour get so bad sometimes that there are no gains from trade and occasional losses.

"Internally, inside companies, the incentives may be to act similarly to third-parties, or they may not. This depends on "culture"."

And here the essay would be improved with an analysis of the compnents of "culture" that matter. It talks about trustworthiness of company leadership, but that is only one factor, surely.

1. Akerlof's "Market for Lemons" paper is the seminal one here.

Re: Why is it hard to buy things that work well?

#198

Post by xeromal »

NickNameNick wrote: post #196
xeromal wrote: post #173

Earlier quoted context omitted.

As a random anecdote, I own a steelcase gesture that runs about $1,100. I also own a 70$ chair from costco and I can "barely" feel the difference. I have all my settings configured correctly. I really don't buy into the the expensive officechair hype. I think it's a level of diminishing returns. I think a $200 office chairs is about what anyone needs. That'll get you 90%+ of the way.

Where would you say you sit on the physical-dimensions bell curve? I'd hope that the more expensive chair has better adjustability to suit a wider range of critical dimensions than the cheaper chair. But that may not matter for a lot of people.

My favorite part about my steelcase is the adjustable lumbar but really all I need is the seat to go up and down.

Re: Why is it hard to buy things that work well?

#199

Post by carlineng »

"Variants of this idea that I frequently hear engineers and VCs repeat involve companies being efficient and/or products being basically as good as possible because, if it were possible for them to be better, someone would've outcompeted them and done it already"

Reminds me of the old joke -- an Economist walks past a $100 bill on the sidewalk, but doesn't bother to pick it up. He turns to his friend and declares smugly: "If it were real, someone would have taken it already."

Re: Why is it hard to buy things that work well?

#200

Post by PragmaticPulp »

scaramanga wrote: post #191
PragmaticPulp wrote: post #169

Earlier quoted context omitted.

> I will note that I often wish Dan Luu was a bit less apparently-uniformly-confident in some of the statements he writes. He does back up what he says, but he'll use the same tone for something he's seen anecdotally and something he's spent a month personally investigating. Dan Luu is one of my favorite currently-active bloggers. He's obviously a smart person with a good amount of experience in specific areas. Howev…

I think that's kind of the point. Dan Luu can change his priorities to do a performative display of what interviewers expect and then Dan Luu can pass a job interview. In which case you are not hiring Dan Luu, but a different guy who has most likely internalised a different world-view and set of values. Even if you assume he can compartmentalize his interview performance skills, he will have at least devoted time to…

> I think that's kind of the point. Dan Luu can change his priorities to do a performative display of what interviewers expect and then Dan Luu can pass a job interview.

This is the problematic logic I was trying to highlight: It's written as though he's infallible. He knows the correct answers to the interview questions, but he also deduces what his interviewers are thinking and why they're wrong.

The other possibility is that maybe the interviewers know what they're doing when they decline him, even when he answers the questions correctly. Interviews are about more than just reciting the correct answers to the questions, but he only discusses them as a sort of pass/fail quiz where the candidate is supposed to guess what the interviewer wants to hear.

My perspective is likely quite different as a hiring manager. The part where he talks about Palantir walking him out the door despite correctly answering the questions as fast as they can deliver is the kind of thing that happens when the interviewers agree that someone is smart, but isn't a good fit for the team. If you do enough interviews, you eventually come across people far more brilliant and successful than yourself whom you would nevertheless not really see fitting into your company's work culture. Someone who openly boasts about dodging interview questions and debating interviewers because they think they can read the interviewer's mind (and they are wrong) fits this description. You know they'll do fantastic things somewhere, but they're not the kind of person you're looking for to fill the open position on your team.

> I don't think Dan Luu is struggling for work.

Neither do I! I never meant to imply as much. Dan posted the long Twitter thread over a period of several weeks, so it appeared in my timeline frequently. I only brought it up as an example of his writing style, not his career.

Post reply on HN