How do you recognize an expert? | Hacker News Reader

How do you recognize an expert?

lemire.me

43 pointsibobev40 comments

For those not familiar with the author, this is Daniel Lemire, a high performance software expert. He has done lots of work about squeezing as much performance out of a CPU as possible; see his work on SIMD JSON parsing[0]. I imagine he wrote this out of frustration for the software industry's seeming inability to identify who actual experts are in the field. It doesn't take much work to see how much books like Design Patterns and Clean Code have negatively shaped the industry.

[0] https://lemire.me/blog/2021/09/25/new-release-of-the-simdjso...

I am currently finising my PhD and will start next month as a software architect. Among others I also looked into these books. What is the problem with them and what would you recommend instead?

The issue with those books is that they don't have any concrete data that affirms that they are worth following. And, in fact, that style of code is largely to blame for why modern software feels so sluggish. They often reduce performance by 10x if not 100x or worse. I like this video on the topic for reasons not to follow SOLID principles[0]. Muratori also has an excellent talk on writing APIs that are flexible and performant[1]. As for books on understanding the hardware performance, Computer Systems: a Programmer's Perspective, is the best example (not the international edition though, according to the author's recommendation).

I'm not aware of any architecture books recommended by anyone that cares about performance unfortunately. Most high performance software is written iteratively, meaning they aren't assuming a code structure from the start. Andreas Fredriksson, a lead engine programmer at Insomniac Games, has an excellent quote on how he writes high performance software[2]:

> Work backwards from hard constraints and requirements to guide your design. Don’t fall into the trap of designing from a fluffy API frontend before you’ve worked out how the thing will handle a worst case. Don’t be afraid to sketch stuff in while you’re proving out the approach.

> The value is what you learn, never the code. Hack it and then delete the code and implement “clean” or whatever you need. But never start there, it gets in the way of real engineering.

> As an industry we spend millions on APIs, documentation and abstraction wrapping a thing that isn’t very good to start with. Make the thing good first, then worry about fluff.

Casey Muratori also has written blogs about his programming style[3]. (He also runs a great course about performance at computerenhance.com). Abner Coimbre has a great article on how NASA approaches writing software[4]. Of course, there is also Mike Acton's famous CppCon talk about Data-Oriented Design[5].

The standard advice usually boils down to this: focus on the problem you have to solve, and be aware how damaging solving the wrong problem can be. It's a good idea to focus on what data your program receives and focusing on handling worst cases.

Since it is difficult to tell who is worth listening to, I suggest always investigating what actual software the person speaking has written. Those that write real time software or software that must not fail under any condition tend to speak very differently about typical industry practices for good reason.

[0] https://youtu.be/tD5NrevFtbU?si=Jkg6VKBHns32_IU_

[1] https://youtu.be/ZQ5_u8Lgvyk?si=tMuPFxKbrboKrBFr

[2] https://twitter.com/deplinenoise/status/1782133063725826545

[3] https://caseymuratori.com/blog_0015

[4] https://www.codementor.io/@abnercoimbre/a-look-into-nasa-s-c...

[5] https://youtu.be/rX0ItVEVjHc?si=buLbaqoc3Zugfwr7

This lecture is a decent criticism of "clean code" and other non-expert advice.

https://youtu.be/7YpFGkG-u1w

"Those who speak do not know, those who know do not speak"

(I say this as a frequent blogger / speaker, so probably says more about me than I'm comfortable admitting)

But in all honesty, many are taken in by the confidence man - the loud person casting easy declarations. But we discount the expertise of the cautious person who admits their ignorance and limitations - despite extensive experience.

I've always felt this was a tad reductive, since there are plenty of people who are both very talkative online/good at marketing themselves that are also experts in their subject areas.

Hence the best explanation would be more like "self promotion/marketing is a skill in of itself, and those that are good at it aren't necessarily good at what they're selling".

So I guess the best strategy is to not take good salesmanship as an indication of anything other than the person knowing how to sell stuff.

“Those who speak do not know, those who know do not speak“

Speak in this context means “tell you I am good at X”

If I have to tell you I’m funny, I’m probably not funny but if I demonstrate it by making you laugh - then probably funny.

Amount of actual speaking around X isn’t super correlated with how good you are at X. Eg marketing X isn’t the same as being good at X.

Okay that's a fair point. Anyone who boasts about how good they are at something should probably be treated skeptically to say the least.

Kinda like how any politician that says they're some pillar of morality turns out to be anything but that.

Purely providing advice about something or marketing your skills isn't quite the same thing.

> Those who know do not speak

I'm sorry you say that frequently, because I disagree. I have from time to time come across "experts" who would offer only vague explanations of what they were doing, or would shield their keyboard from your gaze. I have always learned subsequently that their knowledge was rather limited, which is why they didn't want to share it.

In my experience, the people who know the most have enough knowledge (and confidence) to share it freely.

It's definitely circumstantial. There are certainly many more people who are experts in topics who are not consulted or are left out of discussions that are filled by less experienced people who are better at dealing with human processes, bureaucracy and administration, conflict, etc. Lobbyists come to mind - they tend to be educated, intelligent, driven people who thrive under high stakes and pressure, effective at achieving outcomes but not necessarily nuanced subject matter experts (although many are, or have valuable perspectives).

At work, I hired someone to be my "lobbyist". He has no experience or background, but he trusts that I do, and I use him to get through to people that I can't communicate with. It's backfired a bit, though, because he's so easy and enjoyable to talk to, as intended, that management wants to recruit him. I've encouraged him and offered support because it's certainly within his capacities, and I'd love to stack the deck a bit.

This post treats being an expert as an unambiguously good thing. Richard Hamming's essay Experts has a bit more of a nuanced take on the topic. He talks how paradigm shifts in fields often come from outsiders, while at the same time acknowledging how increasing complexity neccessitates specialisation. I suppose Hamming's essay focuses on scientific and social gain, whereas this post focuses on personal gain.

This for mentioning this, I found an audio of the talk at https://www.youtube.com/watch?v=KWNBzAgAiMc

Update I know see that this is taken from Chapter 26 "Experts" in Hamming's "Art and Science of doing engineering"

> paradigm shifts often come from outsiders

There were two execs on a plane. "My company makes hard drives. We polish metal plates and try to put a thin metal coating on them, but our tolerances are mediocre." "Oh? My company makes telescope mirrors. We polish glass optically flat and lay down coatings measured in nanometers." "Oh, man, we've got to talk some more." Boom, Seagate.

> showing a total lack of practical skill is a status signal.

In some circles, maybe. Perhaps it commands respect in academia. But a lot of high-status occupations demand practicality and experience - expertise. Law, accounting, soldiering, finance - these are high-status occupations that require expertise; you can't fake them on the basis of second-hand knowledge and abstractions.

Apart from politics, dynastic rule and being a professor, I can't think of a high-status occupation that doesn't require practical expertise. And many professors actually have practical expertise. Not all professors are abstract.

It's a mark of people who avoid difficult things, who lack curiosity, and make a habit of evading volunteering and are unable to fathom noblesse oblige.

#osspodcast:

Who are the experts - https://opensourcesecurity.io/2020/04/07/who-are-the-experts...

Experts from a world that no longer exists - https://opensourcesecurity.io/2021/11/28/episode-299-experts...

It's quite easy. You don't. By definition, you can't discern anything above your own level of expertise. If you knew what an expert was talking about, there was no point in consulting them in the first place. You can merely ask people you trust. Stereotypes also play an important role, even though they really shouldn't. A rather introverted, slightly socially awkward, blond guy with a ponytail wearing glasses and business casual clothing like a polo shirt while carrying a laptop with swag stickers from their last conference and a 3D printed keychain probably radiates more tech nerd energy than, say, your stereotypical school janitor.

I have apocalyptic experience with "experts".

I've worked on shaken baby syndrome for 8 years with hundreds of families, doctors, lawyers, scientists, [1] and the major obstacle we're facing is with the so-called "experts".

They are judicial experts, appointed by courts to indicate whether they believe a baby has been shaken or abused, on the basis of specific intracranial findings (subdural and retinal hemorrhage, mostly). They tend to be extremely affirmative in their opinions, claiming that violent abuse is "absolutely certain", is the only possible cause in the entire universe, even when there are obvious DNA-proven genetic conditions known to cause the exact same findings ("these have nothing to do with it, on the contrary, sick babies are even more abused"). That's what judges need: certainty, and experts give them exactly that. People are going to jail every day because these experts are so affirmative and certain, and everyone in the system literally worships them.

When anyone has any sort of question about SBS, they're referred to them. They are the "experts", they have 30 years of experience, they have diagnosed SBS during their entire careers. Who else would know better?

They teach their theories to the entire child protection and law enforcement professions. They write their own papers, they are editors in the relevant journals and they review all submissions — carefully rejecting every dissenting voice. They attack everyone who tries to bring a bit more nuance to the complicated issue of the medical detection of child abuse, calling them out "denialists" and "revisionists" in public.

So, I'm not at ease with this post, because these (pseudo)-experts have plenty of experience, they are the experts, and yet they are so much worse than serious specialized scientists and physicians who really know what they're talking about (but who are not judicial experts because they are too busy, well, treating patients and contributing to scientific knowledge). These experts have diagnosed SBS hundreds of times for decades, no one in the world has more experience. Yet, I can personally attest that the scientific level of their work is catastrophically bad.

Perhaps the way to conciliate things is with this: "An expert has a track record and has had to face the consequences of their work. Failing is part of what makes an expert: any expert should have stories about how things went wrong."

That's exactly what's missing with SBS and child abuse: there's no feedback loop! The experts could be wrong every single time, accusing innocent parents and caregivers every time, they would never know, except that people would systematically protest their innocence — which, horrifyingly, is actually the case. It would be funny if it weren't so tragic, but protesting one's innocence in these cases is actually so frequent that it has become one of the diagnostic criteria.

Now, they're not wrong every time, but I believe they're more often wrong than right.

[1] https://news.ycombinator.com/item?id=37650402

PS: for the anecdote, one of these experts, the top forensic pediatrician in the country, a "child protection champion", is being prosecuted now because a patient has accused her of raping her.

IIRC a TED talk concluded SBS was a now mostly discredited pseudo-scientific moral panic of the 80's and 90's. There are certain pathos microtransgressions in Western civilization that are equatable to Holocaust denial, and one of them is anything that can be (mis)construed as "child abuse". But instead of identifying and prosecuting real child abuse, it becomes a lazy all-purpose "hammer" to find "nails" to accrue statistical "wins" at the toll of unreasonably over-criminalizing nothingburger situations while ruining lives.

PS: Forgive me viewing the child abuse "identification" system as inept and often wrong. I recall disclosing to a teacher how my father was an abusive narcissist and horribly cruel to my mother and myself in ways that didn't leave physical marks, yet nothing was done. Around age 16, I called the police on him because he held me down to my bed and choked me to unconsciousness for not mowing the lawn, but he talked his way out of it.

https://www.ted.com/talks/waney_squier_i_believed_in_shaken_... (Waney Squier, 2017)

Yes, courts are known to invite forensic experts of fields that boil down to unscientific fraud.[1] But I must say your experience with experts is limited to an entire field of con artists.

[1]: https://www.theatlantic.com/ideas/archive/2024/03/courts-jun...

If the person says "I'm an expert" my skepticism kicks up considerably.

https://www.youtube.com/watch?v=sR3f95BGIiA

Interesting in light of this posting’s title; I expected it to be the same thing before double checking, since this was a recent posting (that no one had much to say on):

https://news.ycombinator.com/item?id=40070432

If someone had, say, 10-20 years of solid experience with one thing, would that make you less suspicious?

Often whenever someone quotes how many years/decades of experience they have, I then begin to wonder if they have become complacent and/or adverse to learning, improvements, and change.

That's a lot of negative shit to be assuming based on a cursory comment. If someone was averse to change, they probably couldn't be bothered to stick with anything because just about everything changes. I know I'd rather seek advice from an experienced specialist rather than a rando person with some limited experience.

Except in situations like a court hearing where someone is asked whether they are an expert in something and they say yes, most of the quality people I've run into would take a more modest approach and talk about the subject and let you decide whether you accept their opinion. It's also an acknowledgement that no matter how much you know about any topic there's usually even more to learn. So outright stating that you are are an expert is often posturing.

I think excessive modesty is kind of pointless. People who are looking for an expert in a thing often don't know what constitutes an expert in the thing in the first place. You should not have to recite your whole resume to be able to call yourself an expert. The point of having the word at all is to efficiently communicate an opinion about a person (even yourself). When does it become ok to call yourself an expert? 20 years? 5 books under your belt? The most authoritative book under your belt? What?

I get that there is always more to learn, but there's nothing wrong with being real with people about what you're good at. Maybe people hate experts or are quick to dismiss anyone who dares to call themselves one for any reason they can think of, to disqualify them and inflate their own egos. But I think that's all counterproductive. If you spent like 20 years doing a thing, you better be at least somewhat expert at it lol.

An expert is somebody who knows more than me.

If I know more than him, he can't be an expert.

Income security 101. Great article.

Save you some time, this post never answers the question in the title.

I was wondering what the answer would be, because it's a hard problem. There was only the half answer of "they have experience".

The best answer I heard is that only domain experts have a chance to recognize each other. Other than that you are left with secondary signals like other people paying them for their expertise.

2nd sentence: An expert is someone who has repeatedly solved the concrete problem you are encountering.

EDIT: A kiwi once asked me "what do you do for a biscuit?" for which I required the translation: "and what might your area of expertise be?"

s/encountering/& faster, cheaper, and/or better than you or most people/

PS: I would ask: "What kind of biscuits? Do they have low glycemic and high satiety indexes?"

What? It gives at least a partial answer in the first paragraph:

> An expert has a track record and has had to face the consequences of their work. Failing is part of what makes an expert: any expert should have stories about how things went wrong.

This might sound so obvious as to be a non-answer, but I think it's a good point. There are many "experts" who acquired degrees in, wrote papers on, and now teach others about their area of focus, but have at no point in that process had to, say, stake their employment on being correct about that area.

For example, professors of literature have all written thousands of pages of text about good novels, but there's little evidence that they can actually make good novels.

Writing novels and analyzing novels are completely orthogonal skills. I bet there are many great authors who would completely suck at explaining other authors' works (or even their own), comparing different works, or explaining how literary works fit into and interact with the rest of the culture.

Just like reading code and writing are different skills, I see some of my peers perform the second without ever learning the first (at least other's people's code but I'm following your analogy), with the trigger reflex of 1- rewrite everything always and then 2- tell other colleagues to stop rewriting dammit

Second this notion.

At least in software craftsmanship, experts are best identified working with others. It appears to be an irreducible process that cannot be pantomimed with trivia-based interviews or formulaic problems. Acting in an arrogant fashion or looking smart has zero correlation with performance, but it can fool some people some of the time who lack subject matter expertise.