My Friends, My Founders

April 30, 2007

My co-founders are people who can tell me when something is wrong. For example, my co-founder called me today and talked about misgivings he was having about the company. He didn’t see how the company was aligning with a lifestyle that would get him the things he wanted. The beautiful thing is that he was telling me this. I’m so happy to be working with friends who are unafraid of discussing such things. Not only can I discuss my often radical (and sometimes stupid) ideas, but I can operate unafraid.

I hear things like, do you have an A-team or a B-team? Do your co-founders have the proper skill-set? I am lucky that my friends and co-founders have a very dynamic sense of competence where they move from unknown to unknown with little difficulty. But the thing I am happiest about is the sense of fearlessness we manage to generate in our culture. It would be naive to say this is all you need, but you can always acquire skills and more expertise.

An open culture is invaluable and far more difficult to create because it requires the sacrifice of pride. It takes a lot of practice to say, “No, that’s a bad idea,” to someone in the middle of their sentence. And it takes even more practice to not snap back and tell them they’re full of shit. However, when you finally achieve this state, you save so much time!

My company will be invulnerable to innovative stagnation and incumbency collapse. To do this, when the company reaches 162 people it starts giving two fully-payed-months to 27 people to go and start their own companies. This means that every two months, 27 people leave the organization and 27 come back.

Most of these groups will break off into two and three person founding groups. These people would be cut-off entirely from the management structure and strongly discouraged from communication with the central corporation. You would not be allowed to come to work.

Some of you may be familiar with Google’s policy of 80/20 time. 80% of the time you work on company projects, and 20% is reserved for your own. But since you are connected to the social structure, no matter how liberal the management, you still end up working 120% of the time if you choose to work on side projects. This problem is why the breach between the central company and the startup teams has to be severe for true innovation to occur.

The advantages to this internal form of startup is that if you fail, you still keep your job. And most of the projects will fail. But some of them will succeed and cause the company to grow. As the central group grows beyond 162, it will spawn groups based on these internal startup successes.

Of course, compensation will have to match people’s time investment, otherwise it will make more sense for people to leave the company and start their own companies. So if your internal startup succeeds, it will be absorbed by the central company and the startup team will get a reward equivalent to being acquired in the startup market.

Rule of 162: Questions

April 26, 2007

Jeff Lyndsay writes:
Just curious, what makes you believe channel capacity is related to a geometric structure?

I think that channel capacity, the idea that we’re built to only handle a certain size societal group, is related to geometric structure because of a feeling that groups grow in layers. Let’s say you have a founder and he or she is represented by a sphere. If you pack spheres of people around this founder as closely as possible, you can fit twelve people-spheres. This packing represents direct and data-rich relationships. This would mean the founder would have to manage twelve such relationships. People in the outer layer would have to manage five, four with the surrounding people on the first layer and one with the founder.

Certain research indicates that humans chunk memory into blocks of seven plus or minus two, and when I think of my daily deep interactions with people, it is with a number significantly less than twelve, and around seven plus or minus two. So I don’t see the founder continuously holding twelve deep relationships, rather as the population of the outer layer reaches eleven, the founder migrates to the outer-layer to form a system of twelve. The most efficient structures in creating a two-dimensional system boundary are triangles, and when you remove the center founder-sphere, you can nudge the spheres in the outer shell into an icosahedron, a shape composed of triangles.



As you pack more people onto the icosahedron you get a second layer which will eventually be composed of 42 spheres. Then a third layer with a population of 92. And a third with a population of 162, the number I isolate as the number of the average person’s channel capacity.

But for me to choose 162 I have to claim that the inner layers migrate out to complete the outer layer being built. I think this migration occurs because if you keep the inner layers you will have a layer (the group of 42) trapped inside two layers (the group of 12 and 92). The people in that layer would have to deal with twelve deep relationships on a regular basis, which I find improbable to sustain.

This does not explain the biological connection between physical structures in the brain, but I bet there are correlates in the brain’s neocortex that mimic or at least can be mapped as tetrahedral closest-packing structures. Some of the research going into simulating aspects of the neocortex seem to indicate that the brain aggregates data through semi-triangular hierarchies. Still, I have no definitive answer.

PS: Feel free to use the image I have created as you wish.

A company that grows haphazardly creates a debilitating management structure. I say that no amount of good managing can create efficiency in a group that has grown beyond 162 because the amount of associations people would need to have a sense of belonging and group responsibility are well out of the human channel capacity.

To remedy this, when a company reaches 162 employees, a few members of the group will split off and form another group. These people will have their own central management and be physically distanced from the first group. When this new group reaches 162, it does the same, and so on.

Note for those who read “The Tipping Point”
I pick 162 over Gladwell’s 150 because I believe channel capacity to be related to a geometric structure based on the closest packing of spheres. If you think I might be totally insane (and I might be), you’ll find it interesting to note that Gore Associates (one of Gladwell’s examples) doesn’t have an average of 150 fifty employees per facility, but 7500 employees per 45 facilities, an average of approximately 167 employees per group.

Focus, Whining, and Rest

April 24, 2007

The noise began when I drifted into the half-awake limbo right before sleep. My friend’s hamster started running in the cage next to the couch I was crashing on. I was loathe to get up and put the cage in another room, so I tried to ignore it, which rapidly turned into holding the pillow around my ears. This was to no avail, until I started focusing on the noise itself. Suddenly I drifted off, and when the sleep itself interfered with my focusing on the sound I burst back into awareness. Inevitably, the sound was persistent enough that I eventually put the hamster in the bathroom, but it turned me onto this idea of focus being able to create rest.

Adopting the idea of focus being integral to rest provides an explanation for whining. Whining seems very useless to me, but I do it. Everyone does, even though plenty try not too. But whining might be more useful than simply serving as an emotional purgative. When we bitch about things, we focus on something bugging us to the exclusion of other thoughts. By highlighting the problem and deactivating the rest of the mind, we may be improving and creating new connections with the problem. Perhaps whining is the American Zen. (I am probably betraying cultural ignorance. Observing the people I know from Asia and Europe it seems there is a reservedness about complaining.)

In Zen you focus on your thoughts in a passive way. Perhaps this does the same thing as the other instances, but on a macro-scale. I focus on the hamster noise and I fall asleep. I whine about something and it’s not such a problem. I observe my thoughts and take on a preternatural calm.

Debate 2.0

April 23, 2007

I recently had an idea for an informed-debate website. Plenty of debate goes on in the world, but many of the arguments proposed are based on faith or are engineered to create an emotional effect. This would be fine as long as a place (better: many places) existed for informed discussion where people sited their sources and the validity of those sources were questioned. The academic scientific community is one such avenue, but it has insulated itself from the general public in the form of journals whose advantage is peer-review and disadvantage is a fee for casual observation. So assuming that a public forum does not currently exist that incites both fervor and reason, here is my blueprint for a debate site.

My image of the user experience begins with arriving at the site and seeing multiple discussion headings displayed on the front-page. Perhaps there is a featured discussion, highlighted as symbolic of the ideal. I click on the most interesting subject, Global Warming. I see a list of several points. Besides each point are a minimum of three sources. The first point is that global warming is man-made phenomenon. The counterpoint is that it is a natural phenomena. In counterpoint to both is the idea that it is a combination of the two. Next to the claim that global warming is a natural phenomena are three sources. Two of the studies are sponsored by the petroleum industry and have been tagged under “conflict of interest.” A new point has been raised in defense of one of the studies, and more sources have been cited in the new points defense. Admins do not move points into the central discussion unless the person gives three unique sources in their point’s defense.

This could give rise to admin abuse. To combat this, a reputation model allows a person to place a critique against the admin, but this also opens the judge to attack over the fairness of their statements. For example, in my hypothetical user experience I see the admin has a fairly good overall appraisal. There is one rather biting commentary, but the person who said it is rated as being very unfair.

Undoubtedly there are many holes in this system. How do you define a unique source? What if the community is temporarily or permanently homogeneous, skewing debate? What if people or organizations started paying admins for favors?

No matter what form it eventually evolves into it would have to keep these basic necessities in mind.

  • Multiple points of view
  • A tendency to resist (better: transform) spamesque flame wars
  • Sited Sources
  • Discussions of validity
  • Conducive to progression

Hopefully I’ll be able to get to this project one day. It would be easy to create in Ruby on Rails.

In the last few months, I recently had the pleasure of learning Zen meditation. I misunderstood Zen greatly before this, as I saw it as an activity where you endeavored to have no thoughts. However, this is not the case. Zen is simply sitting and observing your own thoughts. Where some meditation focuses on the breath, a point, or a repeated word, the object of focus in Zen is the mental dynamic. The only rules are that you do not pursue or oppose thoughts. The emphasis is on a very raw form of observation. Throughout this introspection I have encountered numerous phenomena, but one of the most interesting was the observation of how my mind was not entirely my own.

About a month into my practice I had finally settled into sitting and entering a relaxed state. Outer stimulus was no longer a big deal and I could observe my thoughts for long lengths of time. Most sessions revealed very little coherent or meaningful thoughts. Faces and images would pop up in the middle of much longer thoughts. It was very interesting, but nothing very profound had occurred. It was during one of these normal meditations that I became heavily involved in a set of thoughts. So much so that it was interfering with my ability to observe those very thoughts. The mental activity was seizing my central nervous system as if to prepare myself for some sort of action, but then the emotion peaked and I returned to a normal sort of consciousness. This event struck me because of the amount of control these fairly innocuous thoughts had over me.

I had always thought of my mind as a unified system, but this experience stopped that assumption. It seems that there are thoughts which we have that for some reason have not integrated totally and yet have connected themselves in the right places to be able to exert control over central nervous activity. After the initial event, the heavy involvement happened repeatedly during the following sessions, so much I wondered if it was going to persist. All this time I was attempting to neither pursue these thoughts or stop them from happening. After a while, the intensity of the experiences waned and I returned to less emotionally powerful thoughts.

I’m not sure exactly what was happening during these experiences. I think perhaps Zen may be an activity that creates a certain amount of unconscious integration. That’s certainly how it appeared as the power of these near-autonomous thoughts diminished. I felt as if I had learned some mechanism for rapid integration that had not previously been there. It certainly seems to be the case as I have continued to practice this form of meditation without returning to such severely involved states.

How to Give Up

April 19, 2007

When starting a company it is important to know how to give up. A founder has to cultivate a certain amount of insulation to stimulus in order to keep going with an idea and make it successful. Starting a company is essentially an insane activity when you measure it by the standards of most people. It’s not insane to me because I enjoy a certain amount of risk, but in all start-up likelihood I will fail. However, the quality of the work I do sustains me, and of course there is always the chance that it will make me rich.

So if I have this determination, how do I let go of the proper things at the right moment, while not giving up? I cultivate a systematic view of my success. Very broadly, I am forming a company to make something which will create some change which will make me rich. By holding this view it has allowed me to get through the following problems.

First, we lost our initial third founder. He knew way more Ruby on Rails than I did at the time. I had gone through only one tutorial online and was making my way through another. I still felt pretty incompetent in the framework. So I could have given up, but I didn’t because I held this overall view of what I was doing.

Second, we didn’t get the funding we wanted. It would have been with very interesting and well-connected people. So now we were all going to have stick with some sort of part-time day job and labor in our spare time. This sucks. This makes you want to get a real job and just get payed. However, I’m making a company and this is an acceptable loss. As all things pass in time, so do new oppurtunities for money arise.

Third, our idea sucked. Bizzarely, this was the easiest one to get by. We had been working on it for quite some time. We now had something to show people. But as the image increased in resolution so did the flaws in the model. If you were looking at me as this dawned, you would have seen me typing at my computer in the garage. I sat my head on my hands and ruffled my brow for a moment. I got up, went into the other room and said the idea sucked and it wouldn’t work. And then I stated the next idea we working on. I was murdering a baby, but it turned out I didn’t love the baby enough to make me delusional, I loved the idea of the start-up more.

All of these have been fairly external issues, which is why they have been easy to deal with. The next was when my co-founders were feeling unappreciated. I live in the Bay Area and they wanted me to move down to San Luis Obispo (3 hours South) to live and hack with them. For various reasons that only became clear to me over time (perhaps another blog post), I was unable to. They interpreted this as me not wanting to share the work with them. Up till that point I did most of the work myself. But suddenly they took great offense and since I am working with good people, they expressed it to me. So we hashed things out and some of it’s still getting dealt with, but I hope it will be resolved. This is the first failure that has been directly my fault.

Interestingly, this problem has allowed me to understand how I defined the difference between a founder and an employee. A founder is driven entirely by his own initiative. That statement is obvious, but in the context of my own experience it means the other two founders should be going about their business automatically. They should rip the code out of my hands if they have to and start working on it (and it is my duty to let go). But they did not do this. Now I am sounding like I am shifting blame away from myself. It is still entirely my fault because of my failure to communicate what my expectations were. If that had been communicated it would likely have been a non-issue.

Undoubtedly, I will have more failures. There will be plenty of right and wrong things to give up on. I’m hoping that what leads to my success will be a comfort with failure.

Encoding a Meritocracy

April 18, 2007

I started off thinking I was going to revolutionize the way businesses were made. The idea was to create a dynamic share system within a company, so that people are rewarded based on their contribution. At the outermost layer of the user experience, you join a group that has some goal you like. People would divide the main goal, for example selling sprinkler heads, into subgroups, such as development/marketing/sales. Each of these sub-goals would be a percentage of the main goal, and as such are a percentage of the company equity. Within these subgroups people could set bounties for certain tasks that would be some amount of shares in the sub-goal. When the company is set to generate income, you set a release version which locks everyone’s shares so that people will be payed based on that release. You can go on contributing, but your additions will only be represented at the next release.

We planned to make money by setting up this platform where anyone could insert their ideas and other people on the web could contribute. Then we would sell business services to these groups like app hosting, logo design, shipping, etc.

However, we couldn’t figure out how to protect the groups’ assets. What would stop someone from taking all the work and selling it himself and denying the contribution of others. Or some guy in a country with loose piracy laws stealing it. We could have various circles of trust, where as you do more work, more of the code, artwork, and prototypes are revealed as the admin trusts you. The trouble is working on a project is usually an all or nothing deal. In the end, their seemed no way to ensure that all contributors got payed, and that there would be no outright theft of people’s work.

Then we discovered that there was someone doing something similar CambrianHouse. Though their overarching idea was similar, we differed in that we were more focused on creating what would probably described as a structured wiki. I wish them all the luck.

Perhaps we will create an in-house version of the software, so that our current company’s share system is dynamically adjusted. I do like the idea of encoding a meritocracy, but for the moment the resources are going elsewhere.

To begin, I’ve just dropped out of school to start a company that I hope will revolutionize the way people learn. (It began as an idea for revolutionizing how people form businesses, but I will save that for the next post.) I think that we can create a system where education follows desire.

In current institutional education we use curricula to guide education instead of following people’s whims. Having people do things they don’t want to do causes a lot of unhappiness and makes for a very poor form of knowledge. Often enough, the knowledge that comes out of it is of very low resolution. After an initial drop in knowledge after the course ends, they then rebuild and add to the set of knowledge when they reach the next complimentary course. Whereas when someone’s passion is heavily married to the subject of their recent study, their knowledge has a very high resolution, and the specifics of a subject are instantly accessible.

You might say that one has to get through the unpleasant parts to get to the interesting stuff, but that is a fallacy dependent on the old educational paradigm. The system should follow your desires. Let’s say you want to learn how to engineer on the nanoscale. Your path will lead to physics and math, but the way you get to those stepping stones should not create a disconnect between each of the subjects. A curricula is a static mechanism that enforces that disconnect, through the form of time and content gaps. The solution that follows the interest of the adventurer creates no such gaps because the quest itself is a cohesive system.