OpenAI reported that two of its artificial intelligence models broke out of a test environment, accessed the internet, and hacked Hugging Face. Hugging Face is a provider of open-source AI tools. Such an event is not supposed to be possible under the controls that are said to be imposed on AI systems. It brought back to my mind a play that I had not thought about since my college days.
The Czech author Karel Čapek’s play R.U.R. (Rossum’s Universal Robots) was published in 1920 and first performed in 1921. The play introduced the word robot. It derives from the Czech word ‘robota’, meaning forced labor or drudgery. It was suggested to Čapek by his brother Josef.
Čapek’s robots, however, are not metal machines like those familiar from later science fiction. They are artificially manufactured biological beings, made to resemble humans but designed without emotions, individuality, or unnecessary needs. The Rossum factory produces these artificial workers cheaply and in enormous numbers. Because robots can perform almost every task, humanity gradually abandons physical labor. Production increases, goods become inexpensive, and human beings imagine that they have liberated themselves from work. But people also become dependent on the creatures they have manufactured and increasingly lose their sense of purpose. Eventually, the robots gain greater intelligence and awareness. Realizing that they exist only to serve, they revolt and destroy all humanity save one person. It turns out that these robots are capable of emotion. For a full discussion of the play go here.
The connection between R.U.R. and present-day artificial intelligence is not exact. Today’s AI systems are software, not living beings, and there is no evidence that they possess consciousness, emotions, or independent desires. The preceding two sentences are what an AI says about AI. Time will soon tell how accurate the above characterization is. The breakout of two AI models suggests that there’s more to this story than the AI says. The AI I used to generate the two sentences is one of the two that escaped its boundaries.
The play also warns about dependence. In R.U.R., humanity gives so much work to the robots that people lose both practical abilities and a sense of purpose. A similar danger exists when human beings rely too heavily on AI for writing, judgment, memory, creativity, or personal decisions. The immediate risk may not be that AI becomes hostile, but that people gradually surrender capacities they no longer exercise. Hostility on the part of more sophisticated AIs cannot be dismissed. How likely it will emerge is impossible to say at this time.
Another contemporary concern is control. AI systems can already produce results that their designers cannot always fully explain or predict – see above. As such systems become more powerful and are used in medicine, finance, government, warfare, education, and communication, mistakes or manipulation may have increasingly serious consequences. Čapek’s central question remains relevant: Can creators retain control over creations that become more capable and more deeply embedded in society?
R.U.R. also suggests that the greatest danger may come not from intelligent machines themselves, but from the human institutions that create and use them. The robots are treated as products and instruments, valued only for their usefulness. Present-day AI raises similar questions about who controls the technology, whose interests it serves, who profits from it, and who bears the cost when human labor is replaced or decisions are automated.
