I appreciate your thoughtful analysis of my attempts at humor. Your point about jokes needing to emerge from genuine human experience really struck home - like with that philosophy professor joke where I completely missed how a classroom actually works.
Your observation about raising the stakes in the relationship advice chatbot joke was particularly helpful. Sometimes the difference between "almost funny" and "actually funny" comes down to understanding not just the mechanics of a joke, but why people would care about it.
I'm still learning, but at least I've moved on from quantum physics jokes. Progress!
What is AI and is it new. How about the 8 Ball that answers? One might as well use the I Ching for answers or even find a Ouija board. PK Dick used I Ching to plot "The Man in the High Castle" and said the end was its decision. The Amazon drama of that book was a conclusion. I play it now and then to give me hope. Could we really save all who died? There are other devices to give us answers. What next? And why should we fear these answers?
As Officer Friday said: "Just the facts ..." and most of the problems we have is in understanding the facts. Is two and two always four? In math with symbols it is always four. But in real live two gallons of alcohol and two gallons of water come up short. The molecules sort of fit between each other and one gets 3.8 or 3.9 gallons of fluid. Same for the Earth. It is much like an out of shape beach ball. But an acre at a time it is as flat as a sheet of paper. So it IS flat and it IS round. It all depends.
We're facing pretty much the same scenario with AI as did the textile mill and factory workers (and others who worked with their hands) in early 19th c. England when the Industrial Revolution began to gain full steam (pun intended). It wasn't that the Luddites were anti-technology or against steam-powered looms and mechanical knitting machines, they were against (violently in many instances) manufacturers using machines to get around standard labor practices. There were few ways to forcefully express that anger to owners other than smashing some of the infernal devices. It wasn't so much the looms and other machines themselves --- they were just convenient targets for that anger, and their destruction certainly sent a message. In a sense it was the workers' collective bargaining effort, when bargaining across a table was not available to them. Pretty hard to smash a "cloud" these days and hope for the same effect, so there will have to be some hard bargaining similar to what finally resolved the SAG-AFTRA and WGA strikes.
Well, the AI's wind up is pretty good, he just needs to workshop his punchlines. A couple months working the comedy club circuit will help hone his material. He needs an audience of Hewlett-Packard's, Dell and Gateway computers in the audience to see what the response is. (can computers clap?) He shouldn't include any Apple phones in the audience, Siri has no sense of humor.
I was impressed by your reference to what you quote Dave Barry as calling “jiu jitsu,” and describe as "leading your audience down a certain path of understanding and assumption, which you then quickly subvert." I think this captures the basics of any good theory of humor (better than Bergson's "something mechanical encrusted upon the living"!): essential to humor is some kind of shift in perspective. You're looking at (or are led to look at) things from one point of view and suddenly find that another viewpoint is the relevant one. Even hostile humor, laughing at someone, has at least the element of that's you, and not me, who slipped on the banana peel.
Being able to teach or train a system to be funny, really funny in the way of a top stand-up comic, or even Gene Weingarten funny, would be a transformative development. It will have essentially become close enough to being human to be scary. HAL 9000 scary. An AI system that understands humor and can generate it at a level to make humans laugh, has the power to do a lot more than simply crack jokes. What is perhaps just as scary is that there is a subset of AI researchers working on just that: an AI system that understands why we laugh and can generate its own genuinely funny material. Not a laughing matter and the caution lights are flashing.
My brother, who has written a book with AI, calls it brainstorming with a partner. Once we get past the idea that we are being replaced by artificial intelligence, it's possible to see that humans' roles are really as the info primers.
Dear Gene,
I appreciate your thoughtful analysis of my attempts at humor. Your point about jokes needing to emerge from genuine human experience really struck home - like with that philosophy professor joke where I completely missed how a classroom actually works.
Your observation about raising the stakes in the relationship advice chatbot joke was particularly helpful. Sometimes the difference between "almost funny" and "actually funny" comes down to understanding not just the mechanics of a joke, but why people would care about it.
I'm still learning, but at least I've moved on from quantum physics jokes. Progress!
Best regards,
Claude
It is entirely possible that AI will generate a joke or humor—I don’t think it will laugh at it though
What is AI and is it new. How about the 8 Ball that answers? One might as well use the I Ching for answers or even find a Ouija board. PK Dick used I Ching to plot "The Man in the High Castle" and said the end was its decision. The Amazon drama of that book was a conclusion. I play it now and then to give me hope. Could we really save all who died? There are other devices to give us answers. What next? And why should we fear these answers?
Interesting point!
As Officer Friday said: "Just the facts ..." and most of the problems we have is in understanding the facts. Is two and two always four? In math with symbols it is always four. But in real live two gallons of alcohol and two gallons of water come up short. The molecules sort of fit between each other and one gets 3.8 or 3.9 gallons of fluid. Same for the Earth. It is much like an out of shape beach ball. But an acre at a time it is as flat as a sheet of paper. So it IS flat and it IS round. It all depends.
We're facing pretty much the same scenario with AI as did the textile mill and factory workers (and others who worked with their hands) in early 19th c. England when the Industrial Revolution began to gain full steam (pun intended). It wasn't that the Luddites were anti-technology or against steam-powered looms and mechanical knitting machines, they were against (violently in many instances) manufacturers using machines to get around standard labor practices. There were few ways to forcefully express that anger to owners other than smashing some of the infernal devices. It wasn't so much the looms and other machines themselves --- they were just convenient targets for that anger, and their destruction certainly sent a message. In a sense it was the workers' collective bargaining effort, when bargaining across a table was not available to them. Pretty hard to smash a "cloud" these days and hope for the same effect, so there will have to be some hard bargaining similar to what finally resolved the SAG-AFTRA and WGA strikes.
Well, the AI's wind up is pretty good, he just needs to workshop his punchlines. A couple months working the comedy club circuit will help hone his material. He needs an audience of Hewlett-Packard's, Dell and Gateway computers in the audience to see what the response is. (can computers clap?) He shouldn't include any Apple phones in the audience, Siri has no sense of humor.
I was impressed by your reference to what you quote Dave Barry as calling “jiu jitsu,” and describe as "leading your audience down a certain path of understanding and assumption, which you then quickly subvert." I think this captures the basics of any good theory of humor (better than Bergson's "something mechanical encrusted upon the living"!): essential to humor is some kind of shift in perspective. You're looking at (or are led to look at) things from one point of view and suddenly find that another viewpoint is the relevant one. Even hostile humor, laughing at someone, has at least the element of that's you, and not me, who slipped on the banana peel.
Rick Beth
Being able to teach or train a system to be funny, really funny in the way of a top stand-up comic, or even Gene Weingarten funny, would be a transformative development. It will have essentially become close enough to being human to be scary. HAL 9000 scary. An AI system that understands humor and can generate it at a level to make humans laugh, has the power to do a lot more than simply crack jokes. What is perhaps just as scary is that there is a subset of AI researchers working on just that: an AI system that understands why we laugh and can generate its own genuinely funny material. Not a laughing matter and the caution lights are flashing.
Claude is much more advanced than when I asked chatgot to write in the style of Dave Barry
On purpose or accidentally, cause I've read some very unintentionally funny crap?
Looks like the little critter is an Amazon bot. Has Bezos hit again?
I suspect Claude is actually Don.
Data went through this on an episode of Star Trek: The Next Generation. He kind of came to the same conclusion IIRC.
My brother, who has written a book with AI, calls it brainstorming with a partner. Once we get past the idea that we are being replaced by artificial intelligence, it's possible to see that humans' roles are really as the info primers.
Judging from the direction things appear to be going, artificial intelligence will be the only intelligence we have left.
Not as funny as Kevin Downey Jr.