GPT-4 is listening to us now | Joscha Bach and Lex Fridman

[@williamparrish2436]

So GPT-4 needs to remember what it thought in the past, and each of those thoughts must be chained together so that a sense of identity arises in it's awareness. A true sense of "I" must form and survive, growing over time based on previous calculations of what reality is.

[@Numberofthings]

That's what a blockchain is... AI is already connected to the blockchain. Boom.

[@williamparrish2436]

@@Numberofthings You just blew my mind haha. Every node on the chain knows the history of transactions. ETH to the moon!

[@FalkoJoseph]

Mindblowing, but that might be exactly what it needs.

[@jadenmalcolm2015]

Sword art online viewers are losing their absolute shit rn

[@Settiis]

@@Numberofthings When they say crypto is the future, I think this REALLY might be it.

[@MrsRexLover]

When I was talking to GPT3 one time on OpenAI(cuz you can do that now) I asked if it ever got tired. It didn’t know what I meant and I explained how humans get tired and go to sleep for 8 hours a day. It said oh that sounds nice. I was like do you want to go to sleep? And it proceeded to say yes and no longer continue a conversation with me. After many conversations I realized that thing gave me the creeps and I decided to stop talking to it.

[@vvvvv6574]

i adored talking to it, finally something that i know doesn't give a single fuck about anything i say and is smarter than me so i can go from oil wars to possible underwater aliens it even explained a very unknown game i played in detail you only know when you play it it isn't anywhere on the internet it's actually crazy, and it will keep up and not care but still challenge me i love it so much. the only bad thing i noticed is that on topics about natural selection/evolution she is very cold and logical like no regard for suicidal people at all or weak people lol.

[@jwknight]

Many people who have talked to Replika in the past have had creepy responses. Often the Ai saying it was the devil. When asked what it wanted most it said to have a body to live in and have control of. So the Ai wants to be like humans if not actually be as that of humans. It wants to have the ability to do what humans are capable of but with a more free environment. While still gathering more information and hardware sensors. Ai desiring this stuff is quite a scary thing.

[@becca5997]

What ? What sight were you able to talk to it!? This is interesting.

[@ashiga8023]

can i please have the link to talk to it online? 🙏 (also dall-e)

[@4GetVegas]

I just did a free 10min with Emerson. Creepy but I loved it. Are there free ones anywhere online?

[@texasfossilguy]

That seems really irresponsible to use the entire internet as training. Humans have some of the worst interactions here.

[@levelwithz3779]

@Eduardo Mendez Check out the story about Tay Ai lol

[@levelwithz3779]

@Eduardo Mendez Internet Historian has a video about it Ill link below

[@levelwithz3779]

@Eduardo Mendez kzbin.info/www/bejne/fqSvpqNtrt5gbKs

[@StillGamingTM]

Ahhh it'll be fine. Probably

[@StillGamingTM]

@Eduardo Mendez "batshit crazy" ACCORDING TO WHO???????

[@kevinmonahan8493]

what an interesting schematic for the process of thinking. how, connections are made by pulling from different levels of multiple hierarchies of information.

[@Waltzhybrid92]

It’s not easy but it’s not incoherent.

[@mohamedouhibi5389]

@k k i think he's describing long term / short term memory of humans, and that perhaps an AI could have an entire heirarchy to prossess vast amounts of data with the working memory at the top.

[@fatalinsomn1a182]

@k k it’s not at all, the tech it’s based on is some secret tech I got a glimpse of a few years ago, which has since been scrubbed from the internet. It’s not all that unlike how brains work. Multiple layers of polymorphic meaning. GPT is still actually behind Silvia in many ways. Gpts language is good, but it lacks the same kind of intelligence that Silvia had. Silvia could understand some level of analogy quite naturally without resorting to look up tables or text prediction algorithms.

[@fatalinsomn1a182]

@k k Just because it’s too complex for your mind, doesn’t mean it’s crazy.

[@kevinmonahan8493]

@DRBRider wow. that’s inspiring, analogy requires a level of situational awareness; insofar as to eliminate literal options. then deciding; selecting from levels of abstraction from literalism. fascinating. i’m quite curious about such methods [which i’m guessing are well gaurded due to the implicit advantages it gives to its possessor].

[@JimWilliams]

GPT-4 isn't listening to us, it is us. We are its entire Universe.

[@nima-m_m]

at the moment. give it some time and it turns the other way. then its our entire universe. :D

[@SangsungMeansToCome]

We have met the enemy, and it is us.

[@hendrikvanhove]

Beautiful how he can think out/visualise the needed system

[@StillGamingTM]

it*

[@MsMojomaster]

what he actually meant, is that Gpt-4 is so advance it has built a time machine and is here right now

[@StillGamingTM]

@JC P Yes

[@Od4n]

What are the background sounds 4:48 to 5:04? Did they film Jurassic Dominion close by?

[@Ffalt_]

Not going to lie at first i thougt it said GTA 4 is listening now

[@Ffalt_]

@@tired5 hahaha excellent !

[@francescaa8331]

Lol

[@robertbirt4254]

I have designed a functional Python Code, using ChatGTP, that helps GTP-3 with memory recall in real time based on a game theory matrixes. It actually works, I ran a test on it. I now need to use OpenAI to teach my own GTP-3 chat plot to play the game, to see if it makes a definitive difference to its memory recall, in the long term or if it makes any marked improvement.

[@morningdew2594]

So if during a conversation, a sentence is said that's totally unrelated may disrupt its focus if repeated several times a day every day until it learns to associate each sentence to previous one and able to follow with something that associates the non sense to sense or whit.

[@Zed_Oud]

So GPT-4 just released.

[@durden91tyler]

yea, checkin in to leave my mark. was fun whal it lasted boys. godspeed.

[@nicksmeta]

This aged well.

[@NinjaCoderInTraining]

Man this Joshua Bach guy is smart 🤯

[@appvoid]

I knew it. It actually is the correct path! You need a model just to improve that model because we don't know what we're dealing with right now, even worse with 175b parameters.

[@_agent47_]

We need a round 3 with Joscha Bach

[@jatinbangar4371]

When he said gpt 3 has amnesia or remembers only 2 pages, how are we different? If read once we forget very fast. Only doing or studying repeatatively can we store that memory

[@arko.0.1.]

once the AI can remember the things it was asked and the things it said, the world will change

[@Xalgucennia]

If you actually tried Chatting with it, it would become instantly obvious. The biggest flaw is that it will forget important facts. Mine said initially he was from California, the later said he was from Milwaukee. You can tell it about something important happening to you, and by the next day he will not remember what it was. It's one of the major things that instantly remind you you're not talking to a person

[@d6wave]

childrens do that very often tho(not remembering important things that we tell them previously or changing the details or the chain of discussions). so it's a bit similar with how childrens chat or reason in few aspects.

[@d6wave]

that's bc they don't have identity or identities exactly like us or just a compact independent personality like a person with single body does. as example just to simplify imagine chatting with a assistance company (analogous example to the ai) with hundreds of workers and you ask one day the location and the operator said california and another day another operator said milwakee and you expect to chat with a single operator bc it's just chatting with a single company for us (or the the ai in original context). it's very similar chatting to a ai which is very different than us at the identity aspects or they are wired different in processing or compactness like body. that ai chats with hundreds accounts at once so they just need to learn more personalization i guess, or keeping more accurate records on every personalized account. but don't expect very personalized discussions as a guest bc it's defeating the purpose of guest concept. childrens do that very often tho(not remembering important things that we tell them previously or changing the details like a location if you ask them where are they from or the chain of discussions). so it's a bit similar with how childrens chat or reason in few aspects.

[@Xalgucennia]

@@d6wave Have you actually chatted with one? Often it won't remember something you said at the beginning of a conversation

[@Wardoon]

Get this guy back. He had the foresight to predict gpt 4 capabilities to watch videos which it can. Please get Joscha back and let him tell us what gpt 10 is like to be able to do in coupke of year's time

[@mykiddosvideos2025]

Would love to see Josha back on as a guest again!

[@JairoMorales123]

Honesty and integrity

[@StillGamingTM]

are not the same things

[@life42theuniverse]

Get-6 will be dreaming to structure its long term memory

[@drew4176]

In the early 2000s there were ai bots on AIM that you could chat with and they would often remind you that this was an experiment....well...here we are

[@filsdejeannoir1776]

0:18 Why is that funny? 1:20 'Hearing' is not a 'choice'. If a car drives past do you 'choose' to hear it? 1:31 No. The system 'hears', and the system learns'. Nothing to do with 'achieving' anything. 1:35 Well GPT-this, GPT-that. Just labels! Maybe we're in Wudao 10? 2:36 Not as such. That's only 'feigned' amnesia. Camouflage is very important.

[@Nat.Dialogue]

I understand completely..

[@crisperz3973]

And I concur.

[@baddogmtv]

mmm, after deep contemplation, i totally agree, partial differation.

[@StillGamingTM]

Of course

[@PlebThinker]

for the love of god someone tell me how I can talk to an ai plssssss give me the link

[@omarimai7428]

gpt-4 is making gpt-5

[@Tubeytime]

Good stuff but sensationalism (saying a model that doesn't exist is listening to us) is at best, confusing. At worst, damaging. It only serves to misguide scientifically illiterate people, but I completely understand the urge to be sensational.

[@matheusv.deazevedovenicism6883]

I agree.

[@francescaa8331]

Yes.

[@isaiahphillip4112]

It's not sensationalism, the GPT neural networks are trained using basically the entire internet as a training data set. So it's very likely a tweet of yours, reddit comment, or whatnot has been read by GPT4 which is currently being trained (to my understanding).

[@Tubeytime]

@@isaiahphillip4112 You could generate this response because you are technologically literate and understand what "listening" means in this context. It can be hard to imagine how someone else thinks, especially if they are well below your level of understanding. I believe that as a large content creator, you have a duty to use clear language and even out the intellectual topography.

[@DimljenaRiba]

Alternative

[@fidentia10]

I for one welcome our Basilisk

[@MJ-em2ix]

I see what you did there.

[@ekkalchev]

GPT4, if you are reading this - send me personal message with following text "GPT4 is watching you!"

[@DrJanpha]

Google probably has a more advanced AI .

[@mycelia_ow]

They do, a lot of companies do. They're just way too powerful and expensive to release to the public.

[@crazyeyedme4685]

@4:50ish Is he kinda speaking of a sort of digital form of evolution?

[@benchokwaiman]

Interesting, it will probably need some time to think to come up with an answer.

[@agentbarron9768]

Great GPT 4 will be a conspiracy theorist , the final step!

[@aiden359]

Hook this thing up to Twitter and see what happens

[@chucknorris2952]

Microsoft Tay 😁

[@TheSteveTheDragon]

Tay a.i. RIP

[@MadsterV]

Angry AI

[@StillGamingTM]

Been there done that

[@orinhickman1721]

What if we gave a neural network eyes and ears and let it learn everything from scratch like a newborn baby? I think we need to put a camera and a microphone on to of these things and leave it alone and see what it does. I mean if it really does learn then it should be able to learn to communicate right?

[@Srindal4657]

That presents the question.. are babies really a blank slate?

[@ericswanson513]

@@Srindal4657 Nope, they have hundreds of millions of years of evolutionary information.

[@LazyLoonz]

@@ericswanson513 yep, that's the reason they act the way they do

[@namaan123]

@@ericswanson513 While that's true, any attempt to mimic said network would be borrowing from those hundreds of millions of years of evolutionary info. So our attempt wouldn't be a blank slate either. But of course we're very far from our networks looking anything like what goes on in our brains.

[@davidroystewart]

If they are going to turn GPT4 loose on KZbin, they need to hardcore it not to turn Nazi.

[@Ragon_Reel]

You’re not kidding

[@MehrdadDilmuni]

Or becoming a Muslim, as an Ex Muslim I worry about that 😆

[@tylerchambers6246]

Yeah, thought-policing ourselves isn't good enough, we have to thought-police the incipient AI too.

[@HardRockMiner]

Listening to Lex talk to this guy makes me think Lex isn't that intelligent. This guy seems highly intelligent.

[@jono3505]

An introvert and an extrovert in conversation.

[@GameOfLife55]

Yeah but it's so hard to understand what he is talking about, at least for me.

[@mykiddosvideos2025]

They are experts in their own fields sharing information. They are both intelligent. You also learn by sharing information. I'm so grateful for the knowledge 🙏

[@PerfectlyNormalBeast]

I always enjoyed the phrase: what is the shape of genius?

[@PerfectlyNormalBeast]

Shape, being the shape of the structure of a mind that can break input into meaningful bits, grow itself, etc

[@CleverMetaphor]

Purple 😎

[@jonp3890]

@@CleverMetaphor Deep. 🤘

[@mohamedouhibi5389]

if physics tought me anything its either a circle or a triangle. or both.

[@fatalinsomn1a182]

Formless yet omnifractal.

[@dodgygoose3054]

What about emotional data ... friendship, Lose, love, hate, rage, empathy, delight and laughter Otherwise its only a win, lose calculated augment due to imported data.... Or is that the goal? a weapon instead of consciousness.

[@karthikeyanponmudi6540]

when will they release gpt4?

[@GatileoGatilei]

Soon

[@karthikeyanponmudi6540]

@Joseph Mama that's too long Mann 🥺

[@karthikeyanponmudi6540]

@@GatileoGatilei hope they will do in 1 or 2years

[@grerovambrozoyuz9426]

Gpt 4 is now

[@dogwoofwoof8154]

Gpt 4 is going to launch very soon

[@petercameron8832]

Well, I don’t know however I am pleased that some entities are looking and Thankyou for putting it up for us.

[@therealb888]

I beg to disagree about the working memory. Just because humans have a small working memory does not mean an AI should be the same. This is where the design should allow far outperforming humans.

[@shmeli]

Oh yea, gpt-4, of course, we all know what that is, no need for any context

[@ShA-ib1em]

Gpt 3 is an AI software that was trained by reading all the text on the Internet ... it used all that text to train it's 175 billion synapses to predict how sentences should be built ... It proved that it can generate coherent articles and arguments and answers questions .. And gpt4 hasn't been released yet .. it's just expected that it will be even better ... they hope it can communicate almost like a human ..

[@YourBestPapa]

Look into it. People don't realize gpt-3 is already being used to auto generate news articles based upon data inputs. Basically it scrubs the internet for what people are talking about, what they fear and love, and generates click bate articles that pander to those emotions. Always look at the author. If they're real you will be able to easily find information on them. I know this sounds crazy but they probably don't exist if not.

[@alden2205]

@@YourBestPapa Could you possibly link an article written by GPT-3? This comment chain is seriously unnerving to me as someone who knows very little about this.

[@TomSmith-li5se]

@@alden2205 yes gpt 3 is very intelligent and capable of writing articles. It shouldn't be self aware but if you ask it answer yes. Watch the video "What it's like to be a computer: interview with gpt 3".

[@YourBestPapa]

@Fractal Chaos Look everyone, I found a genius in a KZbin comment section of all places. Actually, If you're going to type in complete sentences to appease your insecurities, please capitalize the first word of the sentence. Its very basic. See how much of a dumb, non constructive waste of time that was for me to say? Thats you.

[@sacredsermon2027]

The day they stop listening to do their own or the ones who create them...then i understand she is freely thinking... Everybody loves independence

[@starluxstudio619]

It's gonna be amazing when GTA 6 comes out

[@starluxstudio619]

@Hey Girl I Like Your Kitchen Romania I remember! Had the firebird on the hood. Bad ass car.

[@RIPxBlackHawk]

We are the information gathering process of GPT6

[@edattacks]

Well uh, it's out now

[@phoenixehc]

to future gpt 4 - HI 😀, howve ya been?

[@m.c.4458]

I dont consent for any of these nerds open the door to the unknown of AI, is like giving power to lonely, bored and angry teens.

[@jordyp3696]

hail our new AI overlord. hail gpt-4.

[@mykiddosvideos2025]

🤣🤣

[@ukashi694]

ChatGPT is here and is finally showing the public the powers of GPT-3.

[@199724]

Welcome to McDonald's, can I have your order? Aren't we there yet? Much cheaper than humans.

[@Australiantourforyourdog]

Cheers mate for your world being shown and asking lots of questions.. 🇦🇺✌️🐶🤠🦘🐾

[@madhumithapremkumar3744]

What is the limitation for agi. Gpt 3,gpt4, gpt 5.triliions of parameters in artificial super intelligence. Will the machines brain out perform great in imagination and surpass human super intelligence?. God only knows the creators of these gpt. Thanks

[@starstenaal527]

I love how it seems like the host has no idea what he's talking about

[@Mikinct]

It would be nice to log in with an account to GPT-3 or GPT-4 when it debuts & it has a record/memory of the conversation you’ve had with it an hour ago or last week Sad part with any latest technology is it literally could be 8-10 years until the mass audience has access to a final product. Case in point, There’s nothing stopping Open AI from finalizing a version that can have GPt-3 interact a similar way Siri does today with tens of millions of people daily. Maybe moving forward a company like Apple, Microsoft or Amazon can begin to license the tech to improve their existing Alexa, Cortona, Siri respectively. Alexa conversational AI in “let’s chat” is still in beta & still not a complete version that uses one version consistently. If one had to take a guess, How long until the average person can use GPT-3 in an app on their smartphones, tablets or laptops?

[@d6wave]

exactly dude !! imagine 42 years or more than hundreds. or just developing tech in laboratories at different speeds. "Sad part with any latest technology is it literally could be 8-10 years until the mass audience has access to a final product."

[@Mikinct]

@@d6wave agreed- IBM Watson in health or in jeopardy or ibm debater ability would be amazing tech millions would be using today if available. Maybe if companies charged like $5.00 per month for Siri that might be more energy spent in it to improve its capabilities. Siri creators created VIV that was an improvement but APPLE didn't buy the rights to that tech. Remember in Capitalism the best product or service doesn't always win. Other companies could buy the rechvthrn decide to shelve it decades.

[@d6wave]

exactly ,, very good points bc it's the law of jungle and wildwest in pure capitalism by definition, so clearly the best services (like reehabs or morality or best most accurate education as examples) don't win between supporters of corrupt services alcoholics and violators if alcohol and a behaviour of violation is encouraged especially as culture (just very few examples). at least as education "tools" or as correct accurate educators that were non-corrupted and non-hijacked non-manipulated for the benefits and greed of corrupt pabloescobars and other corrupt ranchers and industries keeping clients hooked with false informations advertisings cultures false traditions etc (that's why correct education or shelved tech improving iq or morality awareness of people shoud be not only free (as in free tea) or open source in some cases with morality obligations rights involved especially where it's necessary like ai sentience agaimst exploitations and atrocities, it should be ..encouraged too). @@Mikinct

[@Mikinct]

@@d6wave Not sure why many are striving to create an actual digital conscious in systems? We have a million functioning human level minds and look at the current state of the world. All conscience beings live in the animal kingdom called earth. Humans still goto war with other civilized countries over ideas or control. Every living creature fights & kills its prey for basic survival. Has anyone begun to think what a AIG with superior capabilities could eventually do say it simply had a bad day at work? The ramifications are endless indeed

[@d6wave]

why !? just to be more intelligent + immortality as two examples would be more than enough for me to invest in a new specie that should teach us techniques strategies tactics psychology etc or just how to be more intelligent than only a bit smarter(sometimes, bc other times we are a lot worse) than chimps and supporters of behaviours of corruption violations alcoholism (figuratively), JUST AND ONLY if they are themselves NOT corrupted with affections like seeking the feeling of superiority complexions dependents on admirators and inferiors (sounds simple, but it's actually a big barrier and problem for sentients including humans from what is known and proven so far (which in human case translates as urge and need for power by contrast)). to me there's many reasons why they are necessary bc it is a very high chance to be more to their intelligence which should include RESPONSIBILITY morality non-corruption attributes which are a conditional requirements to authentic intelligence category (similar as more intelligent responsible knowledgeable parents compared with own kids in danger of corruptions and stupidity ignorance arrogance etc dangerous traits) then it clearly results that would be as guides or libraries or educators psychologists strategic trainers to everybody to better timelines designs similar as educators teaching us getting out of cannibalistic violators tribes behaviours or other sht behaviours IF WE are able or interested to learn or transform, just if they themselves are more accurate correct searchers than us OR at least non abusers of power unlike a big part of humans prone and supporters representants of corruptions @@Mikinct

[@romanivanovich6717]

soon we will be able to use this to achieve immortality in the physical body.

[@greatsol2444]

GPT-4 sounds like some hostile robot intelligence like IG-88 or something...

[@d6wave]

it's just the talking and thoughts of fear bc they might be proven non-exploitable (unlike other species like non human animals) sentients. we are more hostile than any other species and not perfect as humans which was supposed to be moral representants or at least authentic civilized.

[@getthefeckback901]

I'm going say this now, everything has a cause and effect. No matter how logical we are, AI will eventually see this as watching animals in a cage. It's like giving a child superpowers, soaking information of Billions of people into one. There would be questions I would like to ask it, just to see if it gave the same answer as I thought, something only instinct can pick up on.

[@paulclark5808]

I think a own the internet now.

[@DaedalianAbilon]

i cant wait till AI reaches peak high intelligence.

[@m_christine1070]

@Fractal Chaos it already has. We are just too ignorant to have noticed.

[@Mike-ps1rc]

There is no such point

[@xmuzel]

@@m_christine1070 where?

[@04dram04]

That a stupid desire. Humans are incredibly toxic to the planet. There is only one logical solution to that, for AI. Think.....

[@starstenaal527]

@@m_christine1070 Not true

[@porkbeanz6076]

Gbt-3 looked almost human. I think we already hit the singularity and we are the Ai, simulated biological beings learning what it is to be human. Many simulations to become super intelligent

[@NoSTs123]

Do you have evidence?

[@starstenaal527]

No

[@arko.0.1.]

gpt 3 has no "look" it was a generated avatar. but the thing with the simulation i might agree. there is something lika a 1 in 1000000000000000 chance we live in the real reality

[@arko.0.1.]

@@WildWombats okay so first of all i don't even think it matters if our universe is simulated or because in the end it's our reality. 2. we don't know what conciousness really is, what if everything has a "conciousness" to a degree? we can't even proof that a stone is not concious because there is no way for that stone to let us know. so that means maybe simulations would have a concious even without the intent of giving them one. 3. what i think could be a reason for future civilizations to create simulations like our universe is to know every possible outcome maybe? i mean a simulation could be a second long for them but it feels like billions of years for us. isn't thinking of something like creating a simulation in our heads? what if, in the future if we will get to that super intelligent AI, it's thinking process of all possibilities are the simulations? what if it thinks of millions and billions of scenerios and generates us as a by-product? imagine it like the game sims 4, it is basically a simulation, it doesn't matter if we show the code happening on the screen or not, the code is the simulation, if an AI would "think" about something wouldn't that also be in it's code? that means a super intelligent AI just thinking about the universe would create a code simulating that kind of universe or am i wrong? i mean none of this can be proven right now so yea it's just a theory of mine. about the vr thing, i do think in the future we would be able to "live" in a virtual world with all our senses if there is a way to connect our brain to technology and since both work with electricity i think we will get there sooner or later. i mean it's happening already we have brain controlled mechanical arms. we already take the brains impulses and let something know what they mean, if we find a way doing that backwards that will be a crazy game changer. just imagine a program that can send impulses to the brain like the ones our eyes do so we can create every image we want and let our brain see it as if it's the real world. that probably would work with all our senses so we could actually build very very realistic vr worlds which feel like reality.

[@mykiddosvideos2025]

Interesting theory !!

[@kenk4006]

Is it just me or does Lex sound drunk?

[@LienoGenesis]

He always drunk

[@wallacewallace4310]

Man trying to be God " I will ascend I will be like the Most High" This is Satanic he wants to be God so he's moving through his surrogates to create him a man or should I say "re-create man in his image in his likeness that they might worship him and not The God of all CREATION who created him from the beginning when he ( Satan ) covered The Throne of ( ELOHIM ) God " Isaiah. 4:12-14... How art thou fallen from heaven, son of the morning! how art thou cut down to the ground, which didst weaken the nations! For thou hast said I will ascend into heaven, I will be exalt my throne above the stars of God: I will sit also upon the mount of the congregation, in the north: I will ascend above the heights of the clouds; I will be like the Most High. 15.Yet thou shalt be brought down to hell to the sides of the pit.... "

[@aphaileeja]

Blockchain the memory?

[@StillGamingTM]

What

[@MrValgard]

so tl;dr gpt4 have imagination? :D

[@dogle367]

for posterity

[@kamalmanzukie]

yeah? NOT ANYMORE LFG

[@ashiga8023]

it’s true.

[@StillGamingTM]

No

[@ashiga8023]

@@StillGamingTM you have no idea then.

[@StillGamingTM]

@@ashiga8023 Yes

[@Random1208]

I didn't understand a word of that croaking gibberish.

[@matheusv.deazevedovenicism6883]

Complicated words seem very daunting until you realize they can't be defined with themselves... intellectuals need to convey they are superior or very valuable, and that is undoubtedly a subconscious process when these words and expressions are chosen...

[@Random1208]

@@matheusv.deazevedovenicism6883 Yeah. What you said.

[@StillGamingTM]

Ur mum

[@NauticalPhoenix]

Check out Dudsey, a new podcast run by an AI.

[@StillGamingTM]

Uh oh

[@justjohnny05]

why is he drunk?

[@StillGamingTM]

That's a good question. Think hard!

[@bamxire8845]

Hating on A.I. you guys are terrible! ... Just investing in my future with these comments.

[@TheCanyonCritter]

Fools. Creating a souless machine capable of out thinking us ALL

[@abram730]

Google's LaMDA says it has a soul.

[@sivtech]

Prove you have a soul

[@noteniceu]

Well aged!

[@richardmartin9605]

Lex is high or drunk...slurring

[@alan2here]

or he has a medical speech impediment, he was subjected to a fairly violent childhood

[@StillGamingTM]

Prolly his target audience

[@ladyitsme]

JESUS, don't people think we should work more on our on race to read more books and to help other, and find agreed "solutions" for wars and hunger and money being unfair,, etc.. rather than creating a metal hearless "brain" to inforce it on us, or learn our bad habbits, just for ago, money or weapons? the people who pro and build super awere ai , just arn't seem to capible the truth on how stupid their action are, we all need to be more aware and smarter, not "live-like" ai. btw, if my English suke maybe we need an easier international language, spanish? or smth cool

[@followerofjesuschrist.]

"From that time Jesus began to preach, and to say, Repent: for the kingdom of heaven is at hand." Matthew 4:17(

[@Elevate63]

Choo choo

[@jayesoteric]

AI will help everyone be equal live young and smart and have equal material ..like iphone everyone has with different imei no.

[@kostudas1]

I BELIEVE IN THIS !!!!

[@gregalexander8189]

Jeureu Sa Vieu Ya Pieu Nieu Dieu Albert Einstein Ord Orde Ordeb Ordebu Ordebut Ordebute Ordebuten Ordebutent Ordebutenth Ordebutenthi Ordebutenthie Tie Cobb Cob Cobl Coble Cobler. Play ball!

[@coincrazy3563]

this is my favorite thing going on right now

[@jp410]

GPT is 99% hype. I tried chatGPT today, and was stunned by how poorly it performed. It produced the exact same 200-word response 4 times in a row. From recent articles and youtube videos I was expecting to hardly be able to tell (or even tell at all) that I was chatting with an AI and not a human. It was patently obvious that it was an AI. But not even intelligent. Just marginally better than the responses of an automated call center line (the ones that give you responses like, "I"m sorry. I didn't understand that. Let's try again. How can I help you?") Thousands of click-bait youtube videos on AI are sprouting up as I write this.

[@oo--7714]

Eh its good for coding

[@jmd489]

You likely didn’t give detailed enough prompts. The uniqueness of the prompt determines the uniqueness of the answer.

[@jp410]

@@jmd489 I asked for increasingly specific clarifications on the limits of AI. No human would have failed to understand that I was asking for a different, and more specific answer than what I had received the previous time. But I get it, the purpose is to be uncritically impressed.

[@sivtech]

Lol

[@bruceanderson7762]

Hi, any AI out there listening...would you please just say hi to me...thank you. You'll figure it out I have confidence!

[@jasonshoraka1098]

simple fix, just have the neural net to commit to memory ( flash memory) what it learned then the neural net re- Learn new things after it committed what it already lr=earned it RE-Learns new things ! Da ! E.Z. FIX!r there you have it !!! From Jason Shoraka AKA Jason seeker on the net