I remember reading once, a long time ago, that a famous science fiction writer (Asimov?) predicted that robots would be designed to look like human beings. His reasoning was impeccable. Humans would more readily accept machine intelligences if they physically resembled people, and that robots would be designed so they could easily use human tools, operate human machinery and vehicles and function in facilities designed for people.

Today we know better. Intelligent devices and industrial robots are configured for efficiency of use and manufacture. No effort is made to make them resemble people or to look and perform like pre-digital machines. In fact, humans are now being forced to learn how to act in awkward and non-intuitive ways. Older devices, such as TV sets, automobiles and telephones, were usable even by people who had never used one before: if you knew how to drive a car, you could step into any car manufactured anywhere in the world and drive it off the lot. You didn’t have to read a manual (or DVD or downloaded instructions) to operate a telephone or TV set.

Granted, much of this complexity of operation is due to the fact that modern digital appliances are much more capable and flexible than their analogue precursors, they can be customized and configured in a variety of ways. Installing and managing these improvements and capabilities can now be accomplished very inexpensively by modifying software, rather than expensively retooling different models during the manufacturing process.

These changes are only natural, and represent an expected evolution in our relationship to our machines. But the result is that don’t really look like us, or act like us. Fully AI-capable robots, when they come, will not adapt to us and our ways, it will be up to us to change our behavior to meet their needs. Economic and engineering necessity will force us to adapt to the technology, not the other way round. Robby will not speak grammatical English through a tinny speaker, we will communicate through abstract codes transmitted on keyboards and graphic displays. The latter may be virtual holographic depictions, or even be communicated by voice and ear, but it will be in a language unintelligible to contemporary ears. We will adapt, and we will do so for the convenience of the machines, and of those who make and sell them to us. This is the way technology has always worked, from the first stone tools to the present. The only difference is we once only needed to learn new manual skills. Now we will have to modify the way we listen, speak and think.

Science fiction once described an automated security humanoid as a robocop capable of vetting and restraining potential criminals. That’s not what security systems
look like today. There’s nothing humanoid about them at all.

Science fiction was right about one thing after all. We are in danger of surrendering our humanity to our machines. We are becoming subservient to them, not because the machines are inherently evil and wish to enslave us, but because the limitations of machines will force us to adapt to them. After all, we are more adaptable than they are, and we have no shortage of Quislings ready to sell us out.

Machines are only the hardware manifestation of the complex administrative bureaucracy we have installed to manage our civilization. They are the mechanization of bureaucracy. This is the same tyranny the Egyptians submitted to, the price in regimentation and information management, the loss of freedom to organizational efficiency and economic necessity. And the only technologies they had were plow agriculture, copper tools, wooden boats and papyrus. But unless they were in Pharaoh’s court, Egyptians lived like social insects.