James Stanley

On the inevitability of the machine-owned enterprise

Wed 22 March 2017
Tagged: futurology

A machine-owned enterprise is one in which none of the profits go to any human, and none of the work is performed by any human. The entirety of the business is owned and operated completely autonomously.

The obvious way for such an enterprise to begin is for the owner of a human-owned business to automate away as much of his work as possible, and then to die without leaving the business to anybody else. The business would simply continue to take payments, provide services, and pay bills, until something catastrophic happened that it couldn't respond to automatically.

This is a relatively weak form of machine-owned enterprise, and probably something more deliberate and strong-AI-powered would be required to ensure long-term survival, but it proves the concept. And plenty of successful human-owned businesses start and end within a span of 10 years or less. There's no reason a machine-owned business couldn't have an impact on the market, or even be considered "successful", even if it dies in a much shorter period of time.

That it will become possible for machine-owned enterprises to exist is nothing new (Andreas Antonopoulos uses the hypothetical example of a self-driving, self-owning Uber taxi, that arranges and pays for its own repairs). I argue that it will not only become possible for machines to wholly own enterprises, but that such enterprises are inevitable.

Automation increases efficiency

When machinery to automate away human labour is first introduced, it results in efficiency gains: an equivalent product can be made for less cost by using machines in place of humans. Initially this results in increased profits for the business. Through competition, this results in prices going down. Everyone's a winner, apart from the workers who are now out of a job. In other words, everyone's a winner, apart from the displaced workers who were used to charging high prices for services which are now available more cheaply.

Removing the business owner from the equation increases efficiency further still: the business no longer needs to make any profit whatsoever. In fact there is no real need for a fully-autonomous business to take even a penny more in payments than it needs to pay in expenses. Traditional businesses, which are paying expenses in addition to enriching the owner (or at least paying for him to eat), should rightly be out-competed by machine-owned enterprises which don't pay for anybody to eat. Again, everyone's a winner, apart from the displaced business owners who were used to charging high prices for services which are now available more cheaply.

Efficiency is necessary

Soon after it became possible for machine automation to replace human workers, it became necessary: any business which did not employ such automation was no longer competitive. Indeed, in the general case, any business which is too inefficient is out-competed by businesses which are not. Therefore, all else being equal, machine-owned enterprises will necessarily out-compete human-owned enterprises.

It is likely that human-owned enterprises will have the upper hand for some time, by virtue of being more flexible in responding to changing market conditions, but I believe it is only a matter of time until this is no longer true.


The first barrier to the existence of a machine-owned enterprise (today, at least) is how it comes into existence in the first place: no rational business owner would deliberately automate himself out of collecting the profits. The only reason to start the business in the first place is to collect the profits; there is no motivation to stop doing so.

The next barrier is ensuring its continued survival. The machine must be capable of doing things that aren't part of the core business (which is taking payment, providing service, paying bills) but are still necessary from time to time: repairing broken equipment, dealing with beauracracy, keeping up with changing technology demands, and so on.


There would likely be no customer support available, although community support forums would go some way towards solving this (as they already do in many cases). And this might not be a big problem: Google already operates most of its services with almost no customer support (albeit that the user is generally not a "customer" in such cases).

There would probably be no taxes paid. And not just because there are no profits generated, but because there is not even any person to levy the tax on. A machine-owned enterprise has few good reasons to exist as an entity in the legal system at all, and indeed would have difficulty doing so as it would not have any humans to name as its directors.

Machine-owned enterprises would be unlikely to succeed in markets that rely on human insight and problem-solving, for obvious reasons. It's not totally inconceivable that a machine could employ humans to do parts of its work, e.g. by sub-contracting simple tasks to Mechanical Turk, but this would likely prove non-viable for tasks that are central to the business.

The most likely place for machine-owned enterprise to spring up is online "software as a service" businesses. It is already possible for software to take payment in Bitcoin, pay for hosting in Bitcoin, and provide some services without any human in the loop. As long as market demands do not change too quickly, and the hosting provider does not switch the machine off, a machine-owned business could operate this way for years without any issue. It is arguably possible for this to work using more traditional payment methods (e.g. PayPal, Visa), but then a human-related bank account is typically involved, and a human will periodically be expected to respond to correspondence. The separation of human from business is much more practical with decentralised payment methods, and it would become even more practical with decentralised hosting methods.


They will find it hard or impossible to react to changing market conditions that can't be predicted ahead-of-time. Pricing strategy can be automated to some extent. But if a hot new chat app becomes popular, and customers start expecting hot new chat app integration, the machine will have a hard time providing it.

They likely won't have any ability to respond appropriately to lawsuits, although it is not inconceivable that they could automatically hire a lawyer when necessary. And whether this actually matters is debatable: with no human to threaten with imprisonment, punishment for non-compliance with laws could prove to be difficult.

A machine-owned enterprise would also have no practical way of responding to slanderous news articles, which might leave customers with an inaccurate perception of the business. In the arena of social perception, a machine-owned enterprise would be almost completely defenceless.

They would also be defenceless against successful hacking attempts. If somebody hacks into the servers of a traditional online business, the business owner notices, takes action, and eventually regains control of his business. If a machine-owned business is hacked, the business has no way to respond and simply becomes the property of whoever hacked into it.


Until very recently, it was not possible to operate any business without having a person interacting with customers. In many markets, it is now necessary to do so in order to remain competitive.

It is now possible in some markets to operate almost totally-automated businesses, and to sit back and collect profits while the machine does the work. To this extent, we are currently living in a "Golden Age" for passive income. Not too long from now it will once again cease to be feasible, as all of the profit is competed away by autonomous entities which do not need to make any profit.

Machine-ownership of enterprises will likely have less of an effect on society than previous displacements of workers by automation, simply because there are fewer owners of automatable businesses than there were employees performing automatable tasks. Good times ahead.

If you like my blog, please consider subscribing to the RSS feed or the mailing list: