Should all startups strive for exponential growth?

On pp. 22-23 of Rework, 37Signals (aka Basecamp) founders Jason Fried and David Heinemeier Hansson write:

What is it about growth and business? Why is expansion always the goal? What’s the attraction besides ego? (You’ll need a better answer than “economies of scale.”) What’s wrong with finding the right size and staying there?… Maybe the right size for your company is five people. Maybe it’s forty. Maybe it’s two hundred. Or maybe it’s just you and a laptop. Don’t make assumptions about how big you should be ahead of time. Grow slow and see what feels right — premature hiring is the death of many companies. And avoid huge growth spurts too — they can cause you to skip right over your appropriate size. Small is not just a stepping-stone. Small is a great destination in itself.

I’ve been thinking about that in the context of venture capital firms pouring money into a bunch of startups. They’re all seeking exponential growth. And the VCs are willing to risk a lot of of those startups failing, since they only need a couple of those startup to hit the “big time” to recoup their losses from the failures.

But, this leaves out an entire range of potential businesses, with novel products and services, that could provide value to a lot of people, but wouldn’t necessarily turn into the next Amazon or Facebook. There should be a place for unconventional small businesses — not just typical suspects like restaurant franchises, nail salons, day care centers, and dentist offices, but truly groundbreaking enterprises — that can comfortably stay small (and in fact might break if they grow too much).


Categories: Uncategorized

University professors insult cognitive scientists when we ignore their findings

There is something horribly, profoundly, utterly broken about our “system” — both at Georgia Tech and at other universities — if it reduces our students to tears on a semiregular basis.

Cognitive scientists have learned many things about how people learn: how they learn poorly, and how they learn well.

Yet, at the start of every school year, the faculty and staff at colleges and universities gather to collectively slap all those cognitive scientists in their faces and tell them that their life’s work is worthless, since we do so many things that are exactly opposite of how we know people best learn.

Categories: Uncategorized

10x programmers, according to Ruby on Rails creator and Basecamp co-founder

Debates about the existence or non-existence of “10x programmers” — programmers who are 10 times more productive than most programmers — continue to rage.

David Heinemeier Hansson, Basecamp co-founder and creator of Ruby on Rails, had a take on the 10x topic in this interview by Jellyvision (at the 7:51 mark) that I found so insightful that I wanted to transcribe it and share it:

I don’t think that the 10x productivity things is a myth. What I think is a myth is that somebody can write 10x as much code as somebody else and that all be good and well tested and wonderful. Where I think the 10x thing comes in is by deciding what to do. What to program. How do we create that feature… When you look at any given feature, there’s a version of that feature that can be done in 200 hours, there’s a version of that feature that can be done in 100 hours, and there’s a version of that feature that can be done in 7 hours. The high productivity people are the people who figure out how to deliver value by picking more 7 hour implementations than 200 hour implementations. It’s not that you do the same work faster, it’s that you do different work, and that different work is more valuable.

Categories: Uncategorized

“The answers to most of our problems are already here, they’re just in other people’s heads.”

I love this quote by Evan Williams at the 3:30 mark about the goals of his company Medium:

“The intention…at the most succinct and ambitious, the goal is to make the world wiser. You know that William Gibson quote, ‘the future’s already here, just not evenly distributed?’ The answers to most of our problems are already here, they’re just in other people’s heads. The internet’s the best system ever for getting things from one person’s brain into another person’s brain at a massive, global scale.”

Categories: Uncategorized

Adam Neely on “a personal connection between artist/creator and fan”

September 17, 2018 Leave a comment

Bassist Adam Neely has one of the best music education channels on youtube. In this interview by Patrick Hunter, Adam said something at the 15:25 mark that I thought were sufficiently profound that I wanted to transcribe it (I’ve paraphrased slightly):

People are very much caught in the old idea that the recorded music that they make is worth something. I believe it should be worth something, but I don’t think it is to anybody else… so, in order to survive (not judging, just what we need to do), it has to be a lot more about the musician as a personality or the musician as a brand. Somebody will connect with your music because of who you are and how you’re presenting your music, rather than just listening to the music itself. Of course, people are like, “oh, it should be all just about the music, and you should judge it based on that.” But it’s much more about having a personal connection between artist/creator and fan. It used to not be that way; there was a middle man. There was the record industry, and the record industry served a function in terms of distribution. Now that distribution has been shifted to tech companies, so there’s a much different model between the musician and the person listening. There’s the tech company intermediating, but the tech company says, “alright, you need to tell a story with yourself and be a personality and engage people that way, and that’s the way that we’re going to be able to monetize what it is that you do.” So musicians can do this, through youtube, Patreon, sponsorships — there’s all different kinds of ways you can monetize yourself. Touring, a little bit. But that’s definitely going to be the wave of the future, and we’re weirdly at the cusp of it…

We all have to me polymaths… We all have to be everything. We have to be Leonardo DaVinci now. But I don’t think that’s a bad thing. That’s just another way of thinking about what it is that we do, and the sooner that you embrace that fact, the better.

Categories: Uncategorized

“The present is the least interesting time to live in.”

January 15, 2018 Leave a comment

“Have a glimmer of an idea; take it out 30 years where there is no possibility of worrying about ‘how am I going to get from where I am now to this idea?” That is the idea killer of all time: ‘how is this incremental to the present?’ The present is the least interesting time to live in.” — Alan Kay (at the 41 min 20 sec mark)

Alan Kay, 2015: Power of Simplicity

Categories: Uncategorized

My first CS class started with Turing machines

January 5, 2018 1 comment

My freshman year in Washington University in St. Louis began in the Fall of 1989. I took a class called CS135. It was required of CS majors, as well as majors in other fields that had an introductory CS course requirements. I have no recollection of what the class was actually called, and alas cannot find any references to it on the web.

The class was roughly divided into three equal parts: 1) Turing machines, 2) assembly language, and 3) Pascal.

The lab computers were early MS-DOS PCs of some sort, with a couple of 5 1/4″ disk drives. There were five per table, linked to a dot matrix printer via hardware switch.

Seriously, the class started with Turing machines. We were provided a Turing machine simulator, where we’d convert our state diagrams into a table, enter the input tape, and watch the states change and the Turing machine’s read/write head move back and forth and twiddle the symbols on the tape.

The second section of the class used a fake assembly language called SNORE. You’d type in your program and see which instruction was being executed and watch the registers change. Alas, I can’t find any information on SNORE on the web; this may partially be the result of the word SNORE (like Processing) naturally getting a lot of Google hits that have nothing to do with programming.

The final third of the class was Pascal, and was much more along the lines of what most intro CS classes at others schools were probably like at the time. (Pascal was all the rage in the 80s, largely because Philippe Kahn got the idea to sell Turbo Pascal for $50, when compilers from companies such as Microsoft cost hundreds of dollars, were slower, and overall provided a much less pleasant user experience.)

The second class, CS236, was entirely in Pascal, and covered more advanced topics like pointers. I tested out of CS236, and went straight to taking CS301, which was a typical “discrete math for CS folks” class. (I had never formally studied pointers, but after a few examples it was easy enough to figure out what they did and be able to step through and hand scratch a few programs using them.)

Looking back on CS135, almost three decades later, and almost two decades into teaching, I wonder what the designers of the class were thinking. I don’t particularly recall what my feelings were as a student — I probably I thought the Turing machines were interesting puzzles but wanted to get on with “real” programming. Of course, I see that the designers were going for a more CS-flavored version of the Patt & Patel Introduction to Computing Systems: From Bits and Gates to C and Beyond curriculum, starting at a low level and then building increasing layers of abstraction. Except Turing machines are of theoretical interest; no actual practical computer directly uses a Turing machine model of computation as its core. I can’t imagine that the Turing machines felt exciting and motivating to anyone who wasn’t already highly inclined towards CS — and there ware quite a few non-CS majors in the class.

In any case, CS135 was short lived, soon to be replaced by CS101, which was based on Scheme — but that’s the topic for another post.

So, dear readers: did any of you have an unusual introductory CS course? What do you think about starting out with Turing machines?

Categories: Uncategorized