Hey shut up and tell me about Egypt AI man! My step dad listens to the AI YouTube shit. So much of it is History Channel shit, like, I'm supposed to be the one nostalgic for History Channel
I don’t have any friends let alone friends that know how to cook. So I’m going to use AI. I have a lot of basic questions that recipes just assume you know. I never would have known to skin ginger with a spoon. Also, recipes are just one component. AI helps me make an entire meal that works together.
There's tons of online articles, or videos, by real humans. There's probably a reddit thread for cooking suggestions and tips. Hell, you could probably just ask here and some people would post suggestions and help. LLMs just scraped all that human knowledge and just regurgitates it, not understanding it, not compensating the actual human effort that went into all that, and often screwing up the more you use it.
The people who own AI don't want us to be connected to each other. The entire point is getting us less dependent on each other and more dependent on them.
The thing that really irks me about AI is the way they try to sell it basically as a shortcut for the most mundane shit. There’s that one ad where some guy uses it to check which headphones have the longest battery life. Like really? Less than a couple of minutes looking at specs or reviews will get you the right answer.
I'm going to keep saying this. People don't hate AI they hate companies trying to cram it into everything, companies trying to replace people and people who misused it. AI is a tool not a cure all
Isn't that kind of the point of this video? Not just of AI, but us spending too much time engaging online on superficial things (such as imgur) and not reaching out to people we know to build better relationships. Of leaving the house to do activities where we might engage with people with similar interests. A walk in the park.
It actually is great but we need to bring back the third spaces to supplement it. There needs to be parks in walking distance and local lodges for youth to join, etc.
yeah I get that, but when my washing machine died, I did not ask the AI, but my colleague who I knew was regularly tinkering on electrical devices. I went to a short trip to Barcelona and and did not ask AI but my colleague who actually is from Barcelona. In both cases we shared also personal stuff as mentioned in the video. Not mentioning colleagues helping each other with moving or renovating, referring doctors, gyms, restaurants, brokering jobs or even flats.
Staying inside and not being exposed to the sun is unhealthy. Breathing the inside air, the air conditioned air.. again, not healthy. Addicted to social media, lacking any exercise outside. They're trying to k¡ll us.
Thanks for this. Not knowing anything about the voice actor, I assumed his real name was Shawn Smucker. He gives it fantastic gravitas, but Smucker definitely deserves a lot of credit for the beautiful poem/essay.
I was under the impression that the dude was doing a mimicry of certain AI things... so yeah those concerns are valid in that THAT'S WHAT AI does. Correct if wrong
I mean, 75-77 percent of professional chefs and head cooks are men. It would appear that while women were traditionally forced into the role of the family chef, the people who actually love to cook and make it their entire life's purpose tend to be men. And of course those cooking enthusiasts are the folks you'd want to ask for advice and recipes. So this kind of does break a gender norm, albeit one that only exists in practice and not as a stereotype.
I get what you're going for and it's a nice thought, but 65% of professional chefs being men is more due to the same reason a majority of engineers are men.
The field (the schools, the workplaces, and many of the men in charge) is full of "good ol' boys" that make the environment hostile to womenband force them out.
To be clear, there was no sexism in the poem. The commenter claiming that there was is definitely reaching... but there is definitely A LOT of sexism in the restaurant industry.
If AI worked it would sell itself. The fact that they need to shove it into everything to try and make us buy into it is all the proof you need that it doesn't work.
Google tried to sell my company a Gemini add-on to Google Workspace. We declined. The next year, without warning, they just one day added Gemini to our existing subscription and increased it by $2/user/month. I had to angrily call up customer support and get them to provide me a switch to turn all that shit off.
AI does work for a lot of things. They could, in fact, do a great job as selling it as "this is a tool that will allow you to be just as productive and make just as much money as you currently do in 40 hours, but do it in 20, and have much more time to yourself." The problem is that that goes against their drive for 9/9/6, and would involve sharing the fruits of AI with the common people.
You're still missing the point.. it's the human experience that matters. The human connection made between two conversing individuals. The tool that you praise to be so "great" is one that has stolen from all the humans in the world. Not only has it stolen our work, but our natural resources as well. It has stolen our homes, our jobs, and our connection with each other, much like the oligarchs that control our countries. It's capitalism's prodigal devil child. But you're too naive to understand.
I can ask the AI to read and summarize documentation for me; am I missing out on human connection, there, by saving myself a couple of hours? Or am I going to be able to spend more time with my family instead of reading bullshit?
Or would you simply get more work assigned because of your apparent productivity? Or perhaps your position becomes superfluous because AI can summarize documentation so easily.
That is an issue that would be inherent in any sort of productivity improvement, and is not unique to AI. Should we get rid of printing and digital publishing and go back to the days of hand-copying manuscripts, because it put all of those monks out of work?
I'm absolutely not saying it's not an issue, but it's an issue that needs to be resolved through addressing it via the ruling class. I have some suggestions there of the sorts of things that might be effective, in Minecraft.
What really gets me about AI is that it can know the correct answer, but decide to give you something else entirely in it's stead, with no concern for the consequence. Much like the people who promote it. The fact that it steals value from the work of real people simply enables others to use that talent without any ethical consideration or accountability is also very much a reflection of the people who would use it to replace people possessing imagination, skill, and ethical motivations.
It doesn't know the correct answer, all it "knows" is the statistically likely next token to generate, with some randomness thrown in so it doesn't consistently generate the exact same output each time. It's not a knowledge source, it's a human speech simulator that coincidentally makes a lot of factual statements, because it was trained on a lot of factual statements, so those are likely, but not guaranteed.
It's not that it knows the correct answer or that it chooses to give you wrong answers. It only knows the answer that you are most likely to like and engage with, its training has nothing to do with truth, just if the thousands of untrained people paid pennies for their time who are training it like or dislike what it said.
AI doesn't really know what the correct answer is, only what the correct answer sounds like. It's like the dad from Calvin and Hobbes. It says things that, on the surface, sound plausible. It's only when you think for a moment that you realize it's complete nonsense.
My point is that his made-up answers sound plausible, at least to Calvin. Same with AI. What it says is generally complete nonsense, but it sounds good.
It does sound at least semi-plausible that, for instance, they discover the weight limit on bridges by driving bigger and bigger trucks over it until it collapses, weighing the last truck, and rebuilding the bridge. That's not how it works, but it sounds like it could be.
That would be if you aren't using AI as just a tool to do things better. But I swear that it the same argument math teachers used with calculators. "you wont have one in your pocket always" and now we literally do.
Again, using it as a tool, not just blindly using it to cut corners. It IS beneficial whether you like it or not. Also it is going to be widely adopted whether you like it or not. You can resists and go the way of Kodak or embrace it and become more valuable. Your choice.
It's even worse than that, it doesn't pick the most likely answer based on any kind of logic or evidence, it's picking the best answer based on its training experience where people said they either liked or disliked the answer. Maybe some of them went and looked for credible sources before approving it, but most likely, they just liked it if they agreed with it.
It gets even stupider still, because while it does have some feedback to try promote correct answers, because it only produces the most likely next word, it gets measurably worse depending on the *tone* it responds with. If you use it in a very casual tone, like with a friend called Dave from the pub, it will respond the way it thinks Dave would, only Dave actually has *some* critical thinking capability.
A more specific example I can think of is if you ask it maths problems in a very formal tone (and have it respond similarly) it can sometimes be right, but it gets *abysmal* if you tell it to use a "valley girl" dialect - its training data doesn't connect correct answers with this tone, so it is unable to. Not necessarily in a "valley girls are stupid" way, but more simply "you don't have lots of good examples of valley girls doing maths at all"
Frankly, it doing math right is mostly just a fluke, mostly common shit that youd likely find the answer to on the first page of a google search too, libking to some reddit or stackoverflow answer (which likely wont be in a valley girl tone).
But anything more complicated than that and it starts to guess, and if one thing totally doesnt work with math, its guessing and itll produce complete horseshit (unless specifically instructed to use a calculator tool for the job)
AIi is a list of "probably". Ask it 1+1 it won't do the math. It knows that 90% of the answers are 2, 5% are 11, and 5% are window. they're not "wrong" in the mind of the AI but just "less likely answers"
Wrong. That was only the early non-reasoning models. Current reasoning models literally can program entire games for you by a single prompt with completely new invented physics engine.
And the people that invented cigarettes also claimed they were perfectly healthy. Unless it's fully open source and can be verified by independent third parties your assurances mean less than nothing.
when it goes "oh you're right my bad" that's because you probably just told it it was wrong and you probably want an apology, then it gives you one of the other 95% chance two answers
Wrong. That was only the early non-reasoning models. Current reasoning models literally can program entire games for you by a single prompt with completely new invented physics engine.
You're confusing layers for complexity. "reasoning models" are just a second thread going "do they probably want me to check this math" and "they're a software engineer so they probably want this in python" it's blurry which is harness vs model vs data review/tuning, but it's still "probably" based.
Yup, it tries to give you what you ask for, which is rarely truth or fact, as that isnt the assignment. If you ask it to write an essay why trump is the worst president ever, it will do just that. Ask it to write an essay why he’a the best president ever, and he’ll also do just the same, accuracy or truth are not even considered, just what percentage is the most likely that the user finds acceptable
yeancientone
Eloquent.
GravyEducation
Hey shut up and tell me about Egypt AI man! My step dad listens to the AI YouTube shit. So much of it is History Channel shit, like, I'm supposed to be the one nostalgic for History Channel
massdefect3
I don’t have any friends let alone friends that know how to cook. So I’m going to use AI. I have a lot of basic questions that recipes just assume you know. I never would have known to skin ginger with a spoon. Also, recipes are just one component. AI helps me make an entire meal that works together.
marsilies
There's tons of online articles, or videos, by real humans. There's probably a reddit thread for cooking suggestions and tips. Hell, you could probably just ask here and some people would post suggestions and help. LLMs just scraped all that human knowledge and just regurgitates it, not understanding it, not compensating the actual human effort that went into all that, and often screwing up the more you use it.
doctorId
The people who own AI don't want us to be connected to each other. The entire point is getting us less dependent on each other and more dependent on them.
kontroversialkarma
That's why they bought up all the houses.. to force us into apartments. Why do they call them apartments? To keep us all "apart".
Righteousdew
Backwards trail.
hwatL4bloopy
Oh this is a great argument to use ai thanks.
EvilEmperorXurg
Darklinkinfinite
The thing that really irks me about AI is the way they try to sell it basically as a shortcut for the most mundane shit. There’s that one ad where some guy uses it to check which headphones have the longest battery life. Like really? Less than a couple of minutes looking at specs or reviews will get you the right answer.
TheFastpaws
It is more important that the environment be destroyed more and them get a quick answer though! Using your brain? Pttbbb! So last year!
pgdave
Do you think that normal internet use doesn't use data centers heavily? Or streaming TV? Or doing a google search, or using GPS maps?
shigoobee
I'm going to keep saying this. People don't hate AI they hate companies trying to cram it into everything, companies trying to replace people and people who misused it. AI is a tool not a cure all
Clockworkdancerobot
All those moments...
ZircZZ
Lost in time
MikesMakery
Like tears in the rain
ThatOtherGirlYouKnow
...will be lost in time, like tears... in rain.
RavenmoreArt
Time to die.
Widgetboy
TheUnstoppableWampas
You guys have friends?
OnyxTurret
Isn't that kind of the point of this video? Not just of AI, but us spending too much time engaging online on superficial things (such as imgur) and not reaching out to people we know to build better relationships. Of leaving the house to do activities where we might engage with people with similar interests. A walk in the park.
TheUnstoppableWampas
southflhitnrun
Yeah I appreciate the theme of this but he assumes that most people, in general, have a vibrant and responsive friend group.
TheUnstoppableWampas
He should test that assumption.
TheFastpaws
And then there are people like George Lucas shilling for the plagiarism generator.
PresldentClef
He himself was a plagiarism generator too, so of course. It's kind of funny. https://www.youtube.com/watch?v=lNdb03Hw18M
Baskaa
As an introvert, those are compelling arguments.
Also, where have I heard that voice before?
Pirsquar
Have you played Rogue Trader? Because that's Heinrix van Calox
Baskaa
No, but I've seen many Warhammer cutscenes here and on YouTube. And thanks to IMDB, I now know there is a Gothic I remake.
Pirsquar
Yeah there is! And if I had a powerful GPU, I'd be playing it righ now
gemaeuer
many of these arguments I usually use to explain why I think that (constantly) working from home is not that great.
TheFastpaws
It actually is great but we need to bring back the third spaces to supplement it. There needs to be parks in walking distance and local lodges for youth to join, etc.
gemaeuer
yeah I get that, but when my washing machine died, I did not ask the AI, but my colleague who I knew was regularly tinkering on electrical devices. I went to a short trip to Barcelona and and did not ask AI but my colleague who actually is from Barcelona. In both cases we shared also personal stuff as mentioned in the video. Not mentioning colleagues helping each other with moving or renovating, referring doctors, gyms, restaurants, brokering jobs or even flats.
Gas0
Ok, valid reasons. Thanks!
kontroversialkarma
Staying inside and not being exposed to the sun is unhealthy. Breathing the inside air, the air conditioned air.. again, not healthy. Addicted to social media, lacking any exercise outside. They're trying to k¡ll us.
sakm
Source: https://shawnsmucker.substack.com/p/please-use-ai
thatlamer
Thanks for this. Not knowing anything about the voice actor, I assumed his real name was Shawn Smucker. He gives it fantastic gravitas, but Smucker definitely deserves a lot of credit for the beautiful poem/essay.
HeyWhaddaya
Yes, the voice actor was excellent.
rhalm
In the first skit they just assume that your "friend that likes to make meals" is female. No sexism there at all...
rhalm
And the second part, asking a "friend" about camping and going out for a beer. Who is a woman, of course. Psych! It's asking a dude.
GravyEducation
I was under the impression that the dude was doing a mimicry of certain AI things... so yeah those concerns are valid in that THAT'S WHAT AI does. Correct if wrong
shardix
Keep up the good work, Don Quixote.
PresldentClef
Not enough pearls in the world for what you're trying to clutch at.
FiftyShadesOfMascarpone
I mean, 75-77 percent of professional chefs and head cooks are men. It would appear that while women were traditionally forced into the role of the family chef, the people who actually love to cook and make it their entire life's purpose tend to be men. And of course those cooking enthusiasts are the folks you'd want to ask for advice and recipes. So this kind of does break a gender norm, albeit one that only exists in practice and not as a stereotype.
LoudBirb
I get what you're going for and it's a nice thought, but 65% of professional chefs being men is more due to the same reason a majority of engineers are men.
The field (the schools, the workplaces, and many of the men in charge) is full of "good ol' boys" that make the environment hostile to womenband force them out.
To be clear, there was no sexism in the poem. The commenter claiming that there was is definitely reaching... but there is definitely A LOT of sexism in the restaurant industry.
icantbelieveistillhavetoprotestthiscrap
If AI worked it would sell itself. The fact that they need to shove it into everything to try and make us buy into it is all the proof you need that it doesn't work.
pastelteapot
AI does work well for a lot of things. It is a tool. Not an end all solution.
marsilies
Google tried to sell my company a Gemini add-on to Google Workspace. We declined. The next year, without warning, they just one day added Gemini to our existing subscription and increased it by $2/user/month. I had to angrily call up customer support and get them to provide me a switch to turn all that shit off.
graehall
Techbros loathe the concept of consent - you see it so many of their actions
monkeyontherun4
Fucking hell, do they ever
thanatos777
AI does work for a lot of things. They could, in fact, do a great job as selling it as "this is a tool that will allow you to be just as productive and make just as much money as you currently do in 40 hours, but do it in 20, and have much more time to yourself." The problem is that that goes against their drive for 9/9/6, and would involve sharing the fruits of AI with the common people.
rusrsdude
You're still missing the point.. it's the human experience that matters. The human connection made between two conversing individuals. The tool that you praise to be so "great" is one that has stolen from all the humans in the world. Not only has it stolen our work, but our natural resources as well. It has stolen our homes, our jobs, and our connection with each other, much like the oligarchs that control our countries. It's capitalism's prodigal devil child. But you're too naive to understand.
pgdave
I mean, I'm not sure how much human connection I'm getting when I'm doing software dev.
thanatos777
I can ask the AI to read and summarize documentation for me; am I missing out on human connection, there, by saving myself a couple of hours? Or am I going to be able to spend more time with my family instead of reading bullshit?
MagicBeets
Or would you simply get more work assigned because of your apparent productivity? Or perhaps your position becomes superfluous because AI can summarize documentation so easily.
thanatos777
That is an issue that would be inherent in any sort of productivity improvement, and is not unique to AI. Should we get rid of printing and digital publishing and go back to the days of hand-copying manuscripts, because it put all of those monks out of work?
I'm absolutely not saying it's not an issue, but it's an issue that needs to be resolved through addressing it via the ruling class. I have some suggestions there of the sorts of things that might be effective, in Minecraft.
Dannoboyo
What really gets me about AI is that it can know the correct answer, but decide to give you something else entirely in it's stead, with no concern for the consequence. Much like the people who promote it. The fact that it steals value from the work of real people simply enables others to use that talent without any ethical consideration or accountability is also very much a reflection of the people who would use it to replace people possessing imagination, skill, and ethical motivations.
DerpyBestPrincess
arikelrecords2000
I mess those up all the time it's a real problem. Thanks!
RadicalRecluse
stealing value from people is the entire basis of capitalism.
marsilies
It doesn't know the correct answer, all it "knows" is the statistically likely next token to generate, with some randomness thrown in so it doesn't consistently generate the exact same output each time. It's not a knowledge source, it's a human speech simulator that coincidentally makes a lot of factual statements, because it was trained on a lot of factual statements, so those are likely, but not guaranteed.
tinyfootprints
Still better than my wife, who habitually states things from her imagination as well-established facts.
IlsPls
It's not that it knows the correct answer or that it chooses to give you wrong answers. It only knows the answer that you are most likely to like and engage with, its training has nothing to do with truth, just if the thousands of untrained people paid pennies for their time who are training it like or dislike what it said.
neptix
AI doesn't answer questions, it provides answer shaped responses.
Shaodyn
AI doesn't really know what the correct answer is, only what the correct answer sounds like. It's like the dad from Calvin and Hobbes. It says things that, on the surface, sound plausible. It's only when you think for a moment that you realize it's complete nonsense.
Thorketil
Calvin’s dad knows exactly what to answer to Calvin’s questions. He wants to pull his son’s leg and have a great laugh if he falls for it.
Shaodyn
My point is that his made-up answers sound plausible, at least to Calvin. Same with AI. What it says is generally complete nonsense, but it sounds good.
Shaodyn
It does sound at least semi-plausible that, for instance, they discover the weight limit on bridges by driving bigger and bigger trucks over it until it collapses, weighing the last truck, and rebuilding the bridge. That's not how it works, but it sounds like it could be.
Thorketil
Shaodyn
Or the one where Calvin asks how they find the load limit on bridges. Dad's answer is total nonsense, but it sounds real.
pastelteapot
It also creates value for people to do work they otherwise couldn't. Learning AI can make YOU more valuable.
Lachwen
Until the AI is unavailable for some reason and suddenly you have a bunch of people who don't know how to do the work themselves.
pastelteapot
That would be if you aren't using AI as just a tool to do things better. But I swear that it the same argument math teachers used with calculators. "you wont have one in your pocket always" and now we literally do.
Lachwen
Except calculators don't regularly hallucinate answers.
pastelteapot
Again, using it as a tool, not just blindly using it to cut corners. It IS beneficial whether you like it or not. Also it is going to be widely adopted whether you like it or not. You can resists and go the way of Kodak or embrace it and become more valuable. Your choice.
Z0op
Its not a knowledge base or database or search engine, its a probability machine, it will try to guess the most likely answer based on the question.
Which is great if you want it to fill in gaps in information or rearrange existing information, not when you expect it to give you facts
DerpyBestPrincess
IlsPls
It's even worse than that, it doesn't pick the most likely answer based on any kind of logic or evidence, it's picking the best answer based on its training experience where people said they either liked or disliked the answer. Maybe some of them went and looked for credible sources before approving it, but most likely, they just liked it if they agreed with it.
DrSparken
It gets even stupider still, because while it does have some feedback to try promote correct answers, because it only produces the most likely next word, it gets measurably worse depending on the *tone* it responds with. If you use it in a very casual tone, like with a friend called Dave from the pub, it will respond the way it thinks Dave would, only Dave actually has *some* critical thinking capability.
DrSparken
A more specific example I can think of is if you ask it maths problems in a very formal tone (and have it respond similarly) it can sometimes be right, but it gets *abysmal* if you tell it to use a "valley girl" dialect - its training data doesn't connect correct answers with this tone, so it is unable to. Not necessarily in a "valley girls are stupid" way, but more simply "you don't have lots of good examples of valley girls doing maths at all"
Z0op
Frankly, it doing math right is mostly just a fluke, mostly common shit that youd likely find the answer to on the first page of a google search too, libking to some reddit or stackoverflow answer (which likely wont be in a valley girl tone).
But anything more complicated than that and it starts to guess, and if one thing totally doesnt work with math, its guessing and itll produce complete horseshit (unless specifically instructed to use a calculator tool for the job)
snaptacular
AIi is a list of "probably". Ask it 1+1 it won't do the math. It knows that 90% of the answers are 2, 5% are 11, and 5% are window. they're not "wrong" in the mind of the AI but just "less likely answers"
DerpyBestPrincess
Wrong. That was only the early non-reasoning models. Current reasoning models literally can program entire games for you by a single prompt with completely new invented physics engine.
snaptacular
do you know the difference between models with harnesses?
DerpyBestPrincess
I literally worked with developing Qwen 3.6
neithermenoryou
And the people that invented cigarettes also claimed they were perfectly healthy. Unless it's fully open source and can be verified by independent third parties your assurances mean less than nothing.
PresldentClef
thanks for helping pollute the world with yet another hallucinating world-polluting water-chugging theft engine I guess. Sod off.
snaptacular
when it goes "oh you're right my bad" that's because you probably just told it it was wrong and you probably want an apology, then it gives you one of the other 95% chance two answers
DerpyBestPrincess
Wrong. That was only the early non-reasoning models. Current reasoning models literally can program entire games for you by a single prompt with completely new invented physics engine.
snaptacular
You're confusing layers for complexity. "reasoning models" are just a second thread going "do they probably want me to check this math" and "they're a software engineer so they probably want this in python" it's blurry which is harness vs model vs data review/tuning, but it's still "probably" based.
DerpyBestPrincess
I literally worked with developing Qwen 3.6
Z0op
Yup, it tries to give you what you ask for, which is rarely truth or fact, as that isnt the assignment. If you ask it to write an essay why trump is the worst president ever, it will do just that. Ask it to write an essay why he’a the best president ever, and he’ll also do just the same, accuracy or truth are not even considered, just what percentage is the most likely that the user finds acceptable