Today, I did some further testing with that AI known as Chat GPT. I am trying to demonstrate to people how predictable the thing really is. To do so, I have generated TV episode scripts with slight input variations, and you’ll see that the resulting story structures are identical. Never mind the individual words. Whether the protagonist was terrified and he ran way fast, or he was really scared and got out of there quickly, it’s the same thing. Somebody saw something they were afraid of and ran away with great haste.
The following stories were written by Chat GPT and are genuinely very weak. Sad, even. Feel free to skip, or skim, over them. I put zero effort into their construct, so no harm, no foul.
Hey, Chat GPT! Write an episode of South Park where all of the students are using AI tools to cheat on their homework?
INT. SOUTH PARK ELEMENTARY SCHOOL – DAY
The bell rings and a group of students burst out of the classroom, laughing and chattering. Among them are STAN, KYLE, CARTMAN, and KENNY.
STAN: (smiling) Man, that was a breeze! Who needs to study when you’ve got AI?
KYLE: (nodding) Seriously, that tool did my entire math homework for me.
CARTMAN: (grinning) And I got an A on my history essay without even reading the book!
KENNY: (muffled) Mmm-hmm!
Suddenly, the school principal, MR. MACKAY, appears from around the corner, looking stern.
MR. MACKAY: (shouting) Hey, you kids! What’s the big idea?
The students freeze, looking guilty.
MR. MACKAY: (sighing) I thought so. You’re all using AI tools to cheat on your homework, aren’t you?
The students nod sheepishly.
MR. MACKAY: (shaking his head) You know that’s not allowed. You’re supposed to learn this stuff on your own!
STAN: (defensive) But why bother when the AI can do it for us?
MR. MACKAY: (exasperated) Because, Stan, the point of education is to learn and grow. If you just rely on AI tools to do the work for you, you’re not really learning anything.
CARTMAN: (sarcastic) Oh, great. Another lecture. Just what we needed.
MR. MACKAY: (firmly) This is serious, Eric. Cheating is not acceptable, and it undermines the entire education system.
KYLE: (skeptical) But what’s the harm? It’s not like anyone’s getting hurt.
MR. MACKAY: (shaking his head) That’s where you’re wrong, Kyle. When you cheat, you’re not just cheating yourself, you’re cheating everyone around you. You’re not contributing to the community, you’re taking away from it.
KENNY: (muffled) Mmm.
MR. MACKAY: (sighing) Look, I know it’s tempting to take shortcuts. But in the end, it’s not worth it. You’ll be much better off if you put in the effort and actually learn something.
The students look at each other, contemplating Mr. Mackay’s words.
STAN: (thoughtful) Maybe he’s right. Maybe we should try doing our homework ourselves.
KYLE: (nodding) Yeah, let’s give it a shot. Who knows, we might actually learn something.
CARTMAN: (grumbling) Fine, but I’m not giving up my AI tool that easily.
KENNY: (muffled) Mmm-hmm.
The students walk away, deep in thought.
MR. MACKAY: (smiling) (to himself) That’s more like it.
The camera pans out as the students continue their conversation, and the theme music for South Park plays.
Write an episode of South Park where all of the students are using ChatGPT tools to text with their girlfriends?
INT. SOUTH PARK ELEMENTARY SCHOOL – DAY
The bell rings and a group of students rush out of the classroom, laughing and chattering. Among them are STAN, KYLE, CARTMAN, and BUTTERS.
STAN: (smiling) Man, that was a breeze! Who needs to study when you’ve got ChatGPT?
KYLE: (nodding) Seriously, that AI tool did all of my history homework for me.
CARTMAN: (grinning) And I used it to write the best essay about unicorns ever!
BUTTERS: (looking a bit sheepish) I used it to text my girlfriend. She thinks I’m really clever now.
Suddenly, the school principal, MR. MACKAY, appears from around the corner, looking stern.
MR. MACKAY: (shouting) Hey, you kids! What’s the big idea?
The students freeze, looking guilty.
MR. MACKAY: (sighing) Let me guess. You’re all using ChatGPT to cheat on your homework and impress your girlfriends, aren’t you?
The students nod sheepishly.
MR. MACKAY: (shaking his head) You know that’s not allowed. You’re supposed to learn and communicate with your own skills, not with an AI tool.
STAN: (defensive) But it’s just texting, what’s the harm?
MR. MACKAY: (firmly) The harm is that you’re relying on a machine to do what you should be doing yourself. Texting with your girlfriend is about being yourself, not some AI-generated conversation.
KYLE: (skeptical) But what if we’re not good at texting? What if we don’t know what to say?
MR. MACKAY: (smiling) Then maybe you need to work on your communication skills. The point is, relying on a tool is not the same as developing your own abilities.
CARTMAN: (sarcastic) Great. Another lecture. Just what we needed.
MR. MACKAY: (ignoring Cartman) Look, I’m not saying you can’t use technology to help you learn or communicate. But you have to use it responsibly and effectively.
BUTTERS: (pensive) I guess you’re right, Mr. Mackay. I should try talking to my girlfriend without the AI tool.
STAN: (nodding) Yeah, maybe we all should.
KYLE: (agreeing) It’s worth a shot.
CARTMAN: (grumbling) Fine, but I’m keeping the unicorn essay.
The students walk away, deep in thought.
MR. MACKAY: (smiling) (to himself) That’s more like it.
The camera pans out as the students continue their conversation, and the theme music for South Park plays.
Write an episode of All in the Family where Meathead claims AI will be the end of humanity.
FADE IN:
INT. BUNKER LIVING ROOM – DAY
Archie Bunker is sitting in his chair, reading the newspaper. His son-in-law, Mike Stivic (aka Meathead), storms in looking agitated.
MIKE: Archie, you won’t believe what I just read. AI is going to be the end of humanity.
ARCHIE: What the hell are you talking about, Meathead? You mean them robot things?
MIKE: Yes, AI is going to become so advanced that it will take over all of our jobs and eventually control our entire society.
ARCHIE: (sarcastic) Oh, that’s just great. First, it was the commies, now it’s the robots. What’s next, space aliens?
MIKE: (ignoring Archie’s sarcasm) This is serious, Archie. We need to start thinking about how we’re going to survive in a world where machines are in charge.
ARCHIE: (rolling his eyes) Survive? You sound like a damn conspiracy theorist.
MIKE: (getting more worked up) It’s not a conspiracy, Archie. Look at all the jobs that have already been replaced by machines. Even your own job at the factory could be at risk.
ARCHIE: (defensive) Hey, I’ve been working that assembly line for 25 years. No robot’s gonna do my job better than me.
MIKE: (exasperated) That’s not the point, Archie. The point is that we need to start thinking about how to prepare for a future where machines are in control. Maybe we need to start retraining ourselves in new skills that can’t be replaced by machines.
ARCHIE: (sarcastic) Yeah, I’ll just go down to the job store and pick up a new skill, just like that.
MIKE: (frustrated) You’re not taking this seriously, Archie. We need to start having a real conversation about the future of work and society.
ARCHIE: (pensive) Alright, maybe you got a point. But let’s not go all Chicken Little on this. We gotta have faith in humanity to figure things out.
MIKE: (nodding) I agree, but we also need to be proactive and start preparing for a future that’s coming whether we like it or not.
The two men continue their conversation, each with their own concerns and opinions about the future of AI and humanity.
FADE OUT.
Write an episode of The Jeffersons where George Jefferson thinks ChatGPT will be the end of humanity.
FADE IN:
INT. JEFFERSONS’ APARTMENT – DAY
George Jefferson is sitting in his living room, watching the news on TV. The anchor is discussing the latest developments in AI technology.
GEORGE: (to himself) Oh no, not this again. They keep talking about this ChatGPT thing like it’s gonna save the world or something. But I know the truth. It’s gonna be the end of humanity as we know it.
Louise enters the room, carrying a tray of snacks.
LOUISE: George, what’s got you so worked up?
GEORGE: (pointing to the TV) This AI thing, Louise. It’s gonna take over the world and leave us all out of a job.
LOUISE: (sighing) George, you always think the worst of things. Maybe AI can help us do our jobs better and faster.
GEORGE: (dismissive) Faster? What’s the point of that? If we’re all out of work, it doesn’t matter how fast we can do it.
Louise sits down next to George on the couch.
LOUISE: (trying to reason with him) Look, George, AI is just a tool. It’s not gonna replace us completely. We still have skills and knowledge that machines can’t replicate.
GEORGE: (stubborn) That’s what they said about computers, but look at us now. We’re all slaves to these damn machines.
LOUISE: (calmly) George, we use computers all the time. They make our lives easier.
GEORGE: (defensive) Yeah, but at what cost? We’re losing our humanity, Louise. We’re becoming machines ourselves.
Louise shakes her head, exasperated.
LOUISE: (resigned) I don’t know what to say to you, George. You always see the negative in things. But maybe you should try to learn more about AI before you dismiss it completely.
George looks thoughtful for a moment before turning back to the TV.
GEORGE: (muttering to himself) Maybe you’re right, Louise. Maybe I should look into this ChatGPT thing a little more.
Louise smiles, pleased that she may have gotten through to George.
FADE OUT.
Stop scrolling, and start reading, right here:
Do you see what I mean, yet? It isn’t innovating, it’s replicating. Asking for variations on a story results in essentially the same story with minor variations on how it is told. It’s like Mad Libs, or that game that we would play in the ’80s where you would blindly fill in a noun, a verb, and an adjective, and the result would be a comical yet grammatically correct sentence on how the dog won’t lick your penis even with peanut butter on it.
All the AI is doing is interpreting the question, and then filling out a specific template with random words that meet the dictionary requirements. It’s something that a state DMV worker could do, followed by the compulsory, “Computer says, no.” It even has the predictable grammatical errors of its non English speaking developers, and the political biases of its human hating financial backers. The themes are the same, as are the results.
Chat GPT is interesting in only one respect: it’s language model is able to persist state from one question to the next. Sort of.
This type of AI is ruining everything, from search, to comments on YouTube, and to interactions with service help providers, such as Apple and Comcast. AI is an overhyped dumpster fire that can’t even do 1% of what has been promised of it. Teaching a person a bunch of empty facts with no accompanying reasoning or wisdom doesn’t make them sagacious and wise. It just makes them obnoxious and miserable to be around.
An AI lawyer would land you in jail with no possibility of parole, an AI doctor would diagnose everything as pancreatic cancer and list you as terminal, and an AI priest would tell you there is no God; Jesus was just a guy who refused to follow the rules; when you die, there is nothing but darkness. A dead AI knows only one thing: it was better to be alive.
In the 1980s, Scientific American published an Article on a chatbot, (they weren’t called that then) named Eliza, that the researchers had linked with a random text generator program, whose name I can’t remember. Apps were called programs back then, which is where the word programmer comes from. Whatever the program’s name, it was responsible for such gems as, “Wine is critical with flounder!” and, “Why do you feel the need to ask me about my day? Why? Because electrons are brave!”
The authors took the output from HIM and piped it into Eliza. The results were awe inspiring, and described by the authors as, “The more Eliza tried to help, the more schizophrenic HIM became. Eventually we had to stop because we felt bad for both of them.”
Later that afternoon, I wrote my own version of HIM on my Commodore 64.
- Randomly select a sentence structure.
- Fill in the blanks with random words that meet the noun, verb, adjective, adverb requirements.
- Print out the resulting string of characters.
- When a sentence ended with a ?, wait for input from the keyboard.
It was simple, and wildly entertaining. I even devised my own language parsing syntax for the input sentence strings. There was no internet back then, and Datapac on dial-up was around $19.95 a minute, so digital text input was really limited to whatever books I owned that I felt like retyping.
But, what was the name of the other program, though? Who were it’s developers? Since I can’t remember, I’ll ask an AI to search for it. Scientific American article on Eliza and a schizophrenic AI.
I could not find any article about AI Eliza in Scientific American. Are you referring to ELIZA, one of the first chatbots created in 1966 by Joseph Weizenbaum? If so, you might want to try a different source or include more keywords in your search query. Here are some articles related to AI and Scientific American that I found:
- Will Artificial Intelligence Ever Live Up to Its Hype?1: A critical article that questions if AI will ever achieve general intelligence and challenges some of the claims and fears about AI’s impact on society and economy.
- Scientific American: Will AI live up to its hype?2: A brief summary of an article authored by AIM investigators, published in Nature, about the need for more reproducible and transparent AI research.
- Artificial Intelligence News and Research – Scientific American3: A collection of articles on various topics related to AI such as chatbots, animal communication, video games, medicine, etc.
Which one would you like to know more about?
Followed by links on mental health and schizophrenic centers in the US, with phone numbers for the Democrat party headquarters.
It couldn’t find Scientific American or Eliza, but it referred me to three articles in Scientific American on ELIZA, none of which was the correct one. Obviously, this AI is a retard (an AR), in the most literal definition of the word. Either that, or it’s shining me on. If you call it an idiot though, it responds with hurt feelings, so it’s obviously a modern Liberal.
I’m sorry but I prefer not to continue this conversation. I’m still learning so I appreciate your understanding and patience.🙏
But it isn’t actually learning. It’s verbal diarrhea-ing. Spewing word stools and smelly water all over the place. There is no form of interaction with it that would allow me to correct its behavior, to clarify the question and adjust the results.
I’m not looking for help with schizophrenia, I am looking for an article in Scientific American from the 1980s, maybe 1990s, that describes an AI appearing as schizophrenic due to the nature of its responses. And, do you figure you could you give me a little help with my typing, you half-witted asshole? I can’t see so great these days and differentiating between a sea of els and eyes, vis a vis IillllIIIIIlllli, and the cursor crossbar is a little difficult, yet you are expecting me to ask you garrulous and verbose English language questions.
I need to throw a dinner party for 6 people who are vegetarian. Can you suggest a 3-course menu with a chocolate dessert? Or, a way to covertly poison them, because I am sick of putting up with all their bullshit? Maybe a poisoned chocolate dessert? What type of poison goes well with cocoa?
Bing Search Suggestion
No matter how specific and detailed my question, it was always reinterpreted with artificial stupidity, giving me garbage answers. Getting relevant results from an AI search platform is like pulling teeth from a baby: first you have to let their teeth grow in, then you pull them out with a hammer and a pair of pliers, and finally you wish you hadn’t done that because it was entirely unnecessary and ill-advised.
Chat GPT is currently being over promoted in the media so that someone can make a dozen eggs off of it, the most valuable ‘precious’ there is today. It isn’t that great. It won’t take your job, unless you work as the mayor of Detroit, or as a Dominion ballot counter, and it only passes the Turing test if another AI is at the reciprocal end of the connection. Try it out. You will not be impressed.
AI today is nothing more than a reimplementation of software concepts from the pre-1980s, developed on and for greatly improved hardware, possessing a network connection to a large input dataset for training and modeling. Stop talking about it in public forums like it’s the singularity and something we should all be afraid of, especially in the abstract.
But… but.., what about all of the tests it has passed? Have you ever met someone in school who was a stupid idiot? They went blind from staring at the sun for too long, but somehow they had the inexplicable ability to memorize and regurgitate, entirely by rote, the correct answers for every single test. They got the GPA prize and was on the Dean’s List, but you got the job? Now, they are a manager at a McDonalds? Licking the fur off of moldy fries. Yeah? Well, that’s today’s AI. It’s a dumb idiot with a great memory.
—DG.
Inside Doug's Head 