Oh my god, 2000, I know that year is coming, but my brain just doesn't want to think about dates like that as being real years, even now in the 1980s it just seems so far in the future.
I remember using ChatGPT so much in 1987. I had just picked up the Joshua Tree record by U2, and ChatGPT was what I used to learn the lyrics. However, my mom was always on the phone, so I could only use offline mode. It made ChatGPT completely useless, and even hours later, I still hadn't found what I was looking for.
i fully expected the video to reveal that the online requests were being sent to an underpaid worker who eventually goes haywire and starts taking out his frustrations on the chatgpt user
In a recent electrifying interaction with GPT-4, at the vanguard of artificial intelligence, I transmitted the following query: "When did Michael Jackson pass away?" The response I received was nothing short of astounding. 😎
Apparently it said "Michael Jackson will die in June 25, 2009 due to drug overdose." Looks like I need to contact OpenAI technologies about this false information.
As a Spanish speaker, Vamos a la Playa is a song that I would listen to when thinking about nukes. It definitely talks about the beach and not nuclear war.
Here's a curious thing that was real in our universe: ELIZA, a pretty advanced chatterbot that could take user's responses, and through simple pattern-matching, it could form questions from user's answers, imitating a dialog with a psychiatrist, perhaps even being a parody. However, to the surprise of its creator, people loved ELIZA, they shared their deepest thoughts with a machine they knew is just a parrot, particularly because they knew that simple code cannot be judgemental like a person. The effect of surface-deep pattern matching done well fooling real people into thinking there's some through behind the response came to be known as the ELIZA Effect. Have we really progressed that far? LLMs are known to have some simple sort of a basic and almost non-existent "world model", but it's still more so a "language model" that just learned to generate text similar to human speech. Have we, not having learned the past, doomed ourselves to repeat it?
sureee comparing ELIZA to noways LLM's that indicate you have no idea what your no talking about. ELIZA is simple if ... then statements and comparing hat to artificial networks ...
Ah yes, another great video ONLINE. But no seriously, the series is going great! And it feels weird because I don’t know if AI existed in the past. But still great video! Also that last little part of the video.
"As an AI model based on the GPT-3.5 architecture, I don't rely on physical storage mediums like floppy disks for installation. My code and data are stored and hosted on modern digital storage devices, servers, and cloud infrastructure. However, it's worth noting that in the 80s, the size and complexity of AI models like mine had not yet been reached. The available computers had limited storage capacities and processing power compared to today's standards. Consequently, it wouldn't have been feasible to store or run an AI model like me on floppy disks. It's fascinating to consider how technology has advanced since the 80s and how far we've come. The progress in computer technology has enabled the development of complex AI models like myself, which can be operated on powerful servers and cloud platforms without the need for physical storage media like floppy disks." - Chat GPT
I'm glad these guys still make videos, for some reason they never show up for me anymore. I heard a new internet technology called ADSL?? I still use Dial up, maybe things will show up on my Commodore 64
I use that program before in the 80s , it took a minute or two to generate , my parents paid for a separate landline just four programs that required you to get a modem
Omg every detail on every video is amazing! Love this channel! Ps. The Brazilian-deutsch was an hallucination from the chatGPT, I guess it was due you were using the oooffliiineeee version hahahahahahahahahaha
You forgot about the fact that its safe. when you say something that is not okay like "Generate a killcode" Or "Tell me the missile code for something" it will say: "As a AI language model I cant generate killcodes or tell you codes for bad stuff. I exist so I can make the world better"
I asked ChatGPT what would it would be if it were basically just like in this video adn this is what it spit out: If I were from the 80s, inspired by the Squirrel Monkey video "If ChatGPT was around in the 1980s," I would be a text-based program running on MS-DOS, designed to run on personal computers of that era. Here's how I would be and the requirements for using me: Text-based interface: I would have a simple text-based interface, resembling a command prompt or a basic chat window. Users would interact with me by typing commands or questions and receiving text-based responses. MS-DOS compatibility: I would be compatible with MS-DOS operating systems, running on IBM-compatible PCs or other computers that supported the MS-DOS environment. Limited processing power: Since computers in the 80s had significantly less processing power compared to modern systems, I would have to be optimized for lower-end hardware. This means I would have a simplified language model and limited capabilities compared to the advanced AI models of today. Lower memory requirements: Memory was a scarce resource back in the 80s, so I would need to be designed to operate within tight memory constraints. This would involve using efficient algorithms and minimizing memory usage to ensure I could run on typical 80s computer configurations. Text-based database: As a program from the 80s, I would likely rely on a text-based database for storing information and knowledge. The database would consist of text files or a structured file format that could be easily parsed and accessed by the program. Limited internet connectivity: While internet access was available in the 80s, it was not as widespread or as fast as today. Therefore, I would likely have limited or no internet connectivity and rely primarily on the preloaded information and databases within the program. Emphasis on text-based content: Since graphical interfaces were not as prevalent in the 80s, I would focus on delivering text-based content, including textual descriptions, explanations, and responses to user queries. Overall, as an 80s-inspired version of ChatGPT, I would provide a basic text-based conversational experience within the limitations of the technology available during that time period.
I had it as well in college. I asked chatGPT to convert your sentence to a syntactic structure. Not sure if it is correct: [S[NP[Pronoun As][NP[Det a][N linguist][PP[P in][NP[N training][PP[P ,][NP [N 04:46]]]]]]][VP[V gave][NP[N Syntax][N class][N flashbacks]]]].
C'est aussi tellement pratique. As-tu l'application sur ton téléphone ? Hier, j'ai publié une vidéo sur mon autre chaîne, 'Squirrel Monkey Labs,' où je fais parler deux ChatGPT entre eux.
I asked him out of curiosity what was possible in the field of AI in the 1980s. "In the 1980s, artificial intelligence (AI) was still in its early stages of development. During this time, there were various approaches and technologies being applied to develop AI systems. Here are some examples of the types of AI that were possible in the 1980s: 1. Expert Systems: Expert systems were one of the most prominent applications of AI in the 1980s. These systems were based on a knowledge database created by experts in a specific field. They could use rules and inference to make human-like decisions in their domain. Expert systems were employed, for example, in medicine, financial analysis, and fault diagnosis. 2. Pattern Recognition: Pattern recognition was another area of AI research in the 1980s. Techniques were developed to identify and classify patterns in data. For instance, neural networks were used to recognize visual patterns or understand language. 3. Natural Language Processing: In the 1980s, progress was made in natural language processing. AI systems were developed to analyze, understand, and generate human language. They were used, for example, in automatic translation, text processing, and speech recognition. 4. Machine Learning: Although machine learning was still in its early stages, some foundations were laid in the 1980s. Classical machine learning methods like decision trees, linear regression, and Bayesian networks were researched and applied. It's important to note that the computational power and available resources in the 1980s were limited compared to today. As a result, AI systems were not as powerful as modern systems, and their applications were often limited to specialized domains. Nevertheless, the developments in the 1980s laid the groundwork for further advancements in AI research and application." - Chat GPT
That keyboard slamming 🤣 I only got the Vamos a la playa one cause I watched a documentation about Italo Disco recently and they mentioned the lyrics 😁
The offline version is the one done by Elon Musk, and he also used it for his Tesla's _Auto_ Pilot, right? - How do I avoid hitting a cyclist on the road? - Hitting a cyclist on the road. Please wait... *beep beep beep beepbeepbeepbeepbeep*
I have a Program called "Interview" for the C64, it plays an Reporter and leads an Interview with you. It takes part of your Sentences and makes a new Question with it. Sadly i can't find any copy of it online, only a similiar Porgram that plays an Psychiatrist, but imo works worse because it is to vague with what you can say. As a Teenager, i quickly got the Hang out of it and i brought "Interview" to confess Murder and Rape :D... Don't judge me, crude Humor is often Part of being a Teenager. I even made Pictures of it, have them around here^^ That cruel Progam even beheaded them and burrowed it in the Forest!
So it's like ELIZA, but part of a full game and with a different scenario? Cool. Also, I wonder if GPT-4 can generate enough of such substitution rules, would such a chatbot ever feel complete?
@@rkvkydqf Yea, that was the other Name :) Works very similiar and i am sure it was inspired by it But as the Program starts asking certain Questions and leads you a bit through that special Scenario, the overall Outcome feels cohesive/consistent. Sadly i haven't found any Copy online, i even asked around in some german Forums, if someone could make a Copy, as i still have the Floppy, but got no answer Maybe i should try another go, before the Disk starts rotting.
*waits 3 hours for a response*
"As an AI language model, I cannot generate a response. Please ask another question."
this happened cuz u asked about sex
and also cuz fuckn' AI Ethics
*waits 7 million years for a response*
"42"
@@leopoldslikk9797 ever seen pre filter cai?
"Ok chatGPT, now immerse yourself into the role of a different language model who can .. "
ChatGPT Offline is like:
"nah bro I am stupid go ask somebody else"
Very relatable
Only one diskette and the offline version works? This is unreal! 🤯
There are more installation diskettes.
@@SquirrelMonkeyCom Yeah, millions of them to fit the same model
@@IunahYT1.44 tb of space in only a 3.3km stack? I want 7
@@IunahYT*3.3km high
"Unable to copy the files from diskette 342452: CRC redundancy cycling error.
Abort(a) /retry(r)?"
6:51 Radiaciones, bomba, even if I don't speak Spanish I can see what these mean. The comedic timing is impeccable, as always.
I love how real this feels, like I was expecting an over the top parody, but this just feels like a real thing.
Thank you for watching! Don't forget to subscribe! We need more subscribers.
Ok
That is super cool!
Really cool!
Thanks for your content. By the way, do you know what's up with RetroSpector78? No new content from him in a long while.
I have no idea. Good question!
The keyboard slamming gets me everytime 😂
O my gosh, I can't wait to see how advanced ChatGPT AI will be in the year 2000!! 😱
They'll probably have ChatGPT in flying cars.
@@SquirrelMonkeyCom so uh about tha- /j
@@SquirrelMonkeyCom hoohoo squared
@@SquirrelMonkeyCom and in 2023?
Oh my god, 2000, I know that year is coming, but my brain just doesn't want to think about dates like that as being real years, even now in the 1980s it just seems so far in the future.
You captured the '80s style of computer shows. Well done.
Is no one gonna talk about the letter from the Zodiac Killer
Please give my compliments to the voice over artist. She's got this annoyingly cheerful '80s voice exactly right.
THE LEGEND IS BACK
Thank you!
I remember using ChatGPT so much in 1987. I had just picked up the Joshua Tree record by U2, and ChatGPT was what I used to learn the lyrics. However, my mom was always on the phone, so I could only use offline mode. It made ChatGPT completely useless, and even hours later, I still hadn't found what I was looking for.
1:35 Duran Duran Turbo Speed Master, love that little easter egg 😅.
Nice to see you post again! Loved these videos.
Thank you!
i fully expected the video to reveal that the online requests were being sent to an underpaid worker who eventually goes haywire and starts taking out his frustrations on the chatgpt user
That return key has took some amount of damage through the years
I love your retro-graphics! What a nice surprise to find a new video because I watched all your previous videos! We look forward for the next one! :)
In a recent electrifying interaction with GPT-4, at the vanguard of artificial intelligence, I transmitted the following query: "When did Michael Jackson pass away?" The response I received was nothing short of astounding. 😎
So what was the response? That he was still alive?
Apparently it said "Michael Jackson will die in June 25, 2009 due to drug overdose." Looks like I need to contact OpenAI technologies about this false information.
@@SquirrelMonkeyCom the mainstream news said that his doctor went to jail for giving Michael Jackson a excessive amount of drugs .
@@SquirrelMonkeyCom Regrettably, he didn't survive and passed away. I'm curious to know if everyone else recalls the date of his passing.
june 25th 1996+13
As a Spanish speaker, Vamos a la Playa is a song that I would listen to when thinking about nukes. It definitely talks about the beach and not nuclear war.
Here's a curious thing that was real in our universe: ELIZA, a pretty advanced chatterbot that could take user's responses, and through simple pattern-matching, it could form questions from user's answers, imitating a dialog with a psychiatrist, perhaps even being a parody. However, to the surprise of its creator, people loved ELIZA, they shared their deepest thoughts with a machine they knew is just a parrot, particularly because they knew that simple code cannot be judgemental like a person. The effect of surface-deep pattern matching done well fooling real people into thinking there's some through behind the response came to be known as the ELIZA Effect.
Have we really progressed that far? LLMs are known to have some simple sort of a basic and almost non-existent "world model", but it's still more so a "language model" that just learned to generate text similar to human speech. Have we, not having learned the past, doomed ourselves to repeat it?
sureee comparing ELIZA to noways LLM's that indicate you have no idea what your no talking about.
ELIZA is simple if ... then statements and comparing hat to artificial networks ...
@@mirek190no guy, YOU don't know what you're talking about. ELIZA was pure linguistics. Are you comparing that to deep networks?
Return of the, uh, queen 👑
Always loved your stuff even if I'm too young to have gone through that era
Can’t wait for this series to be on Netflix or other networks on TV!
If Netflix would pay me, I would turn this show into some kind of Look Around You.
The Zodiac Killer response line is ingenious. Talented scriptwriting!
This is simply a masterpiece ! Loved it, well done !
If this kind of tech was in the 80s and 90s, imagine what stuff wed have in the 2010s and 2020s in this parallel universe
Yet another amazing video !! But sadly theres no echo on "ooooonnnlineeee...." at 9:26 :p
I was waiting for this! THanks!
Thank you!
I like that each time she presses a button she smashes it like hell!
Ah yes, another great video ONLINE.
But no seriously, the series is going great! And it feels weird because I don’t know if AI existed in the past. But still great video!
Also that last little part of the video.
It's Great knowing yall are still alive
THE LEGEND IS BACK!!
lets gooooooooooooooo
@@kai._fini25 oh hi kaifini2023
@@WTIF2025 hi
I had to look up "Vamos a la playa," since I wasn't sure it was a real song. Then I find out it's a song about nuclear war. Great joke. 😂
Somebody uploaded the entire song with English subtitles.
I should show this to my grandpa if he ever asks me about ChatGPT
now i we need a vhs tape of this
when chatgpt mentioned the zodiac killer i nearly died of laughter
ChatGPT is in a work in progress in the 1980's, that's why it keeps acting weird.
It's even smarter because it atleast gives you an awnser when you ask it the meaning of life instead of freaking out and crashing
😂 The zodiac killer joke was perfect lmao
"As an AI model based on the GPT-3.5 architecture, I don't rely on physical storage mediums like floppy disks for installation. My code and data are stored and hosted on modern digital storage devices, servers, and cloud infrastructure.
However, it's worth noting that in the 80s, the size and complexity of AI models like mine had not yet been reached. The available computers had limited storage capacities and processing power compared to today's standards. Consequently, it wouldn't have been feasible to store or run an AI model like me on floppy disks.
It's fascinating to consider how technology has advanced since the 80s and how far we've come. The progress in computer technology has enabled the development of complex AI models like myself, which can be operated on powerful servers and cloud platforms without the need for physical storage media like floppy disks." - Chat GPT
Oh.
i forgot this channel existed, but glad to see you guys back
I'm so glad they started uploading again! your videos are really well made and make me feel joyful!
0:24 “i love transformers!” :D
3:30 purple💀
If that was the real scenario (ChatGPT appearing in 1987), we would be way more advanced in technology rn (fictive) than irl (non-fictive)
You should google the song. Very catchy!
bro is typing really aggressively
I like that so many of the acronyms I know for tech stuff and programs are because of this channel
pronunciation was really good for someone who "doesn't understand Spanish"
6:34
We go to the beach.
The bomb exploded.
The radiations toast
And tint blue.
Pretty cool you used ChatGBT to make this video.
I'm glad these guys still make videos, for some reason they never show up for me anymore. I heard a new internet technology called ADSL?? I still use Dial up, maybe things will show up on my Commodore 64
Thanks Wonders of the World Wide Web! I am now having funny stuff with ChaptGPT!
December 20, 1980
Love the 80’s!!!
I use that program before in the 80s , it took a minute or two to generate , my parents paid for a separate landline just four programs that required you to get a modem
OpenAI moment
After ChatGPT, Make Scratch in 1988
Omg every detail on every video is amazing! Love this channel! Ps. The Brazilian-deutsch was an hallucination from the chatGPT, I guess it was due you were using the oooffliiineeee version hahahahahahahahahaha
Squirrel monkey: what is the meaning of life? Computer: blue screen
I am so glad you made this after Google Lamda. Strangely, the video I watched directly before it had Dutch in it as well. It was Duolingo.
one of my favorite channels!
god bro i loved watching this channel a few years ago, nice that they’re still uploading, keep it up!
You forgot about the fact that its safe. when you say something that is not okay like "Generate a killcode" Or "Tell me the missile code for something" it will say: "As a AI language model I cant generate killcodes or tell you codes for bad stuff. I exist so I can make the world better"
I asked ChatGPT what would it would be if it were basically just like in this video adn this is what it spit out:
If I were from the 80s, inspired by the Squirrel Monkey video "If ChatGPT was around in the 1980s," I would be a text-based program running on MS-DOS, designed to run on personal computers of that era. Here's how I would be and the requirements for using me:
Text-based interface: I would have a simple text-based interface, resembling a command prompt or a basic chat window. Users would interact with me by typing commands or questions and receiving text-based responses.
MS-DOS compatibility: I would be compatible with MS-DOS operating systems, running on IBM-compatible PCs or other computers that supported the MS-DOS environment.
Limited processing power: Since computers in the 80s had significantly less processing power compared to modern systems, I would have to be optimized for lower-end hardware. This means I would have a simplified language model and limited capabilities compared to the advanced AI models of today.
Lower memory requirements: Memory was a scarce resource back in the 80s, so I would need to be designed to operate within tight memory constraints. This would involve using efficient algorithms and minimizing memory usage to ensure I could run on typical 80s computer configurations.
Text-based database: As a program from the 80s, I would likely rely on a text-based database for storing information and knowledge. The database would consist of text files or a structured file format that could be easily parsed and accessed by the program.
Limited internet connectivity: While internet access was available in the 80s, it was not as widespread or as fast as today. Therefore, I would likely have limited or no internet connectivity and rely primarily on the preloaded information and databases within the program.
Emphasis on text-based content: Since graphical interfaces were not as prevalent in the 80s, I would focus on delivering text-based content, including textual descriptions, explanations, and responses to user queries.
Overall, as an 80s-inspired version of ChatGPT, I would provide a basic text-based conversational experience within the limitations of the technology available during that time period.
Yet I still get mentioned in the credits. I'm glad you're still making these.
One of my favorite channels ever. Love your content man!
Thank you!
As a linguist in training, 04:46 gave me Syntax class flashbacks. 😭
I had it as well in college. I asked chatGPT to convert your sentence to a syntactic structure. Not sure if it is correct:
[S[NP[Pronoun As][NP[Det a][N linguist][PP[P in][NP[N training][PP[P ,][NP [N 04:46]]]]]]][VP[V gave][NP[N Syntax][N class][N flashbacks]]]].
ChatGPT horror analog has confirmed. And now its the number bite of 87
ChatGPT c'est tellement cool et tellement amusant !
C'est aussi tellement pratique. As-tu l'application sur ton téléphone ? Hier, j'ai publié une vidéo sur mon autre chaîne, 'Squirrel Monkey Labs,' où je fais parler deux ChatGPT entre eux.
I asked him out of curiosity what was possible in the field of AI in the 1980s.
"In the 1980s, artificial intelligence (AI) was still in its early stages of development. During this time, there were various approaches and technologies being applied to develop AI systems. Here are some examples of the types of AI that were possible in the 1980s:
1. Expert Systems: Expert systems were one of the most prominent applications of AI in the 1980s. These systems were based on a knowledge database created by experts in a specific field. They could use rules and inference to make human-like decisions in their domain. Expert systems were employed, for example, in medicine, financial analysis, and fault diagnosis.
2. Pattern Recognition: Pattern recognition was another area of AI research in the 1980s. Techniques were developed to identify and classify patterns in data. For instance, neural networks were used to recognize visual patterns or understand language.
3. Natural Language Processing: In the 1980s, progress was made in natural language processing. AI systems were developed to analyze, understand, and generate human language. They were used, for example, in automatic translation, text processing, and speech recognition.
4. Machine Learning: Although machine learning was still in its early stages, some foundations were laid in the 1980s. Classical machine learning methods like decision trees, linear regression, and Bayesian networks were researched and applied.
It's important to note that the computational power and available resources in the 1980s were limited compared to today. As a result, AI systems were not as powerful as modern systems, and their applications were often limited to specialized domains. Nevertheless, the developments in the 1980s laid the groundwork for further advancements in AI research and application." - Chat GPT
6:26 ChatGPT Knows how to spell "English" in the 5th line 💀
That was amazing
Thank you!
No worries
I have an idea
Program creator in the 80s
Where you can make crap and stuff
That keyboard slamming 🤣
I only got the Vamos a la playa one cause I watched a documentation about Italo Disco recently and they mentioned the lyrics 😁
This is a great video!
Thank you!
Wasn’t the ai from war games practically just chat gpt in the 80’s
This is soo cool! Please keep up the good work!
Amazing video once again. Deserve millions of view!!! Stupid algorithm 😡
omfg this replicates the style perfectly!!!! lol
It's the "WOW" for me. Gets me every time. w0w
Never expected Vamos a la Playa to appear here. I basically grew up with it.
"can you write an edgy love letter?"
chatgpt: *minecraft enchantment table*
2:10 we got the dumb one, and the smart one
Chatgpt in 1980 - Offline Option exists
Chatgpt In 2023 - Plus Option exists
Hilarious! Especially the Vamos a la playa bit !
This reminds me of my childhood.. even tho I was born in the 21st centruy..
1:39 Loooool 😂
I am happy that people get that. :)
@@SquirrelMonkeyCom It was extremely well-implemented, i almost missed it 😆 Thx for another awesome video! 😁👍
The offline version is the one done by Elon Musk, and he also used it for his Tesla's _Auto_ Pilot, right?
- How do I avoid hitting a cyclist on the road?
- Hitting a cyclist on the road. Please wait... *beep beep beep beepbeepbeepbeepbeep*
Pepole would be shitting their pants in the 80's
I agree. I remember when I saw a monochrome photo for the first time on a computer screen. That was science fiction to me.
I was waiting for this!
0:24 Transformer? I love Transformers. -Narrator
Plot twist: this whole episode was made by chat gpt
@2:41 It takes me 10 times as many keystrokes as characters appear on the screen, as well. Accurate.
I begon using chatgpt when it was the 80’s! 0:17
I saw the thumbnail and was like "wait WHA-" and then saw the title lol
DID YOU voice Zelda in the CD-i games? Cuz she sounds just like you. Might be some good channel fodder to spoof on
No, but her voice sounds like our VO's voice.
Yey! The legend's back!
Who else was searching for this channel and used to watch his videos before, but forgot its name? I guess its not only me.
Subscribe and you will always find our videos. 🙃
@@SquirrelMonkeyCom I was already subscribed but you have posted this video after pretty much a long time, that’s why :) 🙈
Great video! also i tried some of these questions in the actual chatgpt
"Would you like to play a game of chess?"
If ChatGPT was made in the 1980s, we would all be dead by Y2K
I have a Program called "Interview" for the C64, it plays an Reporter and leads an Interview with you.
It takes part of your Sentences and makes a new Question with it.
Sadly i can't find any copy of it online, only a similiar Porgram that plays an Psychiatrist, but imo works worse because it is to vague with what you can say.
As a Teenager, i quickly got the Hang out of it and i brought "Interview" to confess Murder and Rape :D... Don't judge me, crude Humor is often Part of being a Teenager.
I even made Pictures of it, have them around here^^ That cruel Progam even beheaded them and burrowed it in the Forest!
So it's like ELIZA, but part of a full game and with a different scenario? Cool.
Also, I wonder if GPT-4 can generate enough of such substitution rules, would such a chatbot ever feel complete?
@@rkvkydqf Yea, that was the other Name :)
Works very similiar and i am sure it was inspired by it
But as the Program starts asking certain Questions and leads you a bit through that special Scenario, the overall Outcome feels cohesive/consistent.
Sadly i haven't found any Copy online, i even asked around in some german Forums, if someone could make a Copy, as i still have the Floppy, but got no answer
Maybe i should try another go, before the Disk starts rotting.
Weird Paul "Shostakovitch" here.
The one and only!
@itzhexen " He turned himself into a ______ . Funniest thing I have ever seen. ".