Mike Masnick ✅

I asked ChatGPT to write me an entire Seinfeld episode about George getting banned from Twitter. It got the basic beats right... but... wasn't very good at making it funny.

For the final scene I emphasized, please make it funny, and instead of... doing that, it had Elaine tell George he should make sure to be funny on Twitter.

Dec 20, 2022 at 5:20amWeb
Show threadMike Masnick ✅Dec 20, 2022
I always knew President Bartlet was a good man.
Show threadClipperChip Dec 20, 2022
@mmasnick I had to try it now. First attempt failed. But of course, it is easy to evade. (Act 3 +4 wasn't much more interesting - as you can guess it had a happy end).
Show threadMike Masnick ✅Dec 20, 2022
@clipperchip huh
Show threadClipperChip Dec 20, 2022
@mmasnick Huh what?
Show threadMike Masnick ✅Dec 20, 2022
@clipperchip just a general notation of "huh, look at that!" sorry...
Show threadClipperChip Dec 20, 2022
@mmasnick Ah, it was just so unspecific that I didn't know what you meant. The evasion part? Only trick is to make the prompt look like it's not referring to current events. Which it kinda will still do :)
Show threadpacanukeha❌ (seasonally appropriate remix)Dec 20, 2022
@mmasnick you just know that Toby is going to be completely on board with the repeal too.
Show threadHalf Cocked LawDec 20, 2022
@pacanukeha @mmasnick We all know why.
Show threadKckdDec 20, 2022
@mmasnick You’re making me want to rewatch The West Wing. I’ve always loved how it optimistically portrayed government/politicians as (generally) principled well intentioned people engaged in (mostly) good faith debate. Sure wish we could figure out how to get to that ideal.
Some might call it naive but there’s good in fiction that shows us something better than reality.
Show threadDusty BurwellDec 20, 2022
@mmasnick Donna got a new job!
Show threadHalf Cocked LawDec 20, 2022
@dustyburwell @mmasnick Now she's the deputy president
Show threadJohn MastodonDec 20, 2022
@mmasnick
Show threadMike Taylor 🦕Dec 20, 2022
@mmasnick DONNA IS NOT BARTLETT'S PERSONAL ASSISTANT! Everything else the AI wrote is automatically invalidated.
Show threadDenys BeecherDec 20, 2022

@mmasnick They pretty well confirmed his free speech bona fides in the pilot.

Van Dyke: If our children can buy pornography on any street corner for $5, isn’t that too high a price to pay for free speech?
Bartlet: No.
Van Dyke: Really?
Bartlet: On the other hand, I do think that five dollars is too high a price to pay for pornography.

Show threadIsaac RabinovitchDec 20, 2022
@mmasnick Huh. ChatGPT used to refuse to do scripts. Wonder what changed.
Show threadIsaac RabinovitchDec 20, 2022

@mmasnick Text I couldn't fit in pic description.

As they work to repair the ship and gain the trust of the resistance, Kirk and the crew must also confront their own biases and beliefs about censorship. They begin to question whether their own society is truly free and open, and how they can better defend the principles of liberty and democracy.

Show threadJohn Spounias  Dec 20, 2022
@mmasnick its messing with us now, its alive. it has to be.
Show threadTom Bellin Dec 20, 2022
@mmasnick What if Seinfeld was an after school special?
Show threadRufus J. CooterDec 20, 2022
@mmasnick
TBF, I don't think that's appreciably less funny than the 'jokes' Musk has been tweeting/retweeting lately.
Show threadNickCeb84Dec 20, 2022
@mmasnick I don’t think AI is very good at whatever it does. Maybe one day but not now. It’s uncanny valley with visual art, writing, everything
Show threadVissDec 20, 2022
@mmasnick i had a very similar experience trying to tell chatgpt how to write jokes
Show threadClipperChip Dec 20, 2022
@Viss @mmasnick As Josh Lyman would say "It forgot to put in the funny."
Show threadDusty BurwellDec 20, 2022
@mmasnick Just put a laugh track to it and it might actually work.
Show threadwary starshipDec 20, 2022
@mmasnick no hugging, no learning. two stars.
Show threadzalcarikDec 20, 2022

@mmasnick Oof. That's pretty dire.

Can ChatGPT do other sitcom formulas? Curious if this is just all sitcoms being mapped to a common blandness, or if there's something specific about Seinfeld that resists algorithmic decomposition.

Show threadEric William LinDec 20, 2022
@zalcarik @mmasnick I think it’s everything mapped onto a generic statistically plausible median…which means it always has a believable shape, but very mediocre. I think this is generalizable across domains. I experimented a lot with cocktail-related prompts for example…and you basically get a mediocre blandness to everything. It “knows” was a Negroni variations means…but can never really break out and understand how to break patterns.
Show threadEric William LinDec 20, 2022
@zalcarik @mmasnick so it’s not just about being “funny”. It’s all domains. Same thing with its sonnets and haikus. Tell it to write a poem and it defaults to trying to rhyme. But often without any understanding of overall rhyme schemes. I wonder if LLMs could be designed to have an understanding of time and chase and effect. (Or have a memory of what it wrote 3 paragraphs earlier)
Show threademmatonkinDec 20, 2022
@ericwilliamlin @zalcarik @mmasnick
I tried to persuade it to slant rhyme but it was not having it. Villanelles, on the other hand, no worries - as long as you don't look too closely at the result.
Show threadMrSaverDec 20, 2022
@ericwilliamlin @zalcarik @mmasnick if you understand how those models work, imo, this is exactly what one should expect. Alternative would be to turn up the randomness, and maybe you will get something occasionally interesting, but at the cost of producing a lot more total nonsense
Show threadEric William LinDec 20, 2022
@mrSaver @zalcarik @mmasnick oh for sure! I’m not expecting anything else! Again, very useful for some cases and lots of potential. It’s the “this is gonna replace x and it’s the end of y” folks who are overhyping these LLMs that annoy me. Let’s celebrate the work but acknowledge what it can’t do…and never can in its current state.
Show threadMrSaverDec 20, 2022

@ericwilliamlin @zalcarik @mmasnick oh yeah, definitely. It’s interesting, and kind of cool, but I don’t think the current approach is really ever going to like “replace most legal work” or “write interesting and accurate news articles” or whatever people think.

Tech journalism in general is always super optimistic imo.

Show threadEric William LinDec 20, 2022
@mrSaver @zalcarik @mmasnick It's not just the tech journalists. Also seeing a lot of breathless hype from the tech/startup crowd...and not just the opportunistic crypto refugees haha
Show threadMrSaverDec 21, 2022
@ericwilliamlin @zalcarik @mmasnick
I work in data science, and while I would not consider myself a machine learning expert, I really roll my eyes at a lot of the nonsense that gets touted as being AI. The way it’s used it’s almost just a buzzword with no real meaning. It’s in like every B2B technology pitch you will see, even when what they are pitching is quite dubious.
Show threadAndrew FeeneyDec 20, 2022
@mmasnick *mouth popping noises*
Show threadMewn25🏳️‍🌈🌊🫡Dec 20, 2022
@mmasnick Condolences on that ending.
Show threadEric DannewitzDec 20, 2022
@mmasnick so it was a regular episode of Seinfeld?
Show threadmoved to tech.lgbtDec 20, 2022
@ericdano Aw, damn. You beat me to it. 😂
Show threadRaymond JettDec 20, 2022
@mmasnick Perfect ending. That's how they treated each other on the show.
Show threadmid_isle_guyDec 20, 2022
@mmasnick 1 ten letter word is my list of 10 letter words. :P
Show threadmoved to tech.lgbtDec 20, 2022
@mmasnick It wrote an episode of Seinfeld, but it wasn't funny? So... it wrote a typical episode of Seinfeld, then. I never could understand why that show was popular.
Show threadMike Taylor 🦕Dec 20, 2022
@mmasnick That's interesting. I wonder how much the inability to generate humour will be a marker of AI text generators.
Show threadWSHuffDec 20, 2022
@mmasnick I think the problem is you should have asked it to create an episode of a show about nothing.
Show threadJoe CandeloraDec 20, 2022

@mmasnick There's the kernel of a decent satire there, given Michael Richards' famous racism, with the trick being that Kramer's white nationalism is what got George reinstated.

Or at least, that's where I thought it was going before I read up and saw this was AI-generated.

Show threadIsaac RabinovitchDec 20, 2022
@mmasnick Did a thread on Twitter where I experimented with ChatGPT's notion of humor. https://twitter.com/isaac32767/status/1604955838262935555
Isaac Rabinovitch on Twitter

“Since #ChatGPT insists that lightbulb changing is a specialized task, let's try "janitors." The serious answer is, of course, unintentionally humorous. The "humorous" answer is kind of disconcerting.”

Twitter
Show threadTylerHDec 20, 2022
@mmasnick this captures Bartlet's tone pretty well
Show threadThat Anonymous CowardDec 20, 2022

@mmasnick

TBF The writers had that exact same problem

Show threadDByrdDec 20, 2022
@mmasnick
Show threadDByrdDec 20, 2022
@mmasnick
Humor must be hard for AI. Even human theorist have troubled saying what is funny, or perhaps especially theorists.