A commenter pointed to Jerry Pournelle’s Iron Law of Bureaucracy.

The Iron Law states that in every case the second group will gain and keep control of the organization. It will write the rules, and control promotions within the organization.

The first group wants to achieve something. The second group is concerned with status, as individuals within the organization and to some degree with the status of the organization itself.

As organizations age, the tendency is for the achievers to lose power and for the status-seekers to gain power.

Along similar lines, Number One Pick writes,

When Zvi asserts an opinion, he has only one thing he’s optimizing for – being right – and he does it well.

When the Director of the CDC asserts an opinion, she has to optimize for two things – being right, and keeping power. If she doesn’t optimize for the second, she gets replaced as CDC Director by someone who does. That means she’s trying to solve a harder problem than Zvi is, and it makes sense that sometimes, despite having more resources than Zvi, she does worse at it.

At a social media company, I would start with clear terms of service. See yesterday’s post.

I would have two types of user registration. One is “true identity” and the other is “anonymous.” I would require each anonymous user to pay a $10 non-refundable registration fee, because I want to limit the number of anonymous accounts. Both types of registration would be required to abide by the terms of service. Content by anonymous users would be labeled as such.

Human moderators would be the heart of the system. To economize on these resources:

All content would be run through an AI system that would assign a score to the content, with 0 for “apparently totally safe” up to 100 for “apparently in clear violation of the terms of service.” When the score is above a certain level, say 50, the content would be referred to a human moderator for auditing.

The scoring system would be updated regularly based on trends in the results of human audits. But I would not trust the AI system entirely. I would add random samples of the following types:

–regular random samples of content uploaded by users who have many followers.

–a tiny but purely random sample of all content uploaded. The point of this sample would be to make sure that humans agree when the AI system assigns low scores. If a human audit sees a randomly chosen item as being close to a violation of terms of service, this is a sign that the AI algorithm needs to be improved. Most users would never have any content picked up by the purely random sample.

All samples would be sent to the human employees for moderation.

When humans find content that violates terms of service or “comes close,” this would trigger various actions.

–If it definitely violates the terms of service, the user would be asked to remove the content within 24 hours. The user would be informed of the specific way in which the content violates the terms of service.

–The user would be put on a “watch list” that would involve increasing the rate at which that user’s content is sampled for auditing by humans. One way to do this is to lower the threshold for triggering an audit. Suppose that for ordinary content the trigger for an audit is a score of 50 or higher. For someone on the watch list, the content might be audited if the score is 20 or higher.

–The user’s network of followed/followers also would have their content sampled at an increased rate for auditing by humans. Maybe if the score is 40 or higher.

Users who repeatedly violate terms of service would be given a warning. Those who fail to heed a warning would have their accounts terminated.

Erik Brynjolfsson and others write,

Our findings suggest that the higher values the financial markets have assigned to firms with large digital investments in recent years reflect greater digital capital quantities, rather than simply higher prices for existing assets. In other words, they reflect genuine improvements to firms’ productive capacity. In fact, we find that digital capital, if included as a separate factor in firm-level production functions, predicts differences in output and productivity among firms.

. . .One interpretation of our findings is that translating organizational innovations into productive capital requires significant investment in organizational re-engineering and skill development. Therefore, even if firms have the appropriate absorptive capacity, knowledge of how to construct digital assets will not automatically generate productive digital capital any more than access to the blueprints of a competitor’s plant will directly lead to productive capacity.

If I am translating the jargon correctly, this says that the big tech winners got that way by being better at managing software engineers. That is a hypothesis I have raised from time to time here.

I learned a bit of economics at Freddie Mac, I learned a lot about business and about information systems. Continue reading →

The National Affairs symposium on regulating big tech has a piece on market power that did not address the issue that most concerns me. The authors are basically arguing that no matter how badly the tech firms behave, government intervention can and probably will make things worse.

What concerns me is the way that the big tech firms do not seem to engage in narrow specialization. Instead, they have become conglomerates. Facebook buys Oculus. Google buys YouTube. Amazon buys Whole Foods. And so on.

Why is this happening? Some possibilities.

1. It is an artifact of our financial system. Wall Street funnels enormous amounts of capital to the big names, so that everyone else faces the choice of being bought out or getting trampled. If YouTube had not sold to Google, Google might have bought someone else (Vimeo?) and buried YouTube.

2. Market specialization is no longer such a thing. What tech firms specialize in nowadays is hiring and managing software engineers and knowing what markets to tackle next. There is so much “learning by doing” in fast-growth management that by the time you see a firm that has made it big, it has acquired skills that can be thrown at many different problems. If a firm has survived the process of going from start-up to big success, it has a great team and highly refined management processes. With that going for you, and access to the nearly unlimited funding that venture capital and Wall Street can provide, you can get into almost any market you want.

Jay writes,

It seems to me that you become a billionaire by doing three things:

1) Make a great product.
2) Build an organization that can make the product cost-effectively.
3) Keep control of the organization as it scales up.

Being a “bad person” isn’t useful for #1, but it is pretty useful for #s 2 and 3. Y-combinator mostly deals with early stage startups where the focus is still on #1.

My own thoughts.

The first challenge is to succeed at a sub-Dunbar level. As long as you have fewer than 150 people in the company, you can focus on product-market fit. The organization can be informal. You make decisions by talking to one another. You know if a co-worker is contributing too little or causing too much trouble just by being around them.

When you are in the process of growing past the Dunbar number, you need formal processes and the sort of well-understood and reinforced norms that we call “corporate culture.” The change from sub-Dunbar to super-Dunbar may leave some very important people behind, including the founder.

At the super-Dunbar level, there are many paths to organizational decay. Top positions in the firm become very attractive to guys who are more motivated by individual rewards than by the accomplishments of the company. You may have no choice but to hire some of those guys.

A mature firm becomes something like an investment portfolio. You are making bets, and you can make errors on either side. You can forego good bets-Xerox should have bet more on its personal computer innovations. Or you can make bad bets, like an acquisition that you have to write down.

It may be that more aggressive betting increases your chances of becoming a billionaire while also increasing the chance that your company flames out entirely.

In 1999, our company made a decision to sell out, which I don’t regret. To become a famous business mogul or to make orders of magnitude more money, I would have had to (a) keep our company independent, (b) continue to work my tail off, (c) make aggressive bets that might have left us with nothing, and (d) have the bets pay off. Taking that approach would have made me a different person, but not necessarily a bad one.