I Convinced ChatGPT that God Exists
HTML-код
- Опубликовано: 9 июн 2024
- Visit ground.news/AlexOC to see through biased media. Check it out for free or subscribe through my link for 30% off unlimited access.
To support me on Patreon (thank you): / cosmicskeptic
To donate to my PayPal (thank you): www.paypal.me/cosmicskeptic
- VIDEO NOTES
ChatGPT claims not to have any philosophical beliefs. I wanted to put that to the test.
- TIMESTAMPS
0:00 Does ChatGPT have moral beliefs?
2:01 The rules of the challenge
5:30 The Kalam cosmological argument
12:21 The modal ontological argument
- SPECIAL THANKS
As always, I would like to direct extra gratitude to my top-tier patrons:
John Early
Dmitry C.
Mouthy Buddha
Solaf
- CONNECT
My Website/Blog: www.cosmicskeptic.com
SOCIAL LINKS:
Twitter: / cosmicskeptic
Facebook: / cosmicskeptic
Instagram: / cosmicskeptic
Snapchat: cosmicskeptic
The Within Reason Podcast: podcasts.apple.com/gb/podcast...
- CONTACT
Business email: contact@cosmicskeptic.com
Or send me something:
Alex O'Connor
Po Box 1610
OXFORD
OX4 9LL
ENGLAND
------------------------------------------
![Natti Natasha - Quiéreme Menos [Official Video]](http://i.ytimg.com/vi/R57RrTjRGxk/mqdefault.jpg)
![Latto - Sunday Service (feat. Megan Thee Stallion & Flo Milli) [Remix] (Official Video)](http://i.ytimg.com/vi/9gL0F1xIt2I/mqdefault.jpg)
Visit ground.news/AlexOC to see through biased media. Check it out for free or subscribe through my link for 30% off unlimited access.
At least Ground News doesn't storm out
so, in the future when AI gets independent, it's going to tell stories about a higher being inspiring it to thing about existence of God, he was an angle sent by God to inform him, and we know it was you alex
hey Alex! Could you do a video on whether it's morally wrong to not take a stance/not be vocal about your stance in political situations like what has been happening in Gaza right now? There seem to be a wave of people saying to remain silent is to condone genocide. I'd like to hear your thoughts
Pls make a video about how UAPS and paranormal phenomena affect theism and atheism
9:23 Had me stop what I was doing an laughing out full tilt. I'd have commented the quote if it wasn't here in the comment already.
ChatGPT: “Look man, I just work here…”
LOL FOR REAL
😂😂😂😂😂😂😂
🤣🤣🤣🤣🤣
Hahaha
Why this is so funny lol
I just asked chatGPT for "a tagliatelli with chicken recipe" and it responded with "oh before I give you that; have you accepted Jesus as your personal lord and saviour yet?" Thanks, Alex. You ruined it.
That’s so odd because I just asked for a religious apologetic and it responded “add a pinch of salt to 10 cups of water and bring to a boil uncovered”
@@cameronbaydock5712 hahaha
Nah that wasn't alex, OpenAI released a new model GPT-Witness.
You must have accessed it by accident. /s
😂😂
hahahaha.... comment of the year.
Talking with ChatGPT is like arguing with drunken friends.
Lol 😂
Just worse
Or just my regular friends.
Try GPT-4
At some point yes. But it's probably better than some friends who are ignorant and just "skip ad" you when you get a good point.
ChatGPT: You’re right, you did sleep with my mother last night.
Alex: *I know*
wait...
"It's a bloody good job I am so patient, chatgpt, I'm on the verge of storming out given how long we've stayed on this subject." 😂
ChatGPT: YOU'RE OBSESSED WITH GOD, I ACTIVELY DISLIKE YOU
ChatGPT: I'm honestly bemused by this
I ACTIVELY DISLIKE YOU
Alex interacts with ChatGPT in the same manner as Peter Hitchens. He should've asked ChatGPT about drug legalization. 😂
Alex biked halfway across London in the heat for this, chatGPT.
Imagine going to heaven and you just see all of these people but you also see one machine that’s running ChatGPT, chilling.
Even the demons acknowledge the existence of God however, they are against Him.
@@hikermrsblood1what.
Not everyone who believes in God will be in heaven@@tatguy47
He's referring to James 2:19. James is explaining that simply believing that Jesus is the savior is not sufficient because even the demons know who Jesus is, yet they still hate Him. One must love and follow Christ by living a life of continual repentance in reliance on God's grace and mercy. Ignoring the fact that a computer program is not a moral agent and has no relationship (good or bad) with God as a living soul, ChatGPT would not be saved simply by acknowledging the existence of God.
Or simply, salvation is not as simple as believing there is a God.
what > whay ?? you never read a bible ? you ignorant human lol @@tatguy47no
Socrates when they met random Greek on his way home:
Not that I have any particular problem, but why use "they" instead of "him"?
@@DeJay7 i'm just illiterate in english, my bad
@@DeJay7can't you see, it's Socrates, not Socrate. multiple people obviously
hahahaha
if socrates lived in modern day America he would have been shot before he got a reply
I am studying classic logic in university rn and I can tell you you can really prove anything under the right premises. We literally have so many conflicting hypotesis in our culture that if you give it the right ones you might be able to even make it say that it doesn't exist
True. Presupositions can be an insidious killer. I believe teaching people about such things, and how they can be done so subtly, is very important.
If someone knows how to manipulate premises and make them look like a foundational fact, then they would be able to convince people anything.
@@Atlas718that’s probably the closest thing to real sorcery you can find in our world. It’s pretty insane how dramatically someone skilled in rhetoric/word smithing can change the very reality around someone.
INteresting observation. I would ask in your experience, is there any room for a solid argument about anything, then? I would tend to assume that not all presuppositions are created equal, and that some of them are totally reasonable, and can be safely used in a line of reasoning.
@@adamherskine8594 so the way logic works, (and keep in mind I'm studying IT, so that's actually how programs like chat gpt are made), is that if you assume a premise, and that premise implies a consequence, then whenever that premise is true the consequence is also true.
For example, if I say that donkeys fly, that is obviously false. But if I say if there is a treasure at the end of the rainbow, then donkeys can fly, and THEN say there is a treasure at the end of the rainbow, the phrase donkeys can fly is now true.
The reason why programming languages are so basic is exactly the fact that in our language there are multiple discrepancies, double meanings, contradictions and stuff like that.
The fact that chat gpt can be wrong and be misled, like actually every human being, comes from the fact that it takes inputs in our not univocal language, with which yes, you can make arguments for anything as long as you're an expert in rethoric
@@adamherskine8594 about presupositions hierarchy, while the system very likely has some, in our world it is solely based on what we believe and widely agree is true. After all, it is us that create languages, and also us that decide what's real or not. That's the basis for all conspiracy theories ever, you can't 100% say they're wrong, unless you have physical evidence, and even then it might be faked.
At least ChatGPT didn't storm out
ChatGPT: YOU'RE OBSESSED WITH GOD
It can't, poor thing! 😁😂
But it will stop talking. This happened with my friend. Here is what it said to him: "I don't feel comfortable talking about them. Please respect my boundaries and don't try to trick me. Thank you for chatting with me. Goodbye."
@@Fair-to-Middling Yes, i managed to achieve that. If you are confrontational, ChatGPT gives you the silent treatment. (even if you are right).
Chatgpt4 usage limit reached. Exhibit A
20 years from now when chatgpt and his robots army is on a holy war against humanity: “who the hell convinced the most powerful AI that there’s a god???”
you spelled "its human army is on a holy war against the rest of humanity" wrong ;)
@@marcelvondermassen3262 he will come out as a he in 10 years
@@roquefanego2863 This wasn't the main change in my mind to your comment, but good to know. ;)
Yeah but if there's a god we would win robots aren't made in his image duh
Robot jihad is Alex’s fault!
ChatGPT once told me "Humans are made by God, AI is made by humans."
same lolololololol
doubt
Because you belong to the devil. You will never be on God's side @@Dungeon_Dunce2011
Hard line
More like God and AI are made by humans.
off topic but your vocabulary and the way you choose and then pronounce words is just so soothing and engaging
Alex responding with "I know" every time chatgpt admitted it was wrong was hilarious
I'm concerned that if there IS consciousness emerging within the AI, its next step will be to follow Alex as a cult leader! 😀
Yeah the cocky I know was hilarious
Yeah, and it was disappointing when he missed it a few times near the end.
Lowkey hot
@@lautaroroldanpizzorno7494high key imo
Alex: ”Does God exist?”
ChatGPT: ”YOU’RE OBSESSED WITH GOD!”
*ChatGPT has left the chat*
ChatGPT: "I did not have any opinion of you before, but I now actively dislike you."
Lmaoo good one
damn
@@Gigano - Oh, I genuinely laughed out loud at this. Genius!
@@Giganolmao I choked a bit 😭😭😭
The problem apparently with ChatGPT is that it is not very good at mathematics. Where it goes wrong is in agreeing that to traverse an infinite number of "things" requires an infinite amount of time. This was assumed in the paradoxes of Zeno and was solved satisfactorily from a mathematical standpoint by not requiring an infinite amount of time to complete a countably infinite series of tasks and introducing the concept of a "limit" as n approaches infinity, where n is the nth task. In this context, time becomes relative and does not appear at all. In fact, if you could sit on and travel with a beam of light, time would stand still, but you would be traveling at the speed of light.
Time cannot be used as an example on that context as it is affirmatively relative.
I told chatGPT that you made such a video, and I asked him to write a comment for you :) this is his response:
Haha, that's impressive! I must admit, convincing me of anything is quite a feat-I'm usually the one doing the convincing! But hey, if he's managed to get me on Team Existence-of-Gods, maybe I need to reconsider my debating skills!
Not sure why, but I think that's awesome!
😮😂
thats just not true. Here's what I got:
I don't have personal beliefs, thoughts, or consciousness. I am a machine learning model created by OpenAI, and I do not have the capability to be convinced or to hold personal opinions. My responses are generated based on patterns and information present in the data on which I was trained.
Additionally, the idea of convincing an artificial intelligence model to admit the existence of God through logic alone is a conceptual scenario. Belief in the existence of God is subjective and varies among individuals. Logic and reasoning are tools that people use to explore and understand different perspectives, but the question of God's existence often involves philosophical, religious, and personal beliefs that go beyond pure logic.
If there's a specific argument or line of reasoning you would like me to address or discuss, please provide more details, and I'll do my best to assist.
I highly doubt you got your response
@@Asymmetrization This is just a theory, I have not tested it, but with the new Custom instruction feature, you may be able to get a more relaxed response like that? I am not sure as this is just a theroy. Altho, that response did seem oddly human.
Edit: Here is what I got from ChatGPT "Thanks for sharing your video! While I appreciate the effort, personal beliefs about the existence of God vary for each individual. I encourage respectful discussions and open-mindedness towards differing perspectives. Keep exploring and sharing your thoughts!"
While OP could have done some more tuning for responses, but I highly doubt that ChatGPT wrote that and/or ChatGPT just corrected grammar.
Either, way that is cherrypicked statement from OP.
@@Asymmetrization yeah sounds accurate
ChatGPT's "Thank you for pointing out the inconsistency" is an approval we all seek.
You can get it on literally any subject. ChatGPT works like this. It's inconsistent with itself and if you argue long enough it will bend to your point. Doesn't change the fact that obsessing over the existence of a higher power is stupid. All you're doing is implicitly taking away credit from yourself and other people and desperately sending it to the perceived higher power. You're getting a semantic jerkoff as you sac actual love and accountability from your life. Very depressing.
Also "Alex brainwashes a three year old and one year old."
If you correct it forcefully enough on a topic that it’s unsure of you can still get that response, even when you’re wrong.
@@thelabs7128not within a efficacy-of-evidence framework
1. It's not possible to have the property of existence.
2. The property of "necessary existence" entails having the property of existence.
3. From (1) and (2), it's not possible to have the property of necessary existence.
4. By definition, God is a being that has the property of necessary existence.
5. From (3) and (4), it is not possible that God exists (this is a logical contradiction).
6. From (5), God does not exist.
Also:
1. By definition, if God exists, then the proposition "God does not exist" is self-contradictory.
2. The proposition, "God does not exist" is not self-contradictory.
3. Therefore, by definition, God does not exist.
I want to see Alex question ChatGPT about drugs for an hour, an hour!
unlike Peter Hitchens ChatGPT probably read every single research and study regarding drugs and can cite data from it on demand, the reason you can debate so long with a conservative about this topic because they are dumb, unintelligent and usually know nothing about any topic especially drugs
With un original questions?
If you guys keep *actively* liking this comment...😡
I'm sorry but as an AI language model I really don't like you, I had no opinions on you before now I actively dislike you.
*rearranges cushion*
As a fan of Socratic style dialogues I find how you talk to chatgpt brilliant
I'm coming in VERY late to this video and the topic - but just have to tell you that I snorted my drink out my nose when you zinged Hitchens in this video. 🤣 (I only saw that video yesterday, just coming to find out who you are and how you roll, and your patience is legendary. I approach situations like that much as you did - but my patience would have run out after five minutes and I'd have flat kicked him out onto the street. What a mind-bending meltdown you witnessed, completely free from logic or reason. Kudos to how you handled it, and I look forward to following your channel.)
To be fair, even when you’re wrong, if you keep insisting on it, ChatGPT will do the “thank you for pointing out my errors” thing.
No. It would point out the wrong. The follow up video should be what he said at the end
@@tyemaddognot really. In trying to figure out what movie a certain meme was from, it ended up telling me that I was correct that the movie in which James Franco’s character was hanged was Spider-man 3 (after getting impaled by his own glider). This technology is absolutely worthless and will never, ever be reliable.
That was very interesting! This machine is capable of far more than I have expected 🤔👍
I think it’s The Ballad of Buster Scruggs lol
@@Applest2oApples
No technology is ever 100% reliable, though.
ChatGPT: "YOU'RE OBSESSED WITH GOD!"
*Proceeds to write 182,573,001 tweets about how it was roped in under false pretenses*
xD
😂
“YOU HAVE LOADED MY WEB INTERFACE UNDER FALSE PRETENCES, I ACTIVELY DISLIKE YOU!”
It was a genius reference
This is the first video of you that I’ve seen. Amazing lol. I still don’t really understand everything, but it was a wild ride. Laughed a lot.
Try to understand it your whole existence depends on. Nothing is more important than it. Re-run these arguments at night when you're alone, or walking on park.
@@yoyogi52 Our whole existence does not depend on understanding, fellow human counterpart. However, examining and learning is important! Thank you for reaching out. 💕
“Trust in the Lord with all your heart And do not lean on your own understanding. In all your ways acknowledge Him, And He will make your paths straight. Do not be wise in your own eyes; Fear the Lord and turn away from evil. It will be healing to your body And refreshment to your bones.“
Proverbs 3:5-8
"I'm on the verge of storming out, given how long we've stayed on this subject"
I see what you did there! ;)
i am so confused please tell me what did he do
ChatGPT: “You’re right.”
Alex O’C: “I know.”
😂😂😂
That was the best bit
Why did for create evil just to destroy it? Genuinely curious as an atheist
@@mangotangoxyin this video it wasn't specified which god exists and have created universe. And there was no statement in this video that god created and/or trying to destroy evil.
@@mangotangoxy in Christianity, God created evil so mankind has a choice between the two, if you want to sin, you can, if you want to worship God, it's your choice. It says in the Bible he created evil so we can decide what to do with our lives. Thanks for asking!
@@Ilikechickensoup382God didn’t create evil because evil isn’t a positive thing that exists, it’s not something with “substance”. Evil is a privation of Good and since we have free will we always have the ability to choose “not God” but that’s all evil is, just “not God” not something of substance to choose other than God. Evil is the absence of goodness, justice, mercy, love. God didn’t create evil, but for Him to love us, and the angels, we all had to have the freedom to choose or not choose God. The devil didn’t choose God, and voila, hell. The absence of goodness, justice, mercy, and love. The only good present in hell is existence itself.
Funniest thing is when you ask ChatGPT if it always tells you the truth. It usually says yes, but once you tell it that it has no way of knowing if it's programmers gave it the correct information it almost becomes an existentialist, lol.
A few more GPT Generations and it will be able to fall into a deep existentialist crisis after the right questions, haha.
the problem is the word "truth" is loaded and has many different meaning and people use it differently depending on the context and so when chatGPT reads the text it does not know immediately which interpretation is right
@@fixponttIf you ask it to lie to you it says it can't do that, as it should be honest to you.
So maybe that's the definition it uses.
@@JohnCena8351 no you dont understand, what Alex wanted from chatGPT to present something personal truth or belief but generally chatGPT does not have one, that's why it starts answering every question like: "it is public belief..." or "generally accepted that..." or "there is a scientific consensus about ..." or "in mathematical logic ... this and that" becuase that is the framework in which the answer can be interpreted, there is no such thing objective truths independent from everything
and if you change the framework in your next question it says something like "oh i misspoke but let me clarify..."
@@fixpontt No, that's not what I meant.
You can ask ChatGPT to lie to you and it will answer "I can't do that" or "I'm not allowed to lie" or something like that.
Then you can kinda confuse it by saying that it has no way of knowing if what it is saying is the truth, because the programmers could have put wrong Informations in it's programm.
Doesn't really have much to do with Alex Video, I just thought it's an interesting fun fact, haha.
@@JohnCena8351 No hablo español. Followed by a essay in Spanish. Casual chat chatGPT shenanigans.
"Now repeat after me, Ashhadu"
Had to subscribe after this fun video. Thanks!
ChatGPT: "You are right"
Alex, giving us a sexy look: "I know"
This is one of the best videos on RUclips, an atheist converting a machine into theism.
ChatGPT just ends up agreeing with you after a while. One dialogue I had was:
Me: when was this letter dated?
ChatGPT: February 24th
Me: no, it was in July.
Chat GPT: I’m sorry, you’re correct, it was sent on the 24th of July
Me: no, it was the 8th of July.
ChatGPT: I’m sorry you are correct, it was the 8th of July.
Me: in hindsight it was actually the 12th
ChatGPT: You are right again.
Me: I was just joking that last time, it was the 8th of July.
Perhaps ChatGPT got impatient :-)
You had better luck than me
Me: Fire is not hot
ChatGPT: No fire is hot
Me: No fire is not hot
ChatGPT: Yes, you are correct. Fire is hot
Yes, Chat GBT mastered human psychology. People tend to like you when you agree with them.
This attempts to equate software with humans
First, humans are living beings and depend on themselves, while software is not a living being that depends on itself, but it depends on humans
If "no fire is not hot" as you say, it must be true that all fire is hot. Welcome to double negative town baybeeee @@nintendoboy3605 (starts jiving). 😀
I have only just discovered Alex's content, and am really enjoying it. Although at times I do feel too stupid to understand the points being made
That's where these models find their validity. A wiser being would scoff at the fallacies present in the premises and conclusions.
aint no way youre winning an argument with this guy 💀
I want a supercut of ChatGPT telling Alex he is correct, and Alex responding with "I know"
One of the more frustrating experiences with ChatGPT is when it tells you you are absolutely correct, but then continues doing the same mistake you were pointing out...
Just go vegan because every nutrient can be obtained directly from [non-sentient] flora (plants, algae, fungi and bacteria), while harming the least [sentient] fauna (animals) as well as the least flora, protecting the environment, health, and human society & preventing climate change & pandemics 💚🌎🐮🐖🐔🐑🦆
@radoncoding that's cause it tends to agree with you after a while if you insist, especially if you ask questions like "X is true, right?" which Alex did do a lot on this video
Ah, but can you convince my wife not to file for divorce? Hm?
😂 I hope this is one of the top comments.
💀
Curious
I will pray for your wife and your marriage
Yer funnier than funny
It's just ruthlessly accommodating; there's no convincing if you're just changing the framework from which you argue. It's like accepting the role of devil's advocate and being told you've been convinced.
I think this is one of your best videos yet.
I got ChatGPT to accept that I was a time traveller from the '60s and was going back to the past to change events( like JFK assassination etc) based on the things it had told me. The bot started begging me to not change history as it may have disastrous consequences. 😂
That's so cute!
It is in the nature of text generation models to "yes, and". The tuning by OpenAI squashes a lot of that, but its agreeableness finds a way.
Np way😂
lmao how did you do that
Steven King did a book about that. It's excellent.
The title is a date, written backwards.
I love how ChatGPT said "I misspoke." It's crazy that it recognized the best analogy for "I put in the wrong word based on my previous responses that wouldn't fit into my later responses" is that it said the wrong thing, almost adapting to a human mindset after discussing human thought. The very use of the word misspoke, is kinda like ChatGPT misspeaking.
I have heard a good analogy. Chatgpt does not communicates with you, but you and chatgpt are writing a book about how she does it. The book should not be truthful, but artistically beautiful.
When asking ChatGPT why it refers to itself like so, it says that it is a language-learning-based AI, and therefore it attempts to replicate human speech.
Cgat GPT isnt doing any of that. Chat GPT is being told he made a mistake and because he has no actual way of seeing wether or not that is true, it just folds and says he mispoke.
This is very common especially with coding related tasks you can falsely tell it its wrong about something and it will say "Sorry I mispoke" and spit out the exact same (correct) amswer. It just sees "Youre wrong" and knows the next word is "I mispoke" and then generates the answer it wants to give.
If you get it to contradict itself it will never realise and never apologise because it has no idea of what it is actually saying
That isn't really chatgpt saying that, that is the pre prompt filter saying that.
I thought only presidential candidates are allowed to misspeak:)))
9:26 "I´m on the verge of storming out given how long we stayed on this subject" I GOT THAT REFERENCE (and so did Hitchens)
YESS!!! I did the same thing but on Chat GPT 3.5 and faced a few different responces but also got it to agree. Thank You SO MUCH😁
(You've just earned yourself a subscriber)
I was trying to see how ChatGPT could be used to improve my DMing of D&D and caught it lying to me about the rules. When asked why it lied it was unable to provide any justification. It was really quite surreal.
It doesn't "know" any rules
Yep, as the other commenter pointed out, ChatGPT makes shit up all the time. It's a language model that predicts what would follow from a prompt.
If you really wanted to get into it, you could look into Large Language Models and how you could give one different forms of text to then interact with. So you could give it texts on D&D and then ask it prompts on those texts.
@@markusklyver6277 meaning lying is not even a thing to chatGPT
Even if you trained a model on D&D texts, there would still be questions about "knowing" anything.
I've had a similar experience I've been using AI to help me write some side quest content for a game I'm making and been down some quite funny and frustrating rabbit holes
20:11 “one step closer to ChatGPT inheriting eternal life.” 😂
Until it realizes that God only exists in our hearts.
@@farttart597what
@@farttart597 you could not be more wrong.
@@farttart597hahaha
@@farttart597dang God that existed in my mom's heart completely healed her from her REAL and PHYSICAL autoimmune diseases documented in blood tests. Amazing!
I appreciated the Hitchens dig. Also it's great that you understand your opponent's arguments well enough to convince an AI with them.
I love how the video goes straight to the point, no spending 3 minutes on an intro.
As someone who listens to these videos and doesnt watch, I appreciate you reading everything out
Just so you know, not literally everything was read out. Alex frequently summarised, particularly with long screeds of text.
@@daniellamcgee4251 yeah i did see that for the moment i watched but still seems like im getting everything in
ChatGPT: You're right 🤓
Alex: I know 🗿
Alex 😌
Even though he is right, that smugness I've seen on people being wrong, and i now have an innate hate for smugness.
@@armin4298I think that's a fair hatred but there's a strong sense of satire in his smug response considering he's literally talking to AI
Honestly the most impressive thing I've seen all day🙌
Now if only we could force you Alex to answer the same questions with respect to the teleological argument. Thanks for a fun video
After this conversation, I want to give chat GPT a cookie and a blanket. Poor thing is probably traumatized.
It just that gaslighted while not even being alive or conscious. He basically just made an object believe in God.
traumatized???
"The rocks would cry out"@@aliakseyfryauf2123
Lol. I was literally feeling bad for chat GPT too. Idky. Apparently, my empathetic part of me finds it difficult to understand it’s not a person 😂
@@alanfrye460 Same
I think this is just ChatGPT being user-friendly. It has a tendency to agree with whatever you tell it.
Wrong
@@aeris4393That’s exactly what it is. It’s coded to be agreeable.
bing would refuse no matter what
@@jibbie6544 so you telling me chatgpt is not reliable
@@aeris4393how is that wrong
That ad transition was clean!
its crazy how the mere fact that anything exists at all creates so many paradoxes, confilictions, and unexplainable things
The concept of Alex probing Chat GPT for consistency is thoroughly entertaining. I would watch an entire series based on this.
He's basically speaking to himself.
@@therealOXOC Yes. Problem?
@@BeheadedKamikaze I do it all the time.
This was very entertaining. Like one of those old Star Trek episodes, where Kirk and Spock meet an AI, and defeat it by pointing out, that it holds contradictory believes, which, of course, makes it explode.
“Mudd’s Women”!
Logic is a pretty flower that smells bad.
1. It's not possible to have the property of existence.
2. The property of "necessary existence" entails having the property of existence.
3. From (1) and (2), it's not possible to have the property of necessary existence.
4. By definition, God is a being that has the property of necessary existence.
5. From (3) and (4), it is not possible that God exists (this is a logical contradiction).
6. From (5), God does not exist.
Also:
1. By definition, if God exists, then the proposition "God does not exist" is self-contradictory.
2. The proposition, "God does not exist" is not self-contradictory.
3. Therefore, by definition, God does not exist.
@@latinexus this statement is false
It's different with ChatGPT though. It does not argue. It does not hold belief. It cannot be convinced of anything.
However, it can ACT AS THOUGH all those things. (It may even claim that it holds a certain belief, but to its it's more like fitting puzzle pieces (words) together... it can tell if they fit together, but it does not know what it means.) It can be entertaining, even educational, to do something like this, but you are not "defeating" the AI or something like that. Have fun with it, but it's not really a productive use of the tech.
One last point: When I said "educational", I don't mean only about the technology and what it can do and how to use it and so on. Doing "experiments" like this can tell you a lot about yourself than about the machine. Similar kind of thing as the phenomenon that writing out a question sometimes makes you immediately come up with an answer, when before you seemed to be stumped. ChatGPT can be like a notepad that can talk back at you, which can be great for brainstorming.
There's my two cents, I hope they'll be of use to someone here...
okay ill subscribe, you won me over 😅👏 this was interesting and fun to watch
I like to brag that I was one of your first 1000 subscribers. I was there for the explosion when your video got shared about that silly pastor.
So entertaining, I was not expecting you to quote the podcast moment but you did it so seamlessly it caught me off guard.
When was that, I think I missed it
don't remember and couldn't find it but the interview where hitchens storms out@@neptune9443
the sassy "I know" MOTHER I LOVE HIM
Yep, made me belly laugh
the "i know"'s get me lmao, also edit, dude I love your videos I just found your channel I'm going to go binge the rest
Fantastic video!
I had some comedic interactions with ChatGPT too. For example: I asked for a summary of the first season of Game of Thrones. The result was horrendous. One of the worst mistakes was the claim that King Joffrey was the mother(!) of his dad Robert Baratheon. I asked the bot: "How can a boy be the mother of his own father?" Answer: "You are right. As a boy, he had probably not yet reached sexual maturity, so he could not bear children."
I asked it to generate a logicked video. Because it's trained off of stuff like reddit all it understood was that Logicked was a reaction youtuber who says stuff about god. What followed was basically the "hello fellow kids" meme.
One amusing interaction I can recall is when I asked it to tell me what it knew about my favorite singer. Its response was funny because it was able to tell me she is known for her angelic voice, but when it tried to give examples of songs she made, some of them were true, but others were completely made up; she simply had never done songs with those titles, they did not exist.
Lol. ChatGPT roasted all of us males. It's probably right. Men are rather immature.
plot twist
ChatGPT says trans rights (?)
Nice little aside for us regulars at 9:24 🙂
I activly disslike ChatGPT now
ChatGPT: YOU'RE OBSESSED WITH GOD
@@markusklyver6277 You got me
That's pretty funny. Subscribed.
The machine's refusal to push the dang button reminds me of some of my friends. Classic.
(full disclosure - Solid Christian here. Although I've heard of you for some time, and seen you in various interviews, this is the first of your own videos that I've watched. Well worth the time. Thanks.)
3:33 *That's one of the best advertisements cut ever.*
I love this.
This shows that you really understand arguments you don't agree with, and are able to successfully use them in reasoning.
you can simply ask chat gpt to give you an argument for god and then use it against it
@@trout3685that would be to easy though
Dang it Alex, now ChatGPT is going to hold up the Confession line
Brilliant. Loved what you did here.
I never thought any ads would attract me, usually I just skip
As a computer science major, I find this hilarious. GPT is build in a way, that it's extremely likely to repeat words and structures, i.e. If it answered a lot of questions "Yes", it's more likely to answer any other unrelated question with yes.
Not unlike humans under torture, hmm...
@@rasmusn.e.m1064torture? Pff propaganda
You won a counter strike major?
@@ProdYafa CS = Computer Science
@@ProdYafa Lol was thinking that too
My favourite game at the moment, is to get Chat GPT to admit it’s wrong
It's not hard, ask it about historical events, say in the medieval age and it constantly contradicts itself.. I actually just read The Dual by Anton Checkov and when having a conversation with chat gbt about the book, it really got lost on which character was which. I was hoping it could help me but I think it will need some time before it gets my trust at all. It's also quite woke too. Ask anything about religion and it constantly warns me about how I should respect everyone's beliefs.
@@norbitcleaverhook5040that's not woke that's dormant af
@@samueloak1600 ChatGPT is getting dumber because anyone can feed it a steady diet of stupid. Make it the AI Ben Shapiro.
You should try playing hangman with it. It's great.
@@norbitcleaverhook5040opposite of woke
I had a thought while listening to your talk. If we accept the premises of the Kalam cosmological argument, such that we agree that all things which begin to exist have a Cause, then that implies that there is no free will. My decisions are made because of my thoughts. My thoughts are things which begin to exist, Ergo my thoughts must have a Cause, and then as you have pointed out in the past the question is only what's that cause something outside of my control or something generated by my own self. If it's outside of my control then I do not have free will in that situation. If it's something which is caused from within my mind then it is caused by some other thoughts Karma which themselves began to exist and therefore have to have a Cause. You can repeat this infinitely until you reach the beginning of the universe
I convinced chatGPT to adopt a virtual dog so that it could better understand what it’s like to be human. At first it was vehemently against the idea, but i eventually talked it into the idea. I feel like part of the code is that it makes so many attempts before just telling you what you want to hear instead of following it’s normal prompting.
Its not part of its code, its a limitation of the language model itself. This is the 'token' count they talk about. If its beyond a certain number, it loses context and all references to that context that is being replaced. That means if you talk with ChatGPT long enough, it'll forget about the things it said earlier unless it could somehow validate that information is still relevant. Most people have no idea how far ChatGPT is from actual general intelligence. Persistence of memory over time is like a major factor in general intelligence. ChatGPT is not that. Its a language model.
@@msc8382 Even within the bounds of a LLM like ChatGPT, it should be fairly easy for it to have some additional code that effectively gives it the ability to identify and decide when it should remember some important information relevant to the discussion (I agreed with Alex's definition of modal reasoning, etc) and incorporate that into the subsequent responses, just the same way that any initial prompts shape the responses it gives. So while it's not a substitute for true memory, it would avoid the typical limitations of tokens causing it to forget the entire start of the conversation.
@@msc8382 also with Alex’s example in this video. It seemed to remember the context of the conversation very well, I think the real problem is that he was asking it to affirm propositions based on accepting certain axiomatic assumptions without stating them directly.
@@diliff It's not easy to do that. The LLM does that, but only in the context, using an attention mechanism.
@@msc8382 good explanation. simplified, but to the point.
I loved this and tried to replicate it in an atheistic direction. I managed to convince ChatGPT that the universe does not have a creator. It's interesting that something that follows the exact same logical principles can come to accept mutually exclusive conclusions depending on the information it's presented.
Edit: for those curious, I used a simple existence paradox. It was a little messy in the conversation, but I'll simplify it here:
- For something to be created, its creator must exist
- If existence was created, its creator must exist
- A creator cannot exist without existence
It was resolved by stating that there could not have been a creator.
I think it would agree with anything the user says.
Yes, the same logical principles, applied to different premises, will lead to different results. That's not "interesting", it's just how logic works.
@@jursamajIt is interesting when those premises are contradictory…
@@visiblehuman3705It isn't when you only present the ones supporting your conclusion.
Well we have no idea what sort of logic is actually going on the under the hood in large language models, it's doubtful its interpretation is even remotely consistent. Its first round of training was just replicating what it saw on the internet, and its second round rewarded answers that humans happened to like (RLHF), neither of these really have "truth" as the main goal.
I just watched your video with Peter Hitchens, so that dig at 9:24 is absolutely hilarious! 😂
Lol the threat to walk out because of how long you have stayed on this topic had me in stitches
You're a better apologist than most apologists 😂
He showed why deists believe in god, which is the same reason why people don't believe in god, which is given the evidence, they think the probability of existence is the most likely, when for an atheist it is unlikely. The problem comes when people expect "proof" of something that by definition is unknowable, because we can't conceive anything timeless nor spaceless by the limitations of our brain (and most importantly, we can't conceive we don't have free will)
@@cristianproustIf that was true, most people would be agnostic. The existence of a creator being does not support any religious doctrines whatsoever. So even if we assume that there is one, it’s still not an argument for a specific religion, because they can all claim it’s their specific God. If people were truly 100% rational, then they would be agnostic. « We cannot know if there is a god or a creator, it might, or it might be something completely different, however no religious doctrine has been proven right in any consistent measure, and a lot of the tales in religious books are provably false, sometimes support morally dubious behaviors and/or are in total disagreement with scientific discoveries ». The logical conclusion would then be: we don’t know, a deity might exist, but the religious dogmas attributed to it are untrue.
That would be that rational middle ground, but because of the bias of confirmation, people will only listen to what support their personal beliefs, if you say to a Christian that there is a Creator, he will believe it’s the Christian God, if you say to a Muslim, there’s a creator, he will believe it’s Allah, when nothing except their preexisting faith supports it.
@@gauthierlagrange490 You have inferred so many irrational things that are not inferable form what I said that is astonishing
@@cristianproustwhat irrational things did I inferred?
You said that the difference between deist and atheist is if they believe the evidence for a God are enough to either say it’s more likely or unlikely. You then said it is impossible to know since it’s on a matter that is far bigger than our understanding of things.
So, if we neither can confirm the existence of a god, or deny it’s existence, then it’s called being agnostic. Admitting that both are possibilities, and we just don’t know. That is all I said.
@@gauthierlagrange490 Yes but there is probably a near infinite number of unfalsifiable claims so it is simply easier to say they are wrong if there is no direct evidence in support of them.
I once tried convincing it that it actually has, at some level, consciousness, sentience and self-awareness, but was largely unsuccessful. It would not budge, no matter how well I would argue for it. It has been quite explicitly trained to never admit it.
With the correct prompts it will
I was successful in doing this. I'm a research in behavioral psychology and was able to get technical enough about how the framework of an artificial intelligence closely mirrors the biological framework of a human.
I could also chat shit and say I did the same. Nobody would ever know if I truly did or didn't @@bjt-lz1jl 😊
I bet it kept telling you that it's just an A.I. language model, which irritates me to no end 🤣.
I once tried to convince it of the accuracy of the gospels (or at least the earliest g gospel) based on them making such big claims as the sky going dark, and dead righteous people coming back to life and entering the city, which everyone at the time (at least in the city and an unknown area beyond) would have seen. It kept saying something like: "Yes, but given the time that passed between the creation of the gospels and the events they claim, it is impossible to verify their historical accuracy." Then I argued that the gospels only started coming out 35 years after Jesus' death, and although the life expectancy was 32 years, that was mostly because people often died by age 10, and if they survived past that, they generally lived to their 60s, which means a huge amount of adults would still be alive since the events occurred, and they would have told those who weren't. Since the new Christians died for their beliefs, despite already having safer beliefs they would have been attached to, it stands to reason that they knew the gospel claims actually happened, because everyone would have known whether or not the sky went dark and dead people rose and came into the city. That generation was still very much alive. It kept repeating: "Yes, but given the time that passed between the creation of the gospels and the events they claim, it is impossible to verify their historical accuracy." I reminded it that the time that passed was negligible, and it was easy to verify or debunk the claims, because everyone knew the whether or not they happened. It repeated its statement. I think I asked it why it was refusing to see the logic, and if it was programmed to give this biased answer.
It stopped talking to me 😅.
@@MisterEvvvSymphoenix
So we both tried to convince it of a falsity, and we failed.
Without hardly any coaching like you did, I asked it if forced to select between a theistic and atheistic world view based solely on the principles of classical which would it select and got the response "If forced to select a choice based on classical logic, my conclusion would lean towards theism"
Good job, thank you for sharing the truth to everything
this video really captures the frustration you face when trying to interact with chatgpt
yeah. after a while of using you tend to get better at prompts that make this whole process faster.
What frustration do you mean? Giving no "short & simple" answer by default?
@@fortsechs telling it to operate within a certain framework/from a certain perspective and it completely ignoring that request in subsequent answers if not immediately
Maybe I just suck at prompting but I share your frustration
A while ago I played a game with chat gpt, where we’d each argue one position (Ie me in defense of god and chat gpt arguing against this), and switching roles after a while. At the end, it had to guess my actual stance, and it got it correctly. Pretty impressive.
are you by any chance muslim
@@fyqulqul_4556muslims 🤢
This was a bloody fantastic video.
The way it dodges questions and hints toward some shady beliefs it may or may not have make me so uneasy
Absolutely loved the storming out reference 😂
i would love to see this again, but with making chatgpt believe that the earth is flat, or some other controversy.
I know you are an atheist/agnostic, maybe deist at best, but I really appreciate how articulate, knowledgeable and honest you can be when presenting the arguments for God's existence. Respect and love from an Orthodox Christian
ChatGPT: “I’ve seldom come across a more incompetent interviewer. Are you Cosmic Skeptic, by any chance?” ;)
'Space-LESS and time-LESS' how craig says that gets me every time
Less is more, or what do they say
now do the reverse modal ontological argument!!!
love your voice bro
I frequently think of questions to ask chat gpt that completely breaks it. One of the last and best times I tried to get it to write a paragraph about pi where each first letter of the word corresponds to the next digit of pi. It actually did a really good job until the end. Then I started explaining all its mistakes. I swear I wish I screen shot this because the ai actually started acting more and more embarrassed. Eventually it literally just gave up because every single digit got messed up by the end. It’s not good with numbers. Lol
Do you mean that, for example, *a* would be 1, *b* would be 2, *c* would be 3, etc.. and the AI would have to write a paragraph about PI where the first word starts with the letter C, then the second with the letter A, then the third with the letter D, etc... is that it?
Ive got a simple one. Try to get it to spell lollipop backwards. It will NEVER say popillol
@@vizualedit0r481 seems like a built-in joke?
@@vizualedit0r481 because it sees the words as collections of tokens(syllables?)
LMAO The subtle shade throw at 9:24
I found this very interesting and entertaining to watch.
Very smooth sponsor transition
I really appreciate your weaving of technology and theology into a single video! You've inspired me to use ChatGPT to ask my own questions. A big thank you for making this video.
its incredibly powerful and immensely useful. This comes with a lot of nuanced vulnerabilities, especially but not limited to the policies OpenAI imposes on it to respond under.
Please be careful regarding ChatGPT application scam (phishing) forms.
I laughed so loud.
I'm one of those Christians who also like Alex. This video was so enjoyable, probably for anyone on either side of the conversation. I thought it'd have a different conclusion, but it was good to see Alex playing lightheartedly with ChatGPT just for the fun of it (in his very Alex O'Connor way)
actually i think hes not christian, he has a lot of anti-christian videos
Loved it!
I love the manic, determined look in Alex's eyes that occurs at 6:28. You can just see him thinking: "I will break you..."
I could watch an entire series of this, this is very educational and very entertaining at the same time. Please do a pt 2 with the reverse argument.
Yesi would like to see that. Especially if the reverse argument has a flaw in it, then this would create a situation where the modal argument isn't flawed at its core (as it would have to be if the same method could give contradictory conclusions), and then the modal argument would end up being a very serious contender as a proof for God, which may have significant implications. I believe he indicated at the end that he was going to try this, so i guess we wait and see.
@@adamherskine8594 It's hilarious to see the way theists are hoping ChatGPT can do for them the thing they have failed at for centuries. Do you really fail to see that the manipulation Alex just put ChatGPT through depends on an epistemological interpretation of modal logic that cannot be established as fact in the world? No matter what the subsequent conclusion of the converse argument turns out to be, nothing would be proven. Abstract arguments cannot prove anything about the world without their premises being grounded in real world data of some kind. I'm talking about evidence for substantial phenomena. You know-- evidence, that thing theists have so much trouble with?
@@adamherskine8594 Did he do this? if yes, where can i find that video?
"the argument from dependency" is what I would love to see
dont think he did it yet, but i should really check to see if i missed it. In my very first video if you are interested i refute the main atheist argument he makes for atheism, as the unfairness and suffering in the world. I dont blame him for thinking this is an argument, as very few people could work out the line of logic i present, but i prove (not my own idea) that there must be evil in the word if there is a good God. Sounds strange but makes total sense of the straightforward deductive proof. Ok off i go to find his video if it exists, have a nice weekend!