Asimov's Three Laws of Robotics
Asimov's robots had the Three Laws. What rules & ethics govern today's AI?
Science fiction novel I, Robot by Isaac Asimov is a collection of stories linked through an interview with robopsychologist Dr. Susan Calvin who recounts case studies highlighting robot development and evolution over the course of her 50-year career.
There’s the case of the little girl who grew attached to her silent but friendly robot caregiver, Robbie. On an asteroid mining facility, an advanced robot believing he was created by someone called The Master forms a robot religious cult. A ‘Brain’ builds a spaceship capable of interstellar travel endangering the lives of two engineers. Or the case of a local politician who might actually be a robot.
Each story serves to explore the development, testing, and inherent limitations of the Three Laws of Robotics which are hardwired into each positronic robot brain.
Asimov’s Three Laws of Robotics:
1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.
2. A robot must obey orders given by a human being unless it conflicts with the First Law.
3. A robot must protect it’s own existence as long as such protection won’t conflict with the First or Second Law.
The novel’s robots struggle to balance the three laws when faced with ambiguous or contradictory instruction, or various external factors, leading to abnormal behavior. In each case study, Calvin and her colleagues identify the robot’s issue by troubleshooting the cause as it relates to the three laws to figure out a solution. As robots become more advanced, we see how their evolution affects human civilization.
Ultimately, Calvin is faced with the question of why Machines governing the world economy caused slight errors. The Machines are computing devices (robots aren’t used on Earth at this point in the future) which output data for humans implement.
Calvin realizes that in accordance with the first law, the Machines are moving humanity away from conflict toward a more peaceful civilization, but that it won’t explain itself because to do so would cause harm humans (humans don’t like change).
Ever so slightly, the Machines take the reigns for humanity, perhaps for the better because we couldn’t do it ourselves. Asimov demonstrates that once humans create robots, our futures are linked and humanity’s survival hangs in the balance.
AI has become an integral part of nearly every industry here in the United States. Like in Asimov’s novel, machines could take over the world for the good of humanity, however, this presupposes that the humans who made the rules governing the robots in the first place were good, decent humans who understood the dangers and also the opportunity for humanity’s advancement That’s not to say that there won’t be failures, malfunctions, and problems to overcome, as in Calvin’s case studies.
Similar to the Three Laws, how are AI and robotics companies setting guardrails for their technology’s use? I asked Chat GPT to explain it’s ethics and guiding principles. Open AI’s Charter and Usage Policies emphasizes the company’s belief that their technology should “benefit all of humanity”:
Next, I read the rules and ethics governing Boston Dynamics robots:
OpenAI technology was recently purchased by the Pentagon. It’s not hard to imagine how this technology could be used in the near future. Boston Dynamics says it won’t weaponize it’s robots but that only means some other company will.
There’s plenty of cases where AI acted strange, like when ChatGPT answered with gibberish or when Bing told a reporter it loved him and when Google’s AI convinced a software engineer it was sentient. At times, users can even bypass a chatbot’s programming. On a more positive note, a recent study found that ChatGPT diagnosed patients with a higher success rate than doctors.
In I, Robot, the three laws provided guardrails for robots. What specific software restrictions are deployed in Boston Dynamics or ChatGPT which compel it to do no harm? Are there instances where human-imposed guardrails fall short causing the robot or AI to miscalculate in error?
When AI becomes too advanced for our own understanding, who will be accountable for it? The tech companies? The AI? The broader public?
AI and robots have the potential for failure and misuse but also for enormous positive changes in our society. I hope we’re able to recognize the difference.
Let me know your thoughts in the comments.
Thanks for reading! If you haven’t already, please subscribe for more writing. You can also Buy Me A Coffee to help support Circles In Space.
Until next time,
Keith
P.S. check out this newsletter’s AI policy.