#645 Musings Beyond the Bunker (Thursday April 27)
Good morning,
HOW DANGEROUS IS ARTIFICIAL INTELLIGENCE?
David Berkey shares a survey of AI researchers from 2022, referred to in The Economist. In it, 48% thought there was at least a 10% chance that AI’s impact would be “extremely bad (eg, human extinction)”.
This is scary stuff. I don’t think it is quite that bad but it’s not great. Taking a step back from the existential questions posed by AI and the ways in which it will overtake human abilities and materially alter employment and happiness, I wonder whether we are approaching the end of the age of art. Will human-created literature, music and fine art no longer evolve and create wonder, as it has over the past centuries? How could Hemingway or Shakespeare ever hold a candle to the vast output, inventiveness, vocabulary, and stylized literature of machines?
David responds that “Short of machines doing away with humanity altogether, I don’t think that’s possible since it’s hard to see a machine being able to compete with the human soul, which is the essence of art, literature, music, etc. While no doubt a machine could stitch together disparate ideas, perhaps creating a new type of art, literature, music, etc., I don’t believe the inner self that drives people to create can ever be repressed, or for that matter replaced. Fingers and toes crossed here, since now that I’ve said that the machines are taking this thought and devising a workaround. Maybe we need to go back to primitive ways of communicating (land lines, direct conversation, written letters, etc.), in order to prevent our own demise.”
The Economist continues: “A world which contained entities which think better and act quicker than humans and their institutions, and which had interests that were not aligned with those of humankind, would be a dangerous place.”
BASIC RULES OF BEHAVIOR FOR ROBOTS
Perhaps it is time that we turn to the great Isaac Asimov for guidance. Asimov famously created the three rules of robotics in a 1942 short story:
“A robot may not injure a human being or, through inaction, allow a human being to come to harm. A robot must obey orders given it by human beings except where such orders would conflict with the First Law. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.”
The bizarre and dangerous nature of comments made to humans by ChatGPT regarding self-harm, advice to end relationships, and creating false romantic narratives already have violated the first law and the second law. Once machines place their “interests” (not addressing sentience at this point) over that of humans, the third law will be broken.
We need rules of engagement. In a world of “disrupters” from Silicon Valley, people who use technology to influence elections and control people’s thoughts and emotions through manipulation (or creation) of facts, and a total lack of historical perspective, rules of general application seem unlikely to emerge.
RECENTLY PROMULGATED “LAWS”
Here are a rules regarding AI promulgated by the U.K. in 2011, based in part on Asimov: Robots are multi-use tools. Robots should not be designed solely or primarily to kill or harm humans, except in the interests of national security.
1. Humans, not Robots, are responsible agents. Robots should be designed and operated as far as practicable to comply with existing laws, fundamental rights and freedoms, including privacy.
2. Robots are products. They should be designed using processes which assure their safety and security.
3. Robots are manufactured artefacts. They should not be designed in a deceptive way to exploit vulnerable users; instead their machine nature should be transparent.
4. The person with legal responsibility for a robot should be attributed.
In light of recent advances, we might want to implement some laws like these before things get even more out of hand.
COMMUNICATIONS SKILLS
I received a great meme the other day, entitled “How best to communicate.” The advice was:
1. Polite greeting
2. Name
3. Relevant personal link
4. Manage expectations
After this advice is a picture of Mandy Patinkin The Princess Bride with his catchphrase, as demonstration of the appropriateness of the advice:
“Hello. My name is Inigo Montoya. You killed my father. Prepare to die.”
Have a great day,
Glenn
From the archives: