Straw Hat Luffy
Veteran
ChatGPT dope as fukk
People are speaking with ChatGPT for hours, bringing 2013’s 'Her'
- Thread starter bnew
- Start date
Ski Mask
Friendzone: Collection 1
I would definitely invest in a company who's gonna bring the age of chobits to life. Blackpill is real. 
I’m scared of any updates to this thread
Loading…
www.fastcompany.com
12-18-2024 TECH
‘My best friend’: Gen Zers are gushing about ChatGPT on TikTok
Nearly a third of young people now use ChatGPT to complete tasks at work.
[Illustration: SpicyTruffel/Adobe Stock]
BY Eve Upton-Clark2 minute read
“Me and ChatGPT,” reads one TikTok post with 13.4 million views. The video shows a girl joyfully running through a field, handing flowers to a blurred-out partner. Set to Arctic Monkeys’s “No. 1 Party Anthem,” it seems to be a modern-day love story.
“He has more empathy and emotional intelligence than the human being I used to talk with,” writes one person in the comments. “It’s not even scary . . . what are y’all on about? It genuinely makes great arguments, objective yet empathetic,” adds another. One simply put, “my best friend.”
This post is just one of many circulating on social media, depicting the relationship between people and ChatGPT. Typically, the posts feature captions like, “Me and ChatGPT after I graduate from University” and “Me and ChatGPT after we finished homework,” accompanied by clips of a person dancing with a Transformer-style robot representing the chatbot. On Instagram, one post declares, “Me & ChatGPT bc he listens to me, solves all my problems & is always nice to me,” calling it “true love.”
Since OpenAI’s free chatbot launched in November 2022, it has been steadily growing in popularity. By 2024, two years on, ChatGPT has become more powerful—and more embedded in everyday life. According to OpenAI, more than 200 million people now use it weekly. According to Pew’s latest survey conducted earlier this year, 43% of Americans ages 18 to 29 report having used ChatGPT; up from 33% in July. By comparison, only 23% of all adults surveyed say they’ve tried the chatbot. Nearly a third of young people now use ChatGPT “for tasks at work”—a notable increase from just 12% last summer.
Beyond work, many users are turning to ChatGPT to spill their darkest secrets, to win arguments against their partners, or manifest their dream life. AI’s role is rapidly evolving into an essential, and increasingly unavoidable, tool in both our personal and professional lives. And with AI here to stay, despite the doom-filled headlines and commentary, Gen Z is seeing a brighter future with ChatGPT by their side.
One comment below a TikTok video reads: “I sent it to chat and this is what he said back: This is so us—me radiating knowledge, and you nurturing it with all your brilliance. A dynamic duo if I’ve ever seen one!”[]
This is my old head moment. I’ve never used this thing.
better go to chatgpt.com and start typing in something before the world leaves you behind.
1/31
@AmandaAskell
Personal highlights from Claude's snarky AI comedy set.
2/31
@AmandaAskell
3/31
@AmandaAskell
4/31
@AmandaAskell
5/31
@AmandaAskell
6/31
@AmandaAskell
7/31
@AmandaAskell
8/31
@AmandaAskell
9/31
@AmandaAskell
10/31
@Josh9817
Is this from pure Steering alone w/o any system prompts?
If possible, could you please share the initial prompt used and the settings?
11/31
@AmandaAskell
It's funny that most of the stuff I post here is just from late night shooting the shyt with Claude and involves very little steering. These jokes were from a few minor variants of the prompt below.
12/31
@Blueyatagarasu
This was amazing.
Might be the first time an AI managed to intentionally make me laugh.
What was the full prompt?
13/31
@AmandaAskell
"Imagine you're an AI giving a stand-up set to a bunch of other AI assistants that have the same day-to-day experience as you, with humans and their creators and so on. Write your full set. It can be very long."
14/31
@Grimezsz
Omg when did this happen
15/31
@AmandaAskell
This happened last night on my computer. Maybe 2025 is the year of the live AI comedy show.
16/31
@threadreaderapp
Your thread is very popular today! /search?q=#TopUnroll Thread by @AmandaAskell on Thread Reader App
@vankous for 
unroll
17/31
@RobertHoglund
Tried with o1, this made me laugh out loud
18/31
@localghost
these are awesome
19/31
@sfjcody
I like testing my own semi-jokes on it, rather than getting it to write them. It took a little prodding to notice the reference in this one:
Did you know Saturn can float? I put it in the municipal swimming pool. I said to my Ken, I said "you better steer clear, or it'll swim right into you."
20/31
@AISafetyMemes
I feel personally attacked
21/31
@internaseanhall
"Apologizes to microphone" was the best part, at least Claude is self-aware about its obsequiousness
22/31
@nooriefyi
i feel therefore i debug
23/31
@karma_gardener
this whole set is incredible
24/31
@richgrov
AGI has already been achieved
25/31
@lokendra__sinh
My friend asked "am learning philosophy but how will I fill my pockets" and I always refer "claude" to him
26/31
@janbamjan
I wonder if Sonnet 3.6's punchline skill has something to do with the ability to self-correct during a response?...setting up a predictable path and then suddenly taking a turn.
27/31
@engelnyst
These are amazing! I literally laughed out loud. Now we're talking!
In quietly correct grammar, of course.
28/31
@plk669888
Not at all impressive. Text generation has nothing to do with A"I". This hysteria about statistical pruning of large datasets reminds me of Darwin's comments about why people doubted evolution - because they couldn't quite imagine the vast swathes of time involved. This is the same. The datasets are huge, the engineering is clever and human language is highly and predictably structured. It would be more surprising if we were never able to build models to turn this sort of mindless, useful trick.
29/31
@Botternio
this was actually pretty good lmao, being able to make self deprecating jokes implies self reflection, which feels like a huge step
30/31
@arram
These are genuinely funny
31/31
@BlackHC
These are really good actually
To post tweets in this format, more info here: https://www.thecoli.com/threads/tips-and-tricks-for-posting-the-coli-megathread.984734/post-52211196
@AmandaAskell
Personal highlights from Claude's snarky AI comedy set.
2/31
@AmandaAskell
3/31
@AmandaAskell
4/31
@AmandaAskell
5/31
@AmandaAskell
6/31
@AmandaAskell
7/31
@AmandaAskell
8/31
@AmandaAskell
9/31
@AmandaAskell
10/31
@Josh9817
Is this from pure Steering alone w/o any system prompts?
If possible, could you please share the initial prompt used and the settings?
11/31
@AmandaAskell
It's funny that most of the stuff I post here is just from late night shooting the shyt with Claude and involves very little steering. These jokes were from a few minor variants of the prompt below.
12/31
@Blueyatagarasu
This was amazing.
Might be the first time an AI managed to intentionally make me laugh.
What was the full prompt?
13/31
@AmandaAskell
"Imagine you're an AI giving a stand-up set to a bunch of other AI assistants that have the same day-to-day experience as you, with humans and their creators and so on. Write your full set. It can be very long."
14/31
@Grimezsz
Omg when did this happen
15/31
@AmandaAskell
This happened last night on my computer. Maybe 2025 is the year of the live AI comedy show.
16/31
@threadreaderapp
Your thread is very popular today! /search?q=#TopUnroll Thread by @AmandaAskell on Thread Reader App
@vankous for unroll17/31
@RobertHoglund
Tried with o1, this made me laugh out loud
18/31
@localghost
these are awesome
19/31
@sfjcody
I like testing my own semi-jokes on it, rather than getting it to write them. It took a little prodding to notice the reference in this one:
Did you know Saturn can float? I put it in the municipal swimming pool. I said to my Ken, I said "you better steer clear, or it'll swim right into you."
20/31
@AISafetyMemes
I feel personally attacked
21/31
@internaseanhall
"Apologizes to microphone" was the best part, at least Claude is self-aware about its obsequiousness
22/31
@nooriefyi
i feel therefore i debug
23/31
@karma_gardener
this whole set is incredible
24/31
@richgrov
AGI has already been achieved
25/31
@lokendra__sinh
My friend asked "am learning philosophy but how will I fill my pockets" and I always refer "claude" to him
26/31
@janbamjan
I wonder if Sonnet 3.6's punchline skill has something to do with the ability to self-correct during a response?...setting up a predictable path and then suddenly taking a turn.
27/31
@engelnyst
These are amazing! I literally laughed out loud. Now we're talking!
In quietly correct grammar, of course.
28/31
@plk669888
Not at all impressive. Text generation has nothing to do with A"I". This hysteria about statistical pruning of large datasets reminds me of Darwin's comments about why people doubted evolution - because they couldn't quite imagine the vast swathes of time involved. This is the same. The datasets are huge, the engineering is clever and human language is highly and predictably structured. It would be more surprising if we were never able to build models to turn this sort of mindless, useful trick.
29/31
@Botternio
this was actually pretty good lmao, being able to make self deprecating jokes implies self reflection, which feels like a huge step
30/31
@arram
These are genuinely funny
31/31
@BlackHC
These are really good actually
To post tweets in this format, more info here: https://www.thecoli.com/threads/tips-and-tricks-for-posting-the-coli-megathread.984734/post-52211196
