Hacker News new | past | comments | ask | show | jobs | submit login
Be Specific (especially during PG's office hours) (lesswrong.com)
113 points by Eliezer on Apr 4, 2012 | hide | past | web | favorite | 30 comments

It's easier to ask someone else to be specific, though, than it is to be specific oneself. Sometimes people have something real, but it's instinctive and they haven't yet sharpened or polished it. Under such conditions it's good to evoke it out of them rather than challenging them on lack of specificity.

That may not apply as much to an investor-founder conversation, because a founder should perhaps be expected to say what they mean unambiguously. But in normal conversation, I think it applies a lot.

Even in the founder situation, there are genuinely talented people who lack full articulacy as of yet. Distinguishing those from the clueless requires discernment, and maybe even kindness. I bet the earlier stage you go, the more of an issue this is.

It’s a matter of preference. I’d rather be challenged, even in everyday conversation, because I just enjoy debating and getting to the heart of an issue. The most helpful thing anyone’s done for my ideas has been to say “Shut up. Stop telling me I should care and just show me why I should care”.

It’s like a good elevator pitch: tell what you’re doing, show why it matters. Good and bad ideas get corresponding reactions—“your idea is bad and you should feel bad” versus “shut up and take my money”, if you will. But if your idea is “frighteningly ambitious”, then they’ll probably just say “yeah, right” or “how‽”

For example, I’m writing the legendary “sufficiently smart compiler”; with it, your program will run on many cores and computers as if they were one fast one.

Yeah. Right.

So yes, it’s a shame when talent goes unrecognised, but founders and funders need to meet each other halfway—somewhere around “fɔunders”, I guess. And there’s no reason you can’t be kind and firm at the same time.

This stood out to me: "I whispered audibly enough for a few nearby people to hear, 'Be specific! Be specific!'" because it also applies to writing. I'm a grad student in English lit and teach undergrads. Their writing is filled with generalities about the work and/or author; very few have the training or discipline to talk specifically about the work and to examine particular sentences. So I spend a lot of time doing that in class.

Actually, most people are like this with books or movies or other things in their lives; they'll be able to say if they liked a book, or could "relate" to it (whatever that means), but beyond that they won't have much concrete (another synonym for "specific). Which is okay, since they're not trying to be professional writers or critics. But if you are trying to a be professional x (writer, critic, startup), you'd better be willing to look at details, since details are everything.

The older I get, the more I believe details are everything. Well, maybe not quite everything, but certainly 95% of the thing.

"The details are not details, they make the product" - Charles Eames

One my favorite quotes on design. And as you pointed out, details are key to make something outstanding.

I watched that onstage office hours session on Youtube and the same thing struck me "They need to be more clear and more specific" My startup is based in an industry that I have been working in for six years so specifics are not the issue, it's condensing all the details down into just a few minutes.

I always liked Spolsky's approach to this, described in his "painless functional specs" piece, of using little stories or scenarios to illustrate typical use cases.

> Their writing is filled with generalities about the work and/or author; very few have the training or discipline to talk specifically about the work and to examine particular sentences.

It's easier to bullshit and fluff your paper with generalities than it is with concrete arguments. It also requires less critical thought. Gotta get that word count up. This is a tactic students use on purpose.

Huh? Spamming details requires no critical thought and puffs word count. A tight logical argument can suffer from low word count. Generalities replace details when the writer doesn't know details, regardless of critical thinking.

Students rephrase the sources they copy from to avoid getting busted for plagiarism. Rephrasing details you don't understand is risky because you might change their meaning to something factually incorrect. Generalities that don't say much to begin with are easier to paraphrase without running the risk of saying something original, and if you do accidentally say something original, it's less likely to be obviously wrong.

It's also easy to throw down tropes that are so general that they have a good chance of applying to the point you're supposed to be making, no critical thought required.

Like this example given by pg in http://ycombinator.com/howtoapply.html

"Information is the lifeblood of the modern organization. The ability to channel information quickly and efficiently to those who need it is critical to a company's success. A company that achieves an edge in the efficient use of information will, all other things being equal, have a significant edge over competitors."

A submission by Eliezer Yudkowsky name-dropping PG and referencing YC right in the thick of YC application season--has any submission ever been so perfectly crafted for HN before?

>> for your brain to stop thinking about an unfinished task, you must (1) know and trust that an external system will remind you to perform that task when it is time to perform it, and (2) have chosen the next action taken at a sufficiently concrete level that your brain is no longer trying to plan it out in the background.

Wow, I do this instinctively all the time, but to be frank I feared it was just a stupid obsession of mine and suspected it could be detrimental, that it made me less efficient. I'm relieved to know this is an advised method of managing your mind, even if there isn't of course much science behind it. Anecdotally, this does help with my concentration.

This is a very valuable post for people who are applying to YC. In particular, I think business people who haven't done as much implementation work are particularly susceptible to this specificity problem. I certainly was (and sometimes still am).

One concept helped me catch myself making this mistake... I read somewhere that a good explanation gives a peer the understanding required to reconstruct / implement the idea themselves.

However, some people have success describing ideas broadly in certain instances. For example, I have some friends who have had success selling "vision" to investors, even though after they pitched me I wasn't able to explain how their product worked at all.

I think the general thumb is to be very specific with seed and angel investors, and much more broad and conceptual with VCs.

why is this?

Because usually angels care more about the product and VCs care more about the market. At the angel stage the most important thing is whether you can create a product people want. But by the time you get to the VC stage it's assumed that you've already done this and have traction, so what's important is the market opportunity.

There are of course exceptions, e.g. Fred Wilson is extremely product focused, but if you're going in blind I think that's a decent heuristic.

The analytics platform discussed in the article is http://keen.io

They've since been working hard to nail their product down with more specificity. :)

Isn't this similar to what The Foundation For Critical Thinking talks about in their Thinker's Guide series?


If a person is a critical thinker then they are almost always going to be predisposed toward rationality, and if they're not a critical thinker then no amount of rationality training will ever sink in. I suppose I vaguely support the goal of trying to teach people to be more rational, but I'm not sure that starting out by trying to teach people rationality tricks is the correct approach.

Can you be specific about what it means to be a critical thinker?

Being a critical thinker means that you have the ability to judge the quality of thought, both in yourself and others. It also means that you're able to generate not just opinions but reasoned thinking.

This is much different than rationality training, which focuses on teaching people to avoid a handful of specific non-rational thought patterns that frequently occur in certain situations. But it's not about figuring out what hidden assumptions you have, learning to judge the quality of information, etc.

I've attended rationality sessions organized by the people behind the Center for Modern Rationality, and they do teach about how to find hidden assumptions, and judging quality of information.

The issues that the article covered around 'being specific' not only applies to presenting start-up idea of course. I've seen the same issue appear in resumes, during interviews and conducting performance reviews - all situations where people start talking in abstract/generic terms and fail to specifically mention their skills/achievements.

In my experience, learning how to be specific has the biggest effect on improving your resume, interviews and performance reviews.

It seems this is best summed up by the old "show don't tell" bit. As in provide concrete information, not explanation. Don't tell me you're creative, show me what you have created. Etc.

One of the only valuable takeaways I got from my encounters with Objectivism is the emphasis on concreteness when communicating ideas. Not sure how it fits into the philosophy, but I learned the technique from an Objectivist.

I find it ironic that an article titled 'Be Specific' is made up of 1802 words.

'Be Specific' is not the same as 'Be Concise'

Concision can also hurt specificity. You can express a concept very concisely when you elevate it to very abstract, general terms. Specific, concrete examples often take quite a bit longer to express. This came up a lot when I was tutoring CS: it was easy to give a precise but abstract textbook definition of something, but that’s not useful at all to a beginner—much better to say “you can use pointers to make a linked list and here’s how” than “pointers are referential types used to implement non-contiguous data structures and to reduce copying”.

Here's a tip: when on stage, don't have all four people lean forward onto their knees and huddle into themselves like frightened armadillos.

"Now explain that to me like i'm five"

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact