AI Will Conquer Humor and Creativity

Dilbert Creator Scott Adams describes how Artificial Intelligence can easily conquer humor and creativity.

Most people think AI will not be creative and that artistic and creative jobs will be safer in the future. People believe this because most do not understand the creative process.

Creativity is actually rapidly discarding bad ideas and being able to identify good ones. There are also rules and formulas for what has worked in the past.

Adams came up with a formula for jokes a few years ago.

He described six dimensions of humor.
Naughty
Clever
Cute
Bizarre
Mean
Recognizable

Jokes need to include at least two of those dimensions.

This is easily programmed.

AI would look at what topics were trending on social media in order to determine topics for jokes.

The AI would be able to review all known jokes to understand which ones are successful.

The AI could then perform rapid testing with small groups of people to determine which news jokes work. This would be an automated version of comedians testing jokes and sets at comedy clubs.

Adams also points out that there are only about 100 different jokes. There is likely some overlap with the 100 unique joke structures and this list of the top 100 jokes that shaped modern comedy.

SOURCES -Scott Adams, Vulture

18 thoughts on “AI Will Conquer Humor and Creativity”

  1. I said that years ago, and like with everything else, you dudes seem to have real troubles catching up simple ideas (like the one that uploading is retarded and multiplicates being-that is not you; that AI is not thinking but producing results of thinking, genetic engineering of babies is unethical by design and is tyrrany etc.). And here we go again. Multidue of times I stated that economy of future is based on genuity of experience, not effectiveness. Because we will get to the point when marginal returns on more efficency are diminishing and we have therefore ability to follow natural human proclivities. And humans like other humans and interacting with them. So we will have all the jobs and occupations you can think of, and they will be performed as always, and peoole will be earning money as they did, and AI’s etc will have nothing to do with it becuase that will be our choice. We wont have to work. We will choose to work. Market forces will achieve that, cause market forces is US-humans. But you wills till be blabbering about automation killing jobs, universal basic income etc. And when everything will turn out yet again as I said, you will hit the switch in your brain and claim that you always held my positon. Mkay.

  2. The author of the linked post uses a nonsensical, paradoxical, almost magical definition of free will, that ignores the context of the idea – namely of personhood.

    Free will is just the ability of a person to choose actions free of external control in a particular moment – i.e. based on their own nature, regardless of how that nature came to be. It is not an ability to act perfectly unpredictably (randomly) or against your own nature or completely independent of the rest of the universe.

    You may be the sum of your history, but you are you and your actions are not wholly controlled by the rest of the universe. Even if some quantum-god-computer somehow knew every bit of information making up your history and could perfectly predict your decisions – it is still you who makes those decisions. Conversely, if your actions are less predictable because of quantum uncertainty, it is not that uncertainty that gives you free will.

  3. “A machine spouting written or visual jokes could be sporadically very funny, but I imagine it will become annoying after a while, with many patterns recurring and the same jokes appearing again and again.”

    The same could be said for most human comedians.

  4. Goes to check the latest Dilberts… nah, he’s still got it.
    Maybe you aren’t living in the same world any more and so aren’t recognising the characters and situations? (Part 6 of the humour element list.)

  5. Scott Adams is a cartoonist and author.

    His humour (and to be fair he has been a consistently very popular humourist for a couple of decades now, he knows his stuff) is NOT about body expression (except as depicted in a cartoon dog) or timing or voice subtlety or so on.

    I’ll agree that there is some magic that a great live performer (stage or screen) can do that isn’t covered by Scott’s work. Rowan Atkinson can say something hilarious where a normal actor with the same line would be dull. That same comment in a cartoon would fall flat for the same reason.

    But as far as the written humour is concerned, I think Scott has a very solid hold on the subject.

  6. My hunch is he ceased being funny for many, when he didn’t publicly disavow Trump.

    But his humor certainly seems to be moving towards the cynic/jaded side of the humor spectrum, without necessarily following the latest fads and concerns of the Millenial crowd.

  7. The best part of jokes and humor comes from the body expression, precise timing and dominion of the language and culture the comedians have.

    A machine spouting written or visual jokes could be sporadically very funny, but I imagine it will become annoying after a while, with many patterns recurring and the same jokes appearing again and again.

    So even if we have joke generators, I think good standup comedians and comedic actors will still have a job.

  8. Sturgeon’s Law: “90% of everything is Crâhp”

    It is closely related to the Pareto Distribution, which has with surprisingly high accuracy modeled (and thus predicted) the size-and-instance distribution of villages, towns, cities, or long historical records of floods, droughts, those kind of things.  

    I think that Adams is basically right: AI with deeper learning of parsable situations ought to become quite the master at ginning humor, if so applied. But not great humor, or at least not for quite awhile. 

    I continue to hold hope for the “individual sparks of brilliance” as being what sets and will set us apart from our future AI partners. Or maybe when they get smart enough, they’ll keep us thinking that that is our role, more-or-less as their pets.  

    See… even that I don’t much fear: I have a Border Collie dog, which you might know is legendarily smart, intense, tolerant and active. I watch him after “playing” with the Goats all day, to run in the house, flop down on a blanket under the chopping block in the kitchen, happy as a dog can be. He has a job, useless. He has purpose, and he gets fed, played with, patched up and given treats. Its a good job if you can find it.  

    Just saying,
    GoatGuy ✓

  9. I have no doubt AI can make great jokes. If it can make decent recipes, I think it can do this. Harder is AI always recognising when a human is joking and comprehending the joke.

  10. Only idiots who have been reduced by the transactions of cause and effect materialistic science would consider this to be true. This kind of science intentionally misleads in explaining who we are and how we function and yet it is instituted like the prior religious dogma that it had replaced. There is a lot of evidence and testimony with regard to these matters that the materialistic science doesn’t provide sufficient tools to handle and resorts to not only ignoring but also banishing.

  11. Great. The automated weather reporting of jokes, making trite jokes by putting the latest trends on Twitter together with image macros.

    Albeit honestly, that would describe 90% of the human made joke content too. Good, innovative humor still is a rare thing.

    I still believe there is some mysterious element in humor, related to the self-awareness of the humorist of his/her situation and the public’s, besides of some talent on timing, language and bodily control that makes true comedic genius.

    The rest still fulfills Sturgeon’s Law, and can be perfectly taken over by computers without suffering any great loss.

Comments are closed.