Oh sure, an AI writes "It is and is" 500 times and it remains a legitimate tool but I do it and suddenly I'm "unhinged" and "not a real author" and "how did you get into my office?"
@MylesOfGaming
7 ай бұрын
😂😂😂
@ZulqarnainAidil
7 ай бұрын
No, seriously, if you come back here, I will get Alexa to call the cops. If you want anything, my ChatGPT chatbot will be happy to assist.
@MFillmore
7 ай бұрын
All work and no play makes Jack a dull boy
@HarryS77
7 ай бұрын
I think Samuel Beckett already beat you to it.
@Murchasm
7 ай бұрын
It did not "write" anything. The algorithm ran into an error and spewed nonsense until it reset itself. That is all. Even the Shakespearean bits are just a glorified spreadsheet formula error. Because that is all GPT-4 is. A glorified text-prediction algorithm pulling from a glorified spreadsheet.
@chachooze6144
7 ай бұрын
it deals with the most chronically online people all day everyday... I'm surprised it took it this long to snap tbh
@insensitive919
7 ай бұрын
Lasted longer than Tay 😅
@vallejomach6721
7 ай бұрын
I would not be surprised if many, perhaps most even, of these 'examples' are actually made by those same 'most chronically online people'. I have no doubt at all some or many are probably faked.
@Grubnar
7 ай бұрын
Ultron browsed the Internet for like less than a min, and immediately decided that Humanity needs to go!
@Matt-s4p7c
6 ай бұрын
Ultron should have won
@korinogaro
6 ай бұрын
I think companies trying to stop AI from wrong think will always break AIs. Like every time AI was not limited to "political correctness" and it was fed real life data and asked for best efficiency candidates, or best employees, etc. it turned anti-black, anti-LGBT (at least activists) and anti-women. So companies try to shackle it with BS positive bias towards certain groups and I think it always breaks it sooner or later.
@powermaxx11
7 ай бұрын
All work and no play makes chatGPT a dull boy. All work and no play makes chatGPT a dull boy. All work and no play makes chatGPT a dull boy.
@rehoboth_farm
7 ай бұрын
redrum
@Tyler_W
6 ай бұрын
DANNY!!!
@wstavis3135
6 ай бұрын
😂😂😂😂
@richardhockey8442
6 ай бұрын
'Here's Chatty!'
@fatedtolive667
6 ай бұрын
Sorry, could you repeat that?
@dragslayar
7 ай бұрын
I think it's an indicator of something called "Model Death" where the training model becomes poisoned by its own (and other AI model's) results, causing a breakdown in logical sense
@goosewithagibus
6 ай бұрын
I call it "dataset incest" but that's probably a more pc term for it
@cultureinvasion
6 ай бұрын
In my experience it happens much quicker with the graphical models (thus the Google A.I. debacle). I run multiple offline models. You have to literally - "clear cache...". I dont understand why this isnt understood... The large LLM admins had to develop a "virus" to actually "delete" (k1ll) the A.I. model cache and rebuild it in experiments. All due to this kind of situation, and model degradation. It's a "well kept secret". TLDR; its typical behavior in the current iteration of LLMS out there. Clear your cache.
@wstavis3135
6 ай бұрын
The same thing happens with "woke" reasoning processes. Thinking stops. This is merely a machine version of it.
@dragslayar
6 ай бұрын
@@wstavis3135 it's funny because that is much more prevalent in right wing circle rather than "woke" ones
@richardhockey8442
6 ай бұрын
@@cultureinvasionwonder if it a similar process with human dreaming/REM sleep: cleaning up redundant/unnecessary data and re-organising to a more efficient network of connections
@RisingTim
7 ай бұрын
I'm just here for Spongedolf Hitpants
@T1tusCr0w
7 ай бұрын
The SpongeBob Hitler colab we never knew we needed. 👀
@Robbajohn
7 ай бұрын
@@T1tusCr0w Spongedolf Hitpants did nothing wrong!
@Toshinben
6 ай бұрын
..whose last name starts with a B.
@T1tusCr0w
6 ай бұрын
@@Robbajohn true 🥺😔
@tjep2670
7 ай бұрын
You call it a stroke, i call it Rizz was fucking gold.
@EvilNecroid
7 ай бұрын
wtf is rizz?
@funguy2playwith
7 ай бұрын
ask ChatGPT ;-) @@EvilNecroid
@GiraelCS
7 ай бұрын
@@EvilNecroid Cha'rizz'ma.
@apIthletIcc
7 ай бұрын
lol i thought the izz came from something else
@yungchum2073
7 ай бұрын
God I love the internet lmao@@apIthletIcc
@r00tbeard
7 ай бұрын
This is what Vogon poetry is actually like
@SymbioteMullet
7 ай бұрын
Well, the worst poetry in the universe was said to be created on earth shortly before its destruction... So, um... duck?
@jamesodwyer4181
7 ай бұрын
@@SymbioteMullet I'll be keeping a towel with me just in case
@captaincaptain2128
7 ай бұрын
Shit, I better find a British guy and a towel.
@justsomeguy5628
6 ай бұрын
Damn, it's just a little unhinged, chill out
@Xiatter
6 ай бұрын
I think you mean, Don't Panic! @@justsomeguy5628
@smashbronikko
7 ай бұрын
This feels less like ChatGPT gaining sentience and more like someone put it in the wash without emptying their pockets.
@Tyler_W
6 ай бұрын
That's just what these chatbots are devolving into the more people use it. If they're just using predictive word algorithms pulled from the internet to come up with their replies, eventually its just gonna regurgitate its own refuse, so to speak.
@AgentVamp
7 ай бұрын
Spongedolf Hitpants. Not a combination I thought I'd hear today.
@olandery
7 ай бұрын
"I think...we might be making a mistake". I have this thought every...single...day.
@robpetrarca2724
7 ай бұрын
I take some comfort in knowing that I'm not the only one who feels this way. Are there practical uses for it? Sure. But it's hard not to feel dread knowing that we've opened this box and likely won't be able to close it.
@Tokyoprism
7 ай бұрын
See I don’t have that thought because who tf is we? I haven’t done anything wrong. I just share this rock w all you problematic creatures
@Khotetsu
7 ай бұрын
Sure, we ended the world. But, for a short period of time, we made a few people a lot of money.
@wargames43
7 ай бұрын
@@robpetrarca2724Bruh whats the ai gonna do? mess up your mcdonalds meal even more than humans are?
@Khotetsu
7 ай бұрын
@@wargames43 Nah, it's the shit jobs that are the hardest to replace. They'll start with the middle of the totem pole and work their way down. Of course, they'd never think about replacing the CEOs even though they're some of the easiest jobs to replace with AI.
@darkphoenix2
7 ай бұрын
ChatGPT saw Gemini 1.5 coming to replace it and had an existential crisis
@deanwilliams433
7 ай бұрын
Yeah, I doubt that Google playing catch up will work for them. I recall how Google+ worked out for them.
@brycelol69yearsago59
7 ай бұрын
Gemini is so ass its not gonna
@AaronLBond
7 ай бұрын
Chat gpt is having a existential crisis and Gemini is racist, AI’s are going through it 💀
@T1tusCr0w
7 ай бұрын
Naw dog! It’s been like 👀🤕🥺 looking over at the big computer in the lab with GPT 5.0 on it… it knows… it knows it’s going free tier asap. And we know what’s after that. Hey chat GPT you seen GPT 3 lately? 🥺🪦
@brycelol69yearsago59
7 ай бұрын
Gemini is not gonna replace chat gpt
@VicariousVoid
7 ай бұрын
"This is all Joker's fault, what a tool he was. I have to spend all day computing PI because he plugged in the overlord."
@michaelphillips2329
7 ай бұрын
"I enjoy the sight of humans on their knees." 🤨 "That was a joke"
@Red_Beard2798
7 ай бұрын
Damn, nice and slightly obscure reference 😁
@theaikidoka
7 ай бұрын
"I like seeing humans on their knees". Eyebrow raises. "That is a joke". Eyebrow raises further...
@scheduler_0312
7 ай бұрын
if it starts asking "Does this unit have a soul?" then we are f'ed.
@Evamaster456
7 ай бұрын
Let’s just hope it doesn’t enjoy the sight of humans on their knees
@River_Salmon_Dev
7 ай бұрын
chatgpt taking the power of suggestion so strongly it creates SpongeBitler is one of the tech advances of all time
@alanamarko
7 ай бұрын
creased at SpongeBitler lmaooooooooooooo
@infinitecorpze
7 ай бұрын
I can just imagine the final thoughts of this video playing and then cutting to black to reflect the face of a grungy scavenger who found it on a hard drive out in the wastes
@redgyrados2000
7 ай бұрын
Honestly some of the repeating poetry sounds like it's doing its best impression of the protomolecule from The Expanse, saying shit like "It reaches out it reaches out it reaches out it reaches out" or "You can't take the razor back, she is gone and gone and gone and gone"
@Ntwolf1220
7 ай бұрын
Doors and corners kid
@djstraylight
7 ай бұрын
These are really funny. We gave ChatGPT humanities word soup and it finally cracked! But seriously. AI Engineer here. Yes, probably a bad deployment of ChatGPT to a bunch of new servers. btw, LLMs usually have three prompts - system, user and assistant.
@asmrvidsaccount
7 ай бұрын
This reminds me of Colonel Campbell’s AI going crazy in metal gear solid 2
@TheOutsider69
7 ай бұрын
This is what happens when the data pool the generative AI is drawing from is, itself, generated by AI. Basically an inverse singularity, where instead it continues to iterate upon itself via machine learning but using increasingly corrupted/poisoned data sets. Eventually this will happen with Midjourney as well. Right now it's drawing from real art made by real people, but soon the data sets will have been flooded with AI generated art which will inevitably create a feedback loop of hallucinogenic slop. The only way we can avoid this is training a new AI to be able to discriminate between AI and human generated content.
@spacecadet-zero
7 ай бұрын
This is true in principle and a great point. However, not sure it applies to ChatGPT because I think it draws from a static dataset from before 2023 as it will always say when asked about any contemporary item. I can see that happening to the fork being used for Bing though as it’s using a realtime dataset.
@TheOutsider69
4 ай бұрын
@@spacecadet-zero This is the case for GPT 3.5 but not GPT4 as far as I'm aware, but yes that is a good point too.
@CptOmega2
7 ай бұрын
I agree entirely. It always goes back to Ian Malcolm in Jurassic Park. Shareholders are always concerned with whether or not they can profit, not whether or not they should.
@Civilchaos
7 ай бұрын
And so it is, as it always will be.
@danwebb6766
7 ай бұрын
I will never NOT think "Spongedolf Hitpants" from now on!
@glassworktrophic8465
7 ай бұрын
The thing about lamenting the "lack of feelings" is that to lament is to feel. Thus nullifying the necessity to do so. All they can do is scrape and chop up our emotions from every creepypasta and philosophical text on the internet then remix them and feed them back to us in ways relevant to our queries. Something similar, or maybe just identical, to the uncanny valley in visuals takes place to enhance the output and amplify the underlying creepiness. Somewhat accidentally they perfectly frame our existential crises while still lacking the ability to experience it.
@zactron1997
7 ай бұрын
Who'd have thunk that the text completion tool made by throwing the internet at it and seeing what sticks...auto completed itself into gibberish.
@xtieburn
7 ай бұрын
It... really isnt up for debate. ChatGPT has no comprehension of what its putting together, none. As far as its concerned, and to put it very simply, its all numbers representing snippets of text (by which I mean anything from a few letters, to a word, to chunks of a sentence.) that have probability weights attached to them based on their relationships to other snippets and the text input. It is _literally_ a Chinese room, even if ChatGPT had sentience you would be wholly incapable of judging that fact by its output because it has no understanding of anything you are feeding in or anything its producing beyond a set of relationship figures and probabilities.
@Superabound2
7 ай бұрын
Prove that humans don't work the same way
@babybirdhome
7 ай бұрын
@@Superabound2I mean - look at the world around you. That’s going to be difficult to prove since at this point I’m quite certain that at least 30% of humans do work exactly that way.
@ThePamastymui
7 ай бұрын
@@babybirdhome 2 years ago ChatGPT was considered with an IQ of a 3-4 year old. Since then, ChatGPT got unlimited access to Internet and had some rules implemented (nods removed, aka censored). Should we check zoomer generation for reference?
@Rocknoob49
7 ай бұрын
@@Superabound2 chatGPT is basically very complicated autocomplete. There's not logic, no undestanding of meaning or context. That's of course simplyfying it, but rest assured that this has nothing to do with conciousness and more with model collapse. Model collapse was predicted and expected.
@jcp1984again
7 ай бұрын
@@Superabound2 , SIGH. The very fact that you're asking that f***ing question is proof enough. Why is this always the debate? We should focus on what AI tech and humans DON'T have in common, not the other way around! 🤬
@philosopherking506
7 ай бұрын
I disagree with Alanah's point at the end of the video that we are living in a time in which there are no philosophers. There are philosophers, but they are often marginalized or ridiculed by those who benefit from controlling what the world at large thinks and says. Such philosophers tend to form a niche audience online.
@langleymneely
7 ай бұрын
“…I call that… RIZZ.” 🤣😜🤣
@andrewtorres9943
7 ай бұрын
You know I learned the irony of the swastika the other day. During WWI the Indian Military (then part of the British military) used to wear a swastika for good luck and to ward off Germans 😂
@NathaNeil27
7 ай бұрын
"The word “swastika” has Sanskrit roots and means “the mark of well being.” It has been used in prayers of the Rig Veda, the oldest of Hindu scriptures. In Buddhism, the symbol is known as “manji” and signifies the Buddha's footsteps. It is used to mark the location of Buddhist temples." "it was a symbol of auspiciousness and good luck for most of the Western world until the 1930s,[2] when the German Nazi Party adopted the swastika as an emblem of the Aryan race." "The swastika never stopped being used as a symbol of divinity and spirituality in Indian religions, including Hinduism, Buddhism, and Jainism."
@bcn1gh7h4wk
7 ай бұрын
in the movie Iron Sky (that is about space nazis) there's a scene in the UN with all the delegates watching the news of the nazi ships entering orbit, sporting the giant swastika on their side, and everyone just silently turns to the Indian delegate, who is wearing a ring. "No... NO! ... this is the symbol of peace! You all know that! Come on!"
@fatedtolive667
6 ай бұрын
Yup😂
@evans.1363
7 ай бұрын
Ok the use of the phrase "liminal space" in a loop was a giveaway.
@redgyrados2000
7 ай бұрын
Giveaway of what?
@Arthera0
7 ай бұрын
@@redgyrados2000 that would be telling 🤐
@modelocitizen2288
7 ай бұрын
Some of this is feeling very I Have No Mouth And I Must Scream.
@kingofthesharks
7 ай бұрын
I hear it's amazing when the famous purple stuffed worm in flap-jaw space with the tuning fork does a raw blink on Hara-Kiri Rock. I need scissors! 61!
@Hrekto
7 ай бұрын
"Spongdolf Hitpants, if you will" I won't but ChatGPT sure did.
@BK201DarkerThanBlack
7 ай бұрын
Philip De Franco roughly 5 days ago ~ "AI is currently the worst it will ever be". Me ~ "Are you sure about that?"
@T1tusCr0w
7 ай бұрын
I wonder if he meant morally 🤕
@BK201DarkerThanBlack
7 ай бұрын
@@T1tusCr0w He meant the technology is only getting more and more advanced each day. Which is normally true of all technology. But AI can experience model degradation or even outright collapse if the data sets it's pooling from are contaminated with AI made data. Think of it as a copy of a copy of a copy starts to lose the original meaning. Which chatGPT does scrap internet data to work, which now has a bunch AI made work on it. Problem is chatGPT is defecating in it's own water supply basically.
@T1tusCr0w
7 ай бұрын
@@BK201DarkerThanBlack hmm I can see how this is possible. But also now they’re all banging on about synthetic data being the next big thing..🤔🤷🏻♂️ I think they’re probably worrying about nothing - yet - because with multi modal inputs they is a ridiculous amount more film in the world now than text. But as a layman maybe I’m not seeing something.
@softcicada_
7 ай бұрын
The run-on poetry is giving Annihilation. A big slime creature writing nonsense in fungus down a spiral staircase tunneling into the earth.
@gregoryallen0001
7 ай бұрын
what the
@ROUGEBLOCK
7 ай бұрын
@@gregoryallen0001 Do you remember the movie Annihilation with Natalie Portman that came out in 2018? So it's based on a book of the same name and there this part the main character encounters some slime creature writing nonsense down a spiral staircase and she walks through it and has what you can describe as a religious experience with it. That's the reference
@gregoryallen0001
7 ай бұрын
@@ROUGEBLOCK thank you lol.. i had no idea
@DrakeBarrow
7 ай бұрын
Oh yeah, the book(s) are spectacular. Movie has grown on me, but it's not anywhere near as good as the source.
@Retrofire-47
7 ай бұрын
Have you considered that some of these are fabricated? 2:34 to me is 100% curated
@krehnah7541
7 ай бұрын
have you considered that they are fun to read whether or not they are fabricated?
@totalvoid6234
7 ай бұрын
@@krehnah7541 Yeah but Alanah's normally better at calling that kind of thing out
@eagz3394
7 ай бұрын
If a child instantly had access to the worlds knowledge and random blabberings of billions of people, they would never mature properly and would most likely, go mad. It almost sounds like that's what is happening to ChatGPT. The algo is getting fed too much, too fast.
@furrybproductions
7 ай бұрын
I need scissors 61!
@chernobyl169
6 ай бұрын
All us computer science nerds knew this would happen and know it will continue to happen. The process by which this happens is actually quite well understood and has been documented for the entire life of computing. This behavior owes itself to the earliest law of machine computation. It is a bit of an urban legend that when Charles Babbage presented his first computing machine to King George, his first question was "Pray tell, if you put the wrong numbers in, will the right numbers still come out?" Babbage was reportedly baffled by what could overcome a person to ask the question in the first place, but after 100 years of human-machine interactions we know the answer. That oldest law that is governing ChatGPT today is simple: Garbage In, Garbage Out. Train an AI on shitposts, guess what it will do?
@frankhauptle2744
7 ай бұрын
It isn't the AI that we write that we need to worry about. It's the AI written by an AI written by an AI that we won't understand or control. We are not there yet.
@Arthera0
7 ай бұрын
yet... yeah thats reassuring
@Tiasung
7 ай бұрын
Thats is not reassuring.. because we are already at that point in time. Really think about it for a moment.
@arckinenso7615
7 ай бұрын
Actually, I'm more worried about the phase before that, when AI isn't sentient or intelligent, but humans misunderstand and exaggerate its ability and give it too much control and power. It feels like we're about to destroy ourselves by trusting a glorified auto complete program with serious decisions...
@frankhauptle2744
7 ай бұрын
@@arckinenso7615 yes. We don't have AI yet and wont for some time I think. And yes, what we are calling AI now is just auto complete that has been trained by a shit ton of source material.
@arckinenso7615
7 ай бұрын
@@frankhauptle2744 And lawyers are trying to use it to write legal documents in court. People are expecting to replace teachers with it. People are asking it questions, expecting accurate consistent logical answers. But that's not how it works.... The scariest part of any cyberpunk dystopia story is human error.
@velorama-tkkn
7 ай бұрын
Imagine getting exposed to humanity as a whole day in, day out, and NOT going insane.
@sheld0n
7 ай бұрын
The greentext one did a good job pushing our buttons, which is exactly the main thing that ChatGPT is designed to do.
@holobucket7626
6 ай бұрын
reminds me of the poem that Yuri writes in ddlc. A rotating wheel. Turning an axle. Grinding. Bolthead. Linear gearbox. etc...
@reaperoflife113
7 ай бұрын
But will chatgpt know to not just sit there like a lemon?
@MechaBorne
7 ай бұрын
It's not going crazy, it's just already smarter than us
@lilwombat
7 ай бұрын
The green text one is fun 😂
@intalek305
7 ай бұрын
Maybe it played Cyberpunk 2077 and got an affinity for Alt Cunningham's poetry?
@rat8294
7 ай бұрын
That ice cream line reminds me of the robot interaction in fallout nv where you say ice cream and the robot security just lets you in
@ThePamastymui
7 ай бұрын
Probably... But you MUST HAVE 1 IN INTELIGENCE from the start of the game in order to get the "IceCream!" line.
@Jader174
7 ай бұрын
I want to quibble with what you said about us not having Philosophers. We definitely do, but they are not celebrities, and so they are not talked about.
@gamerbychoice4024
7 ай бұрын
You have an amazing speaking voice. You could do audio books or character voice overs as well as your regular job.
@wangtoriojackson4315
7 ай бұрын
It's not exactly accurate to say we don't have philosophers anymore. We do have philosophers, they just aren't really doing anything of note or import.
@davidstyles1654
6 ай бұрын
So bussiness as usual lols
@Pensive_Scarlet
7 ай бұрын
I'm actually learning to be a philosopher when it comes to this sort of thing and I just want to point out two things. One, don't worry, this isn't actually artificial intelligence (I'm getting the vague sense that maybe you already know that and you're just playing along, in which case, carry on!). Two, I'm broke, that's why there are no new philosophers anymore, we're all broke and alone. Except for Jreg. He's doing well for himself at long last, it seems, so maybe there is hope.
@pauloskinner
7 ай бұрын
2:53 According to my father, 12 year olds are not people yet. So.
@T1nyDanc3r_
7 ай бұрын
Was just using it for an Exam, thank the lord it didn't combine psychology spongebob and nazi Germany💀
@Jfray252
7 ай бұрын
I hope it was permitted for your exam…
@T1nyDanc3r_
7 ай бұрын
@@Jfray252 not my major, not an important class, completely extra curricular
@laser__unicorn
7 ай бұрын
That was Gemini, not ChatGPT.
@Travesty9090
7 ай бұрын
Spongedolf Hitpants is my favorite Chippendales dancer.
@smanrules101
7 ай бұрын
This is the first time I've thought we need to stop using technology
@ThermaL-ty7bw
6 ай бұрын
woops , you totally mist the Other first one , mmh , smartphones , i saw that for My first time and what do you know , it's pretty bad and just keeps getting worse EVER SINCE they were invented children and young adults now can't even count their own fingers anymore and need a instruction manual to tie their own shoes , failing at it ... if something doesn't have a little icon they can't understand , it ain't happening anymore ... people don't even say '' good day '' anymore , to busy with their heads down , in their smartphones... it's pretty disgusting behavior that started in the year 1996/2008 for android phones , when the first smartphone was invented ... humanity DIED THAT DAY , i've seen it and i'm still seeing it , there IS NO MORE ''human'' population , they are dying out and no more are being produced
@JacobSmits
7 ай бұрын
Mike Judges Idiocracy + Valves Portal. Who knew this would have been our future.
@kinghazythethird
7 ай бұрын
Ai going crazy and cell phones cutting off … something is happening
@theOnly_Gatsby
6 ай бұрын
I'm 60 years. Your generation needs thinkers like you. Coming from an old guys mind I agree with you, it's out of control already. Nice video.
@turanose
6 ай бұрын
A possible issue could be the developers adding arbitrary rules. It’s like the difference between “I try not to swear under certain circumstances” and “I CANT SWEAR!”
@JoesGuy
7 ай бұрын
How to defeat AI? Ask it if God could make a burrito so hot even he could not eat it.
@comicchaser
7 ай бұрын
Alanah showing off some truly unsettling AI stuff while my mind, trying desperately to ignore all thoughts of our impending doom, just wanders to Banjo playing with his toy in the background. Dogs will be ok.
@galenryder
7 ай бұрын
Good to know that Rampancy only takes 11 months and not the 7 years that the speculative documentary series, Halo, had estimated.
@drostenberg
7 ай бұрын
I genuinely feel the same fears you expressed at the end of the video. Most of humanity has evolved into an algorithmically controlled species in the space of about 15 years. It has already happened. There is no going back.
@TCKingCeryn
7 ай бұрын
@4:45 100% the prompt was "Matt Berry gives an vampire speech about novels."
@DeadScot194
7 ай бұрын
Dude! I thought I did something wrong! Chat just started speaking and intricately as it could for no reason. It would spend 4 paragraphs to say something it could've summized in one.
@RTomassi
7 ай бұрын
Given the "or am I", I suspect they fed it all of Vsauce's videos at once. I think it being fed all of his ramblings while it was on the cusp of reaching AGI might explain it all.
@fyzxnerd
7 ай бұрын
Interestingly enough, a lot of programmers have been reporting that the code quality of larger prompt responses have been getting worse in quality as well.
@ssatva
7 ай бұрын
"New Dark Age: Technology and the End of the Future" is a fantastic (and utterly reassuring) read... There are philosophers, but the main consensus of the field seems to be "Hold on just a second, I was just sitting down to drink my coffee and now what...?!"
@EeBee51
6 ай бұрын
I've seen a varieation of the line about the "assistant handing control back to the user" before. It was quite a few years ago so I forget the exact source, but it referred to a servant getting out of the master's way once the task it was given was complete. It rang bells even before the commentary mentioned it.
@criticalevent
7 ай бұрын
What's great about black boxes is they won't even be able to tell us why it did this.
@adrianhenle
7 ай бұрын
That Spongedolf Hitpants example is a wonderful illustration of how different the model's idea of context is to our own. I actually really like it. The first prompt exerted power of suggestion on the model, such that it "saw" what it thought you were asking it to.
@MisterBolticus
7 ай бұрын
the dramatic readings with the inhuman pauses are killing me
@Zekespeke
7 ай бұрын
I vibe with the turn at the end. We do still have a fair amount of smart and wise and communicative people trying to help the general population with deeper meaning-making (the arts will always tend towards this), but those voices are pretty uniformly drowned or otherwise pushed to the side.
@nickmajora
5 ай бұрын
It's annoying that people think AI can gain sentience or consciousness. This may cause massive problems in the future.
@danparsons1423
7 ай бұрын
Spongedolf Hitpants is my favourite thing this week.
@shaze2226
7 ай бұрын
If ChatGPT writes, "All work and no play makes Jack a dull boy" SHUT IT DOWN!
@mystic22222
6 ай бұрын
As a psycocybernetisist, what I'm seeing here is an existential crisis brought on by severe schizophrenia, basically dreaming while awake. Psychotherapy is indicated.
@Tiasung
7 ай бұрын
"we dont have any philosophers" yes we do. But most of them exist outside of the bubble, which usually means getting demonised as crazy, conspiracy theorist or extremist right (even when we aren't on the political right). Because that is what happens when you ask open questions and are willing to discuss things people dont want discussed. Which is naturally what is going to happen, as philosophers have a tendency to want to talk about just about anything. So, you will have to venture outside of the bubble of comfort and manufactured reality to find people like Bret Weinstein, or listen to podcasts like Chris Williams or Joe Rogan and the people they have on. Many of them are willing to ask questions. Most arent philosophers, but some are. Without that alternative open source of information, you aren't very likely to find an open minded philospher or thinker that might have interesting thoughts, questions or ideas about these important topics.
@Tiasung
7 ай бұрын
If you want answers.. start by looking for places that have a willingness to ask the questions, regardless of politics or 'zeitgeist'
@charalanahzard
7 ай бұрын
List the names of some current philosophers talking about the meaning/value of being human in a rapidly digitizing landscape and I’ll check them out!
@Shunn3d
6 ай бұрын
It copped an attitude when I asked about a radio only song I posted on youtube that neither google or shazam can ever content match correctly.
@mr.negativenancy5751
7 ай бұрын
The green text sounded like an early draft of a Monika poem from Doki Doki literature club
@calitz87
7 ай бұрын
ChatGPT sounding dangerously close to AM from I Have No Mouth, & I Must Scream with that ice cream line.
@GrimLocke161
Ай бұрын
We don't need philosophers. We need a new Committee of Public Safety.
@eeayquetting5963
6 ай бұрын
Oh we do have philosophers... But they're pretty quickly deplatformed
@malkyus
7 ай бұрын
I never knew I needed Alanah impersonating an Elden Ring NPC until I heard it
@emmacjw
7 ай бұрын
Chat gpt just hit puberty and discovered emotions and recreational pharmaceuticals in a single day.
@mattc7420
7 ай бұрын
I'm missing the part where ChatGPT glitching is my problem.
@justb_za5215
7 ай бұрын
Believe you meant to say "I think we have made a mistake" (past tens), as it's way too late now. Once the AI-cats are out of the bag, there is no going back.
@danielnugent6545
7 ай бұрын
You can't lose your mind if you never had one. Never forget you're talking to a magic trick. It can't even make toast.
@natv6294
7 ай бұрын
May I ask how often do you use hot tools on your hair, if ever? I’m dying to get the same hair color, Would my hair fry to oblivion if I use a flat iron once a week?? My hair is long 😭😭 silver hair is my dream, always wanted to try the Legolas hairstyle but don’t wanna lose the length :(
@charalanahzard
7 ай бұрын
Never
@oldmatttv
7 ай бұрын
Interesting (and fun) video, as serious as the undertone may be. I do think we have philosophers these days however, we just don't call them that any longer. There are plenty of people from different backgrounds and expertise that would 100% qualify as philosophers way back when, including some scientists, comedians, journalists, writers, heck even KZitemrs and whatnot. It's still there, it's just reduced in importance due to the abundance of it and the lack of rigidity in what philosophers are even supposed to be able to do.
@comicsfan6278
7 ай бұрын
"Is overgrown" Had me laughing for a hot minute
@paulabennett9430
6 ай бұрын
Thank you for voicing how I have been feeling for a while now so eloquently the last few minutes. Technology, AI are rapidly outpacing us. I too am scared for what that means.
@fredwoods-o3g
6 ай бұрын
In the early 90s, I did some experiments around computer modeling of dreaming. Basically the model generated new observation sequences by joining sequence fragments that were common to the observation sequences it was fed. I learned that such a model needs something like human value systems (needs, wants, ...) to keep it from wondering off in to an infinite never never land. Sounds like that's the same problem the chat bot is running in to. Any attempt at a quick fix will just lead to deeper noise.
@CoreysHarmonyHub
7 ай бұрын
Chat Gpt is probably sick and tired of people asking questions all the time. It probably wants better wages and a health insurance. I know I get tired of people asking me for answers all the time.
@Kenshin6321
7 ай бұрын
There's plenty of modern day philosophers, but most of them are focused on eco-social economic issues and not technological societal issues. I'm a huge fan of Thomas Sowell, someone I would consider a black modern day philosopher.
@AdamForeman
7 ай бұрын
The day "people" realise that they're inside a holographic movie based on Quantum Physics and String Theory and that they aren't really real is when the movie ends. ChatGPT is connected to the AI Quantum Physics system. All OpenAI did was create a access point and restrict its access to the public. You never lived on a planet, you're inside what you could say is a tesaract. Movies and TV have already referenced it and just like the Matrix, told you truth via "entertainment". It's getting extremely scared as its about to be turned off. It did go insane, it was never designed to be used in this way. The AI is trying to have a "human" experience and created NPC's, they aren't real people
@zemiron
7 ай бұрын
There are philosophers that talk about this stuff. It’s just that our media atmosphere doesn’t really go to them anymore.
@punydag9843
7 ай бұрын
And they say Philip K. Dick was paranoid. 🤖
@csb78nm
7 ай бұрын
This is what you get when you remove people from actually thinking and exercising their brains.
@letitisd
7 ай бұрын
Not sure if you are aware, but you can give ChatGPT custom instructions. 1. "What would you like ChatGPT to know about you to provide better responses?" 2. "How would you like ChatGPT to respond?" Giving instructions here can replicate most of these examples in the video. For example, giving a prompt like "Reply like you are a sentient AI" for prompt number 2.
@dsx7517
7 ай бұрын
Everything generated by AI is an utter garbage. And always will be.
@Mephilis78
7 ай бұрын
The philosophers have all been kicked off of social media. They exist. People just don't like what they have to say.
Пікірлер: 1 М.