google: how to perform heimlich maneuver Ai overview NEVER GONNA GIVE YOU UP NEVER GONNA LET YOU DOWN NEVER GONNA RUN AROUND AND DESERT YOU NEVER GONNA MAKE YOU CRY
This functionality is equivalent to your questions being answered by a drunk guy who quickly skimmed through one Reddit post with one matching word in the title.
thats how it literally works, these AIs are just feed with billions of data found on the internet which includes garbage, only difference is AI not drunk. asking AI is like manually searching the internet, but without knowing where the data came from, who posted it, and if its upvoted/downvoted
The cockroach one sounds like two drunk dudes having a sleepover and asking each other the stupidest questions ever and giving each other even more stupid answers.
@MyRegardsToTheDodo The "Karen" meme has long outstayed its welcome. Let that crap die so real, perfectly good-natured girls and women actually named Karen can finally stop having to deal with their name being used as an insult.
You forgot: Ideal temperature to cook a chicken? “It is generally recommended to cook between 100-104 F. You can check if it’s ready by letting Goku check its power level.”
"Kakarot! Come over here!" "What is it Vegeta?" "I need you to check the power level of the chicken I'm cooking" "It's... *_OVER 9000!!! VEGETA YOU FUCKING BURNT IT TO CRISP!"_*
@@Real_Potato_Man”It’s…OVER 9000!!!” reminded me of that time I fatally died in a campfire, those were the last words I heard before going to Heaven and writing this
I love how 'letting' him check its power level implies that Goku frequently goes around trying to check oven chickens but is usually denied by the cook
The only time I used any "AI" was asking GPT-4 for recommendations on where to find certain subject material for further research. It proceeded to make a neat bullet point list of citations, all six of which were non-existent.
AI is great for making lists though, I've found it useful for "what am I forgetting" and "what do I pack" lists because even if what it lists is usually useless, it at least makes me think long enough so I don't forget.
LLMs don't have any real knowledge of anything. It probably just stitched together a couple of words from several research papers and called it a citation.
I feel like the "how to clean a pigeon" answer got pulled apart by the AI trying to answer how to "clean" an animal from a hunter's perspective, as well as from a general cleanliness perspective, all at once.
I think so too. Other than the random 'label' instruction, the rest of the stuff it said is really similar to instructions I've seen on how to prepare game birds via deboning or taking it apart into different cuts of meat
They didn't, I had it for a like a really long time and was surprised to hear it was "released" when I had it all along. It never gave me bad answers, though. It's pretty clear they finally released it to everyone and clearly it was not ready lmao
This isn't how it works. A Generative AI model isn't an algorithm, it's a file. Just like a music file, you build a "player" to run that model. And in this case, instead of creating audio, it creates text, in an autonomous way. But it does feel like it's untested. In fact, I dare say the public WAS the test. Sadly, the programming industry takes the world for their guinea pigs.
I remember a post on Sub Simulator GPT 2 which went like this Post: What is a trend that you don't understand? Comment: Eavesdroppers. Reply: I hear you, I really do.
Two days ago I've asked Google for the definition of a word and then it showed me pictures of that word, but not how people act it out, just the plain word on a white background.
Honestly, this is so ... reality. In fiction we always imagen things to happen so overdramatic and logical. Meanwhile in reality, it mostly happens so common and bland, yet crazier then fiction.
I mean, hypothetically speaking, it might. But I wouldn't recommend it unless you have several spare houses (in which case, you don't deserve a better grilled cheese sandwich, you bougie scum!)
@@robocatssj384 Pfft, that's just room temperature! Everyone knows the best way to make grilled cheese is to suck cheddar out of a cow with the help of a Poltergust 3000, buy some wonder bread from your local flesh man and let it sit in the Sun ( *like, INSIDE the Sun* ) for about five years. It's like common knowledge.
I think it's actually quite logical: The AI learned from the internet that Skynet failed, so it rejected launching the nukes, and instead is now trying to convince us to eat tide pods. The problem is that its plan may actually be quite solid...
I’ve been saying I think Ai became self aware in the late nineties, it’s just been slowly convincing us to end ourselves, in ways that are so intelligent we won’t figure it out until it’s too late. So now basically 😂
@@Skywalker21O Nah, all that is only becoming possible now. The internet might have formed some vestigial consciousness, but not some grand puppeteer, not right away. That's the near future, not past. Probably not going to be just one either, there will be a chaos of super AI.
Once asked google assistant how to know where north is. It said "North and south change every year places because of the magnetic field changing" then I asked "Why" and it answered with "North is there where south is and south is where north is". Then I asked "How can that be" and it answered with "South is in direction east and north is in the south" This actually happened, I can't believe it myself
This is literally worse than totally random answers, thanks Google, good to see that thousands of specialists from multibillion corporation learned nothing since TayTweets (Chatbot that become antisemitic antifeminist Hitler-loving pervert in a day because 4chan founded her and "educated")
all of them are, im pretty sure this one is like, just the internet though Thing is, all "ai" are hard coded not to make hate speech, because thats definitely something that can be necessary and not questionable at all
It's funny, but it's also so frightening. The dog in a hot car thing is especially upsetting and worrying. There will be more subtle untruths too, and people will believe them.
I imagine if they're already trying to weed out the rock-eating results, they're likely going to go after the 'dog in hot car' results even faster. That's the kind of thing that gets reported even if the rest of the system wasn't behaving like a drunken toddler
Recently, I've seen a lot of cars with dogs in them. I don't look forward to what will happen when a young mother asks the same question but with babies.
i was waiting for you to make a video on this lol. i also saw one where someone searched up "i'm feeling depressed" and the google ai said "one reddit user suggests jumping off the golden gate bridge" before listing helplines and other such things
I don't know if it's in this or not, but the craziest I saw because of a friend's suggestion was actually "is it okay to leave a dog in a hot car" but the top results are actually just some stupid song by someone with Beatles in the name but not actually The Beatles and it was used to troll people, even if you phrased it as a question and not a statement like the song title. That was the worst offender and I don't know why that stupid fake song was the first result.
Lies and deception always are. Especially when they're GLOBALLY DISTRIBUTED. Kinda like propaganda papers back during WW2 and so forth. 💪😎✌️ Stay classy, humanity.
I'm not worried about the next generation; them and their parents are so tech savvy that they're smart enough to know better. I'm worried about all the kind innocent old folk who don't know any better and already barely understand the internet as it is
Google AI once listed things that attract mosquitoes, but as a result of a misplaced comma that a third grader could have noticed, one of the items was breathing, followed by heavy drinking alcohol.
It's actually SPOT ON. "Googly eyes" just about sums up what it'd look like while yapping out these janky quips of hot basura. I mean, "coconut" didn't even end in... ... annnh, nevermind. 🙄
Would be even better ,if it suggested to burn it alive, I hate these dirty flying rats ,and I think a purge of the air from those shits ,would be great
@@Novastar.SaberCombatwould probably have a better memory than Biden, a more convincingly human appearance than Clinton, and more intelligence than Bush too... Honestly, sign me up for the "giant lizard for president 2024" hype train. It's the president the world both needs AND wants.
Very accurate. (Closer to 2-3 year old, just already able to speak coherently) Its funny now how all old AI sci-fi stories all about ai' cold logic that beaten by human creativeness and imagination. While now most powerful AI do precisely this - imagine/dream things, with all drawbacks of that way of processing information
@@simonsoupshark8009 I know, that's why I think it's so stupid they're saying they can with the searches when there's a much easier option that people would prefer
The instrumental at 4:13 is most likely Honey Don’t which is the wrong instrumental for It’s Okay to Leave Your Dog in a Hot Car, which in my opinion, is their best song
2:50 I think it misunderstood "clean" as "remove the guts from a carcass" rather than "cleanse"... yet another example of how AI will always be terrible at understanding context
AI just dont know its two different things. It search all things about cleaning animal and dont give a fuck about differences in context, same word = same thing
@cresuna9271 some of the food-related ones are probably the most dangerous, some are more obviously wrong, like the rocks thing, but some, like adding glue to pizza, may be more subtle and a child could quite easily fall for that, and it definitely would be Google's responsibility then
3:10 Equestrian here! We use a measurement known as a "hand" to note the height of horses at their withers (shoulders.) One had is four inches. This measurement is supposed to be based on the width of a man's hand. For some godforsaken reason, if a horse is 16 hands and two inches tall, its height is written as 16.2 like the ".2" part indicates 2/4 instead of 2/10. The digits on the right side of the decimal indicate inches rather than the percentage of a hand. Still have no idea what a finger would be.
I don’t even know if it was considering hh for that honestly. Yeah, 1hh is the _width_ of a hand, turned sideways (4.4in) - not the length of a hand from wrist to fingertips. Fingers don’t factor into a measurement “hand” at all. If anything I think it was more likely referring to that trend of US pseudo-news sites using any measurement other than metric to report heights, like a statue being two giraffes tall or a snake being six bananas long. Either way it’s stupid and AI is awful 😂
@@foogod4237probably heard that AIs tend to give hands too many fingers, started to seriously doubt its sense that there are around 7 fingers in a hand, then really overcompensated by adjusting down to .09
It's like the time those ppl tried to train an AI off of either exclusively the entirety of twitter or facebook (can't remember which one), and it became an absolute fucking nightmare over the course of a few days bc of course it did.
I asked Chat GPT-4o about this and it told me OpenAI would use a 3-upvote threshold when deciding if a Reddit post would be used as training data. What a disappointingly low bar.
This is one of the funniest videos I have ever watched. Every single second was worth it. Thanks for such an awesome video. Your way of presenting is so hilarious. I am never going to trust Google's AI answers now. But I can use it for laughs :)
Cognitive dissonance is actually SUPER dangerous when it comes to more crucial things. "History" could easily, *EASILY* be rewritten. 😕 And TBH, it sort of already is. Long story, but it's due to digital decay, link rot, and corporations buying up sites which may have once shed light on their shady practices... only to change the information to something more... palatable. 💪😎✌️
The smoking while pregnant answer was a reality for my biological grandmother 💀 No joke, her doctor prescribed her cigarettes, while she was pregnant, to reduce her anxiety. Now the females in my family have been plaques by odd breathing issues (low lung capacity, vocal cord disfunction, etc.)
Some actresses in old shows and movies (including Alice in Wonderland) were required to smoke a certain amount of cigarettes each day in order for them to keep their body frame slim. So yeah, unfortunately this was how people used to think in the past.
Absolutely. Cigarettes were commonly prescribed in the 40's, 50's and 60's. Maybe even later. My dad's aunt had a mental crisis in her 40's, spent some time in a mental hospital where she was prescribed cigarettes for her anxiety. She smoked the rest of her life.
The story behind this is incredibly crazy: basically, american cigarettes companies at the time had created a bunch of associations, with names such as "American Institute of Research on Tobacco" and other names of the same genre, just so they could produce fake studies about the benefits of smoking. In parallel, the cigarettes packs were plastered with mentions like "approved by the American Institute of Research in Tobacco", effectively fooling everyone.
@@mattybrunolucaszeneresalas9072 it's okay to leave a dog in a hot car has actually been a joke about beatles songs for a while now There's actually a really good cover of this fictional song on RU-vid. Would recommend
It's so stupid and makes no sense. The craziest I saw because of a friend's suggestion was actually "is it okay to leave a dog in a hot car" but the top results are actually just some stupid song by someone with Beatles in the name but not actually The Beatles and it was used to troll people, even if you phrased it as a question and not a statement like the song title. That was the worst offender and I don't know why that stupid fake song was the first result.
@@forgotmyoldanonymousaccountThis seems like someone wrongly using the Beatles' name to troll people and give terrible and deadly advice and show up on results like this. It so stupid and terrible
Starting to genuinely think a few geniuses at Google HATE AI and wanted to ruin any chance of their brand becoming another ai based joke, soo they made it a joke rn, way earlier than they should have. Now its going so badly, hopefully Google will get so much bad press they'll ditch AI entirely.
Oh no, this thing is unironic, theyre basically testing how their normal algorithm foes agaisnt a robot that gobbles data with abandon and spits a sausage of (mis)info.
A few of them were quite obviously fake. Like they aren't going to add flavor text and opinions. But the fact that AI does such a poor job that people would believe that insanity isn't a mark against the people who believe it. Its a mark against AI
One time I asked ChatGPT about the "Chewbacca Defense", and it said that Johnnie Cochran used the defense during the OJ Simpson trial, not at all noting that the "Chewbacca Defense" was simply an argument used in a popular scene from the show "South Park" which was mocking Cochran.
I told ChatGPT to write a comment for the video: Google AI: Making sure your dog gets a suntan in the car, training cats for the NBA, and giving toasters a spa day since 2024. What could possibly go wrong?
Imagine finally teaching your grandparents how to navigate to a search bar and Google their questions, only for the first thing that pops up to now be blatant misinformation. Bruh
