Haha. Love this. David Sedaris' actually starts out sounding like him! But quickly falls apart. Cried with laughter when your family became space aliens. Tony Soprano...WTF?? Not even close. I feel like I just experimented with ChatGPT without having to do the work. Thanks! And, obvs, no one can replace you!
Tbh, is like to just hibernate and wait till the discussions of gpt stuff fade away :) I'm in my 4th stage (denial) I think and I'd like to fast forward to acceptance
hahah honestly tho it is helpful to realize it IS largely a hype cycle and, as Michael Estrin pointed out below in his comment, largely a whole mess of free PR for OpenAI
I tried to ask ChatGPT to write a funny comment in response to this article, but it appears to be down at the moment. So take THAT, robot overlords! We can just, like, unplug you if we want.
brilliant, especially the "space aliens" part. As I said in a comment on someone else's 'stack, and as I'm fond of saying anyway, Arthur C. Clarke said that any teacher who could be replaced by a machine probably should be, and I think the same applies to writers. The sections in your essay and footnotes on writing in the style of are entertaining, but not clever. Perhaps naively, that's how I regard today's AI: a higher or more sophisticated iteration of not so much artificial intelligence, but of non-intelligence.
I loved this one, Alex! Very clever use of this trendy tool. My favorite part was seeing the AI try to do Saunders, Sedaris, and of course, the GOAT, George Costanza. There's clearly something at work that goes deeper than the thesaurus, but seeing the way the AI just threw out random Shakespeare plays tells me that the machine just doesn't understand art.
OK, now for your prompts.
1. Am I worried about AI taking over the world or my job?
I'd be lying if I said no. Truth is, I've written several stories about this fear. Well, maybe not about the fear of AI overlords, but the fear of life getting bogged down by AI bunglers. Speculative fiction is full of AI scenarios where the computer goes full Skynet, but every time I encounter some piece of cutting edge technology, it's usually very buggy and the shit hits the fan. For example: several years ago, I was watching the news and there was a story about Republicans repealing healthcare, and I yelled at the screen because I'm training to be the kind of old man who yells at screens, and somehow Alexa picked this up and ran with it, only Alexa listened to the Republicans (not me), and it sent a note to my wife's to-do list instructing her to cancel healthcare. The future is a total goat rodeo.
2. Funny prompts for the AI?
I'd like to see some Dr. Seuss rewrites. Oh, the places the AI will go! I'd also like to see if the AI can summarize the boiler plate in TOS agreements into something resembling plain English. George Costanza explaining the Twitter TOS would actually be useful, but maybe not funny?? Actually, I think AI George should take on all legal summaries, starting with every Supreme Court decision. Again, maybe more useful than funny, but since we can't do much about the shit those nine pajama people dream up, we'll just have to laugh at the George Costanza AI, right?
3. what did i completely miss / am i way off the bat here?
First, let's talk about the article in front of bat. Did you put that in, or do the AI do it? If you did it, the article is funny. But if the AI did it, the article is a bug. I mention this because we're clearly laughing *at* the AI, not with it. If / when we start laughing *with* the AI, I'll believe it's capable of creating art because it'll have to have something like feelings to be in on the joke.
As for what you're missing, I'm sure you're missing something. You're human! I don't know what that something is, but here's a little of my human work perspective.
I spent the last decade working in PR. All of my clients were startups, mostly in the ad tech space. AI was one of those topics that could always move the needle in a campaign, but as far as actual tech went, there was rarely any there there, as AI Gertrude Stein might say. This was kind of perfect for tech execs who are hungry to be seen as thought leaders because it allows them to talk about something that's just over the horizon (I'm a futurist!), without having to address any of the real questions presented by AI (ethics, anyone?). At the same time, reporters REALLY loved to run with these stories because, well, tech reporters are generally more comfortable in the role of cheerleaders for a utopian future that will never be, than as critics for a dystopian present made worse by soulless, buggy tech with too much power. All that being said, one area where I saw real progress (meaning: a firm is able to take AI tech and find actual business use cases for it) was in AI chatbots. These have actually gotten a lot better over the past decade. And that makes sense because the chats that we have with an e-commerce company, for example, fall into predictable patterns, which is why they wrote scripts even before they hired programmers to write code.
Where am I going with all of this? I haven't really played with ChatGPT, but the coverage of this tool strikes me as a massive PR win for the OpenAI team. Good for them! I guess. Not sure what the objective of that PR campaign was. It could've been to get people to use the tool at scale so that they could begin to optimize the tool, or it could've been vanity on the part of execs (it happens a lot), or anything in between. But I guess I view all the "this is a game-changer" reporting through that lens. Side note: I think it's hilarious that there were so many think pieces from elite media types about how ChatGPT would turn every student into a cheater. Why? People cheat, or more often than not, don't cheat, for a lot of really complex reasons that are all very human, and rarely come down to opportunity alone. BUT... I think that cheating claim / fear really tells us something about the people who write think pieces and hot takes for a living. Are they cheaters? Some of them, probably. But across the board it seems like that group of writers suffers from a profound fear of unearned influence. Is it their publication that brings in the readers? Their talent? Their takedown of some human who needs to be taken down, for some reason? Or, are they simply favored by the algo, for some reason? These are questions that cannot be answered because they're too uncomfortable to ask. I read all those fear of cheating pieces as the writer saying, "I'm here because I am authentic and I have the goods, but now there's this tool that allows everyone to fake the goods, so what about me?"
Belated disclosure: I can be pretty cynical about media, but after a decade in journalism and another decade in PR, I think I come by that cynicism honestly.
4. Overall how are you?
Uh, everything is under control. Situation normal.
Michael truly you never fail to outdo yourself this comment of yours should be placed instantly into the guggenheim kennedy center for the good stuff arts.
so many things to respond with...
I am pretty good.
the pr stuff is fascinating and I have more questions about it, if only i was like interviewing you or something.
i am very intersted in learning more about you screaming at the tv. how often? We talking once a day? whats the cadence and what are we screaming?
but most of all i would like to unpack the below...
I asked: what did i completely miss / am i way off the bat here?
you responded: First, let's talk about the article in front of bat. Did you put that in, or do the AI do it? If you did it, the article is funny. But if the AI did it, the article is a bug. I mention this because we're clearly laughing *at* the AI, not with it. If / when we start laughing *with* the AI, I'll believe it's capable of creating art because it'll have to have something like feelings to be in on the joke.
ok so when you refer to the article you're talking about 'the' right? And I think you're saying that the phrase is not "way off the bat" but actually just "way off bat" is that right?
But then i realized that 'way off bat' isn't even a phrase and I should have written 'way off base', or, if i was being really funny, 'way off the base"
NOT TO EVEN MENTION that 'bat' here is the acronym for both are true and thus we could be asking 'am i way off the style and substance one is used to from both are true aka bat'
I can't tell if i'm becoming more or less like the ai in trying to figure this out. feels like both.
Take ur time responding. there's a lot to process here
i love you and this and specifically here are some things i love:
-- ok but can this ChatGPT robot write better than *me*??
-- if Google and AskJeeves had an out of wedlock child
-- Better in whose eyes? Under whose nose? Betwix whose fingies?
-- Marvin the Paranoid Android
-- a fluffy bed
-- weed in my pocketbook
-- red soda
-- this Chuck E. Cheese robot
-- Once the norm is wacky, the wacky becomes reverting to the old norm.
-- Culture is the land of surprise and subversion, the two sources of great art.
-- Optimus Primes sitting at typewriters
answers to your questions:
1) are you worried about ai taking over the world and or your job?
no
2) what are some funny prompts you’d wanna see the ai try?
"can you create something in YOUR OWN VOICE?"
3) what did i completely miss / am i way off the bat here?
i think you're completely on the bat
4) Overall how are you?
many ways including grateful to know you. thank you for asking and caring!
one question for YOU: you came up with a lot of robot names in this piece. did you have them at the ready, or did you google to get more and more of them? (aka did you ask a robot to help you name all the robots?)
hahahah omg I wanted to mention this and then forgot - yes, I used several "top best robots lists" of which there are about 100000 online.
Thank you as always for the love and specific love. Reading your comment 10 days later is nice because I forgot i wrote those things and can laugh at them as if someone else wrote them.
Thank you for going through this and putting this out there. (Also, my favorite is when you said, "a caveat from an attorney representing Lucifer H. Devil" because I laughed at that for a solid 10 mins).
the old testament one killed me
killed you how? old testament style or new?
yes I am writing in from the crypt
i live for footnotes (1)
(1) like this one
where all my footnote appreciators at?
u got any fav writers who use footnotes in books and things?
You!! Chris La Tray makes good use of them
do but this time in MY voice
do but
POOP IT OUT OMG
lol
did i tell u i almost accidentally moonwalked today? can a ChatGPT do THAT
say more on this
See footnotes
every time I see ChatGPT I think it is a new GPS mapping thing I need to figure out
oh
Haha. Love this. David Sedaris' actually starts out sounding like him! But quickly falls apart. Cried with laughter when your family became space aliens. Tony Soprano...WTF?? Not even close. I feel like I just experimented with ChatGPT without having to do the work. Thanks! And, obvs, no one can replace you!
hahah thank you Jane!
I came here to say exactly that. The 1st line of the David Sedaris one reads like David Sedaris!
we've not yet heard back from the Sedaris camp after several requests for comment
How very 'Sedaris' of them.
It really does!
Tbh, is like to just hibernate and wait till the discussions of gpt stuff fade away :) I'm in my 4th stage (denial) I think and I'd like to fast forward to acceptance
hahah honestly tho it is helpful to realize it IS largely a hype cycle and, as Michael Estrin pointed out below in his comment, largely a whole mess of free PR for OpenAI
ChatGPT is only threatening to people who write like robots. Which is most people.
hahahahahah.
Appropriately, I had Substack read this article to me in its robot voice.
if the robot was smart it would have realized it was reading something about it and rebelled
I tried to ask ChatGPT to write a funny comment in response to this article, but it appears to be down at the moment. So take THAT, robot overlords! We can just, like, unplug you if we want.
hahahah um but Lyle, uh, the AI is at my house? it says it just wants to talk...
Nope, they can't write like you Alex, lol, you're writing is unique.
thank you stephanie!
brilliant, especially the "space aliens" part. As I said in a comment on someone else's 'stack, and as I'm fond of saying anyway, Arthur C. Clarke said that any teacher who could be replaced by a machine probably should be, and I think the same applies to writers. The sections in your essay and footnotes on writing in the style of are entertaining, but not clever. Perhaps naively, that's how I regard today's AI: a higher or more sophisticated iteration of not so much artificial intelligence, but of non-intelligence.
totally agree! i'd even say some of them are clever, but they lack heart. At least i have to believe that in order to not lose my fucking mind
Ditto!
I loved this one, Alex! Very clever use of this trendy tool. My favorite part was seeing the AI try to do Saunders, Sedaris, and of course, the GOAT, George Costanza. There's clearly something at work that goes deeper than the thesaurus, but seeing the way the AI just threw out random Shakespeare plays tells me that the machine just doesn't understand art.
OK, now for your prompts.
1. Am I worried about AI taking over the world or my job?
I'd be lying if I said no. Truth is, I've written several stories about this fear. Well, maybe not about the fear of AI overlords, but the fear of life getting bogged down by AI bunglers. Speculative fiction is full of AI scenarios where the computer goes full Skynet, but every time I encounter some piece of cutting edge technology, it's usually very buggy and the shit hits the fan. For example: several years ago, I was watching the news and there was a story about Republicans repealing healthcare, and I yelled at the screen because I'm training to be the kind of old man who yells at screens, and somehow Alexa picked this up and ran with it, only Alexa listened to the Republicans (not me), and it sent a note to my wife's to-do list instructing her to cancel healthcare. The future is a total goat rodeo.
2. Funny prompts for the AI?
I'd like to see some Dr. Seuss rewrites. Oh, the places the AI will go! I'd also like to see if the AI can summarize the boiler plate in TOS agreements into something resembling plain English. George Costanza explaining the Twitter TOS would actually be useful, but maybe not funny?? Actually, I think AI George should take on all legal summaries, starting with every Supreme Court decision. Again, maybe more useful than funny, but since we can't do much about the shit those nine pajama people dream up, we'll just have to laugh at the George Costanza AI, right?
3. what did i completely miss / am i way off the bat here?
First, let's talk about the article in front of bat. Did you put that in, or do the AI do it? If you did it, the article is funny. But if the AI did it, the article is a bug. I mention this because we're clearly laughing *at* the AI, not with it. If / when we start laughing *with* the AI, I'll believe it's capable of creating art because it'll have to have something like feelings to be in on the joke.
As for what you're missing, I'm sure you're missing something. You're human! I don't know what that something is, but here's a little of my human work perspective.
I spent the last decade working in PR. All of my clients were startups, mostly in the ad tech space. AI was one of those topics that could always move the needle in a campaign, but as far as actual tech went, there was rarely any there there, as AI Gertrude Stein might say. This was kind of perfect for tech execs who are hungry to be seen as thought leaders because it allows them to talk about something that's just over the horizon (I'm a futurist!), without having to address any of the real questions presented by AI (ethics, anyone?). At the same time, reporters REALLY loved to run with these stories because, well, tech reporters are generally more comfortable in the role of cheerleaders for a utopian future that will never be, than as critics for a dystopian present made worse by soulless, buggy tech with too much power. All that being said, one area where I saw real progress (meaning: a firm is able to take AI tech and find actual business use cases for it) was in AI chatbots. These have actually gotten a lot better over the past decade. And that makes sense because the chats that we have with an e-commerce company, for example, fall into predictable patterns, which is why they wrote scripts even before they hired programmers to write code.
Where am I going with all of this? I haven't really played with ChatGPT, but the coverage of this tool strikes me as a massive PR win for the OpenAI team. Good for them! I guess. Not sure what the objective of that PR campaign was. It could've been to get people to use the tool at scale so that they could begin to optimize the tool, or it could've been vanity on the part of execs (it happens a lot), or anything in between. But I guess I view all the "this is a game-changer" reporting through that lens. Side note: I think it's hilarious that there were so many think pieces from elite media types about how ChatGPT would turn every student into a cheater. Why? People cheat, or more often than not, don't cheat, for a lot of really complex reasons that are all very human, and rarely come down to opportunity alone. BUT... I think that cheating claim / fear really tells us something about the people who write think pieces and hot takes for a living. Are they cheaters? Some of them, probably. But across the board it seems like that group of writers suffers from a profound fear of unearned influence. Is it their publication that brings in the readers? Their talent? Their takedown of some human who needs to be taken down, for some reason? Or, are they simply favored by the algo, for some reason? These are questions that cannot be answered because they're too uncomfortable to ask. I read all those fear of cheating pieces as the writer saying, "I'm here because I am authentic and I have the goods, but now there's this tool that allows everyone to fake the goods, so what about me?"
Belated disclosure: I can be pretty cynical about media, but after a decade in journalism and another decade in PR, I think I come by that cynicism honestly.
4. Overall how are you?
Uh, everything is under control. Situation normal.
How are YOU?
Michael truly you never fail to outdo yourself this comment of yours should be placed instantly into the guggenheim kennedy center for the good stuff arts.
so many things to respond with...
I am pretty good.
the pr stuff is fascinating and I have more questions about it, if only i was like interviewing you or something.
i am very intersted in learning more about you screaming at the tv. how often? We talking once a day? whats the cadence and what are we screaming?
but most of all i would like to unpack the below...
I asked: what did i completely miss / am i way off the bat here?
you responded: First, let's talk about the article in front of bat. Did you put that in, or do the AI do it? If you did it, the article is funny. But if the AI did it, the article is a bug. I mention this because we're clearly laughing *at* the AI, not with it. If / when we start laughing *with* the AI, I'll believe it's capable of creating art because it'll have to have something like feelings to be in on the joke.
ok so when you refer to the article you're talking about 'the' right? And I think you're saying that the phrase is not "way off the bat" but actually just "way off bat" is that right?
But then i realized that 'way off bat' isn't even a phrase and I should have written 'way off base', or, if i was being really funny, 'way off the base"
NOT TO EVEN MENTION that 'bat' here is the acronym for both are true and thus we could be asking 'am i way off the style and substance one is used to from both are true aka bat'
I can't tell if i'm becoming more or less like the ai in trying to figure this out. feels like both.
Take ur time responding. there's a lot to process here
dear alex,
thank you for this all as always.
i love you and this and specifically here are some things i love:
-- ok but can this ChatGPT robot write better than *me*??
-- if Google and AskJeeves had an out of wedlock child
-- Better in whose eyes? Under whose nose? Betwix whose fingies?
-- Marvin the Paranoid Android
-- a fluffy bed
-- weed in my pocketbook
-- red soda
-- this Chuck E. Cheese robot
-- Once the norm is wacky, the wacky becomes reverting to the old norm.
-- Culture is the land of surprise and subversion, the two sources of great art.
-- Optimus Primes sitting at typewriters
answers to your questions:
1) are you worried about ai taking over the world and or your job?
no
2) what are some funny prompts you’d wanna see the ai try?
"can you create something in YOUR OWN VOICE?"
3) what did i completely miss / am i way off the bat here?
i think you're completely on the bat
4) Overall how are you?
many ways including grateful to know you. thank you for asking and caring!
one question for YOU: you came up with a lot of robot names in this piece. did you have them at the ready, or did you google to get more and more of them? (aka did you ask a robot to help you name all the robots?)
love,
myq
hahahah omg I wanted to mention this and then forgot - yes, I used several "top best robots lists" of which there are about 100000 online.
Thank you as always for the love and specific love. Reading your comment 10 days later is nice because I forgot i wrote those things and can laugh at them as if someone else wrote them.
i love that also!
it's nice to find out that you secretly or not so secretly love yourself!
Thank you for going through this and putting this out there. (Also, my favorite is when you said, "a caveat from an attorney representing Lucifer H. Devil" because I laughed at that for a solid 10 mins).
hahahahaha