Things Paul Graham is wrong about
rondam.blogspot.com
rondam.blogspot.com
1. A couple quotes from the essay in which he claims I said efficiency doesn't matter (http://paulgraham.com/hundred.html):
"And yet some applications will still demand speed."
"Efficiency is important, but I don't think that's the right way to get it."
2/3. He claims I said x, then at the end concedes I actually said -x. This latter I "grudgingly conceded?" It was arguably the single biggest theme of that talk, that the super-angels might have a winning model without getting big hits.
4. I said assholes are an instance of people who invest late, not that the two are identical. (http://news.ycombinator.com/item?id=1809670)
5. I agree note-taking has benefits other than recording things. In fact I've written about the phenomenon before. I just meant that people didn't have to take notes to record what I said.
1. I may well have misunderstood or misrepresented what you've said. But it was not deliberate.
2. I concede that point #5 was unfair. It occurred to me as I was writing and watching the video, and I decided to add it on the spur of the moment. I should have thought about it more carefully. I've added an update to that effect.
The above is a stupid exaggeration. It only approximates the truth if your average return is more than 3x per year. (Ok, the 10 year example works if your expected safe ROI is 30% per year over 9 years also works, but good luck with that.)
This seems like a very common problem on the Internet, perhaps even more so than in the rest of life. Most of the HN comments I disagreed with on this thread: http://news.ycombinator.com/item?id=1729501 regarding "How Universities Work," an essay I wrote, concern misreadings.
The same thing happened regarding an essay about the movie Avatar. In fact, it was so common that I wrote another essay examining the reaction to the first: http://jseliger.com/2009/12/30/trolls-comments-and-slashdot-... .
I think one reason pg says things like, "I don't mean..." in his essays is avoid this kind of thing as much as possible. I'm a grad student in English and assign a couple of his essays to freshmen; the most common disagreements they have are with things not actually said in the essay. A favorite question I ask in class or on their journal entries: "Where do you see that?" The usual answer turns out to be, "nowhere."
"Most data structures exist because of speed. For example, many languages today have both strings and lists. Semantically, strings are more or less a subset of lists in which the elements are characters. So why do you need a separate data type? You don't, really. Strings only exist for efficiency. But it's lame to clutter up the semantics of the language with hacks to make programs run faster. Having strings in a language seems to be a case of premature optimization." [Processing strings is a huge part of today's computing landscape. There are premature optimizations to be sure, but efficient string processing isn't one of them IMHO.]
"But it was mysterious to me that the super-angels would quibble about valuations. Did they not understand that the big returns come from a few big successes, and that it therefore mattered far more which startups you picked than how much you paid for them?" [The implication here is that because picking the right companies matters more, then it doesn't make sense to "quibble" over valuations. But this is a false dichotomy. Picking the right companies matters, of course. But valuation matters as well, and questioning a ~10M valuation for a company that has no revenue is not "quibbling."]
"... jerks who give you lines like "come back to me to fill out the round." [I have no problem with advising founders to avoid wasting time on reluctant investors. I do have a problem with impugning their motives and character.]
If you really want to engage in a debate about this I will hold up my end. But to be perfectly honest, I'd just as soon see this bruhaha fade into obscurity. It has already gotten much more attention that it deserves IMHO.
More broadly, that comment comes across as symptomatic of a peculiar attitude among some startups and programmers towards investors, angels, and other "business" people.
Regardless of the intended meaning, there was no reason to use that word there. It didn't enforce or strengthen any point; putting "investors" in its place does the job just as well.
It was not that comment per se that "set me off", it's more the general trend at YC towards (how shall I put this to not fan the flames?) rhetoric that implies that there is something wrong with investors who exhibit reluctance, caution, or selectivity. It's been going on for some time now. I really started to notice it at the last Angel Day. The "asshole" comment was certainly the most egregious example of this, but if it had been an isolated incident I'm pretty sure I would have let it slide.
2. Whether programs process strings and whether they're visible in your source code are separate questions.
3. That passage explains my thinking as I was figuring out the part I explained next.
4. There is definitely a subset of opportunistic investors whose motives and character deserve to be impugned. As I said, it's a proper subset of those who invest late.
Not at all. YC gives you incredible leverage to diversify your portfolio, making it that much more likely that you would have a piece of the next Google.
> obviously the sentence about Google, as phrased, is a joke
It's clearly hyperbole, but there's a difference between hyperbole and a joke. You've said many times you believe the outcome of early stage investing is bimodal. In your Startup School talk you even made a point of saying you "did the math" that led you to this conclusion. You said you didn't even start to question this conclusion until you started writing the talk for the last startup school. And this seems to be as close as you get to retreating from that conclusion:
"So I think at least some super-angels are looking for companies that will get bought. That's the only rational explanation for focusing on getting the right valuations, instead of the right companies. And if so they'll be different to deal with than VCs. They'll be tougher on valuations, but more accomodating if you want to sell early."
This is the passage I referred to as "grudging acknowledgement." A non-grudging acknowledgement would have gone something like this: "Because angels do not operate under the same constraints that VC's do, it opens the door to exit strategies that VC's are not able to undertake because of the constraints they operate under. This has the potential to change everything, including the current bi-modal distribution of outcomes, which is caused at least in part by the fact that VC's have to hew to a schedule and can't reinvest proceeds from early exits. So you as founders might be able to provide very nice ROI's for your investors without really dramatic wins, as long as you can add value quickly and reliably." Or something like that.
> Whether programs process strings and whether they're visible in your source code are separate questions.
Of course surface syntax and internal representations are separate issues. I challenge the claim that "it's lame to clutter up the semantics of the language with hacks to make programs run faster." The semantics of the language determine, for example, whether strings are mutable, whether it's possible to take the CDR of a string, whether taking the Nth element of a string is an O(1) operation or an O(n) operation. "Cluttering up the semantics with hacks to make program run faster" could save you a lot of money, and is therefore not "lame." The chances that it will save you a lot of money is proportional to how successful you are.
> There is definitely a subset of opportunistic investors whose motives and character deserve to be impugned.
OK, but maybe you should draw the distinction a little more specifically, particularly when your audience is a bunch of aspiring first-time founders. It is very easy to interpret what you said as, "Anyone who gives you this line is [very likely to be? almost certain to be?] a jerk."
If you are under the impression that you a writing for an audience that want to understand what you mean, you can safely assume that you are talking about the realm of common scenarios. If you start feeling like you are writing for critics of this sort, you start to spell things out in a way that is, I think, worse writing.
the author clearly seems to misunderstand that PG's "rules" are simply ways to mentally lower the barriers to making something:
1. efficiency doesn't matter...as the author says, it matters later. so then get to work NOW
2. the returns on angel investing are bimodal...again, the simplification is used to motivate early stage startups to look for funding
3. valuations don't matter...the author's counterexample is ridiculous. PG's point is don't obsess if you're worth 14 mill or 16 mill, just get to work
4. investors who don't want to be first movers are assholes...i suppose this is the post the author wanted to write but didn't think it had enough content
5. you don't have to take notes...all PG was saying was: I'm posting this talk online. take notes if it helps you, you can think for yourself
The sort of operations commonly done on strings only have a small overlap with list operations, though. I'm all for orthogonality in a language when it's practical, but I'm not sure that one is a good trade-off.
Erlang does that, and aside from space issues (on 64-bit, a linked list of character code ints means 16 bytes per char, due to alignment), it's kind of awkward. While using pattern matching on individual chars is nice, in those cases it probably makes more sense to tokenize the strings into atoms/tuples via REs, and then pattern-match on those instead. Binaries can also be used as pseudo-strings, but to me it feels like two half-solutions. (And its shell's printing a list of ints as chars when they're printable 7-bit ASCII characters is a bit odd.) Having atoms (AKA symbols) helps, though.
Incidentally, Lua's strings are all immutable atoms. That's practical most of the time, but occasionally a major liability. Lua has the obvious option of just using a different string implementation via userdata when it's a problem, though, so it can make such a daring trade-off.
Of course, just saving strings as arrays with a \0 at the end is also pretty brittle.
A lot of people also complain about the performance of strings in Haskell, so this is right where the tradeoff is today. Tomorrow it won't likely be an issue.
A really smart compiler should be able to optimize most of the inefficiency away; I'd bank on compilers improving fast enough to make seemingly inefficient things cost much less.
If I'm designing a language today for the future, I'd err on the side of usefulness.
Lua does almost no optimization at compile time* (in part so it can compile data dumps rapidly) but I find it easy to predict the performance trade-offs in various implementation choices. In a roundabout way, it still ends up easy to tune.
Still, usefulness and performance are only indirectly related. I think that was PG's original point.
* Yet oddly it ends up quite a bit faster than Python, and there's also LuaJIT.
Even if it can be made configurable, adding all those knobs and dials may add a lot of complexity.
It's useful from a convenience of programming point of view. The preferred underlying implementation is now something more sophisticated than a linked list of characters. See Data.ByteString. The programmer does not need to care much, though.
Instead of "strings represented as lists", I think it is more useful to say "strings and lists implementing a common interface." For example, in Clojure, it is perfectly fine to treat a string as a sequence.
> (doseq [ch "abc"]
(println ch))
a
b
cYou know, people love to say this, but in every company I've worked at (sample size 3), past the first year or so of operations, programming language / framework speed actually was a very significant concern for the business.
This may be self-selection; scaling and performance are two of my areas of specialization. However, whenever I'm on the job market, there doesn't seem to be any shortage of companies trying to hire people to help them handle their performance and scaling problems.
THAT statement is dead wrong. Or at least, in dead opposition to my approach to learning. If someone says five things, and four are wrong but the last one teaches me something new or gives me a new perspective on something I already know...
Reading the four things cost me maybe two mintes. The last one cost me thirty seconds to read, plus another couple of minutes to savour. The total investment is maybe ten minutes and I learned something.
Of course, if I skip everyone who seems to be wrong right off the bat I'll avoid some situations where all five are wrong. But I have something to help me out with that: The post is on HN and some other nice people are voting it up. So something must be right, why don't I navigate into the cave, around the pit of vipers, past the deadfall, and onto the intellectual treasure?
And in fact, I read past your first statement and liked the rest. See how that works?
;-)
Your strategy for filtering information obviously works for you, and my guess is that if you missed something good, it'll come back to you later when someone you respects says "Hey, I know Raganwald is usually a waste of time, but did you see X?"
So... Rock on! And yes, I enjoyed reading your comment and the reply to my thoughts. Thanks for taking the time to post.
I hate it when people make this assumption because for me this is not true. When I take notes I stop comprehending what is being said because the act of transcribing becomes mechanical. I also write slow so I start to lag behind what's being said. Is there any research for this because I would like to be a test subject and prove them wrong.
I've personally found that intensely concentrating on what is said and actively engaging in thinking about it is the most effective way for me to not only remember what is said, but come up with interesting side-thoughts and insights about my own situation in the process.
Writing is just mechanical to me and prevents me from properly concentrating on what is said.
OTOH, at conferences I've been trying to close the laptop and take notes using pen and paper. I get more out of the talks this way, but it works because the information density is lower than my math course.
And I don't take a whole lot of notes, but try to jot down key points or things that might help trigger some recall later. I figure I can Google for stuff if all I have are some keywords written down, so I don't have to notate whole sentences.
Sadly, there's very little emphasis on what it means to take good notes, or how to go about doing it. I suspect that good note-takers don't even realize that there's a difference between notes and transcripts, so when they promote note-taking as a study tool they neglect to mention a critical step. You probably could become a good note-taker with some practice, and would benefit from it, but if you're not still in school I doubt it would be worth your time.
When I was forced to take notes I never actually transcribed, but in effect what I wrote down was practically transcribing because I could not keep the processing in real time to write down only the important points.
It takes a few weeks/months to develop your focus, attention span, and memory enough to be able to get near 100% short-term retention, but you start to see benefits almost immediately. And the long-term benefits to your memory and focus can't be understated.
Also, it's really interesting how many other people are expressing that they never took notes in college. I was famous for it in my college (my advisor wrote a paragraph about it in grad school recommendations), whereas I always thought it was common sense, seeing other people who spent the whole class scribbling, but didn't really understand a bit of what was actually taught.
[1]: http://gapingvoid.com/wp-content/uploads/2010/10/rackspace-1...
That said, I think pre-studying and post-studying immediately before or after class is really, really helpful - regardless whether notes are taken. The only challenge is that it takes much more time and a lot of effort. This is another clear example of the relationship between effort and reward.
This way I have a path to the new information from old information, and I'm unlikely to forget it. For me at least, atoms of information are much harder to remember than connections. And they're much less useful, too.
I disagree with PG on a number of issues, but this post completely missed the point.
Are these the valuations that they're getting from YC, or from investors post-YC? If post-YC, couldn't this be due largely to more of the companies being Ramen Profitable? If they raise money after four months rather than three, they've been growing (perhaps at an exponential rate) for 33% longer.
For the $500-800k raised that is still popular, "OK" was a $2M pre valuation and "amazing" would have been $5M.
The $10M valuations today for $500-800k (if there are any) are definitely in the "amazing" range, so it's somewhere around a 2x increase.
That could definitely be explained by a more mature set of companies, or companies attacking larger problems.
Five point list plus argumentative assertions plus picking a fight with pg? Absolute memetically engineered brilliance. Congratulations, Mr Rondam, whoever you are.
"If you wanted to be that accurate, you could prepend 'I think' to every statement in every essay written. That's implicit, except in proofs."
I agree with that one, I love taking notes at events and conferences. After a one-day conference I'll end up with a few pages of notes, and when I get home I type those into a few actionable items, bam!
In my experience PG errs toward the side of putting too much emphasis on taking notes, honestly. (This is probably because he works better when he takes notes, and he's seen a lot of YC founders forget/ignore his advice when they don't write it down.)
You're misinterpreting PG's stance here. This only applies this to VCs and super-angels, not to regular angel investors like yourself. VCs and SAs are obligated to have a much larger return on their investments in order to make the funds they raise a success. This is done via one or two big winners, so they're hoping for 1000x.
>> 3. Valuations don't matter
Similarly, this is aimed at VCs and SAs, not you. And of course you have to interpret the phrase as "Valuations don't matter [within reason]." Clearly you can think of a high enough number where the valuations should begin to matter to these people, but the point is that they're not at those levels right now.
>> I have a net negative return because of one spectacular failure (from which I have learned a lot but that's another story).
I'd love to read a blog post on that experience, and what made it different from your other successful investments.
Maybe I was just a bad note taker, but whenever I took notes in college I always ended up focusing more on writing the notes than paying attention to what was being said and thinking critically about it.
Angel investing is like betting on long shots. You're arguing with the bookie over whether the odds should be 95-to-1 or 100-to-1, when you should be worrying about whether the horse can win the race at all.
So perhaps the point was to debunk "misinterpretations" of these statements. Unfortunately done as a personal commentary.