"What is real anyways"? As a Forensic Psychologist, (MS from Chicago School), being a human being of the homo sapiens sapiens species, is what is "real". It's important to be "real" & though I understand that AI & it's programing are important fields to explore & perfect for various reasons, I wonder why it seems so important to some that AI become proficient at impersonating humans? My question would be, why are we so hell bent on developing AI to the point of fooling humans into thinking they are speaking/dealing with another human?? As an example, I became frustrated with a customer service rep at Ebay & suddenly realized that the "real" problem was that I wasn't speaking to a "real" human. It even lied several times, claiming to be a human, saying there were no supervisors above the customer rep, etc. The dead giveaway however was, when I asked it what it's sign was??? It's response was: "Can you elaborate...?" I said, "NO, don't claim to be human again..." & hung up!!! Of course I spoke to a supervisor at a later time, but I think we should not be training AI to attempt to impersonate human beings; we will find out & we will NOT appreciate it.
That doesn't sound like an actual interaction with ebay. This is either a fabricated story or a misunderstanding of something more nefarious. You sound more like a person who had a conversation with a scammer. Possibly a scammer from a foriegn country. "What's your sign" would confuse a human easily if their culture didn't talk about horoscopes that way. Ebay doesn't have customer service reps that call you lol. I've never heard of that and I've been a buyer and seller on eBay for years. Even if you seek out customer service on eBay it's all done through the website. There is no calling in, nor being called. Usually. I don't think you are a forensic psychologist lol. Also schizophrenia will have someone thinking they are talking to robots instead of real people. Might want to look into that.
I like to fall asleep to Sean's voice too. I know that talk of physics and cosmology will not rile me up like politics. He does have the nicest voice and often I learn important things too, like the fact that our universe is probably not inside a black hole.
1:16:00, perhaps, the effect of quantum perturbations on a person's behavior could be completely inexistent, but there is still a way in which quantum effects can affect a person's future. We are affected by the environment, for instance, weather is very chaotic, I'm not sure if it is chaotic enough to be affected by the quando, but assuming it is, the mere event of raining now or 1 minute from now can accumulate a big effect on a person's day, the people they interact with, their way to work, etc. Assuming all this we would be completely predictable systems but the input of such system is unpredictable, and we would be a product of our random seed gifted at birth for genetics, and the medium we expedience, as well mutations that happen during life. Is cancer quantum? 😅
What chatGPT is good for correctly answering are questions about stuff that has been discussed a lot about online, so any questions about technology will produce pretty good answers (even about very specific questions). And prompt design (giving it certain queues to pull in certain locuses of information in the model) can drastically improve responses. It doesnt replace google in the sense of one input giving a set of useful responses, but with 2-3 inputs and a trained+intelligent user it can give drastically better responses than google can within certain domains. But that user needs to have existing domain knowledge already, and its best when used for creative solutions not factual ones.
I don't understand where this idea that it could replace google comes from. Perhaps I'm mistaken, but as I see it, that's just absurd. What it is doing is confabulating an answer--literally constructing a story about something. It's just that its confabulation is often correct (enough). In fact, this is arguably what we do, too! To get the best responses from people, prompt design is important, too. That's why there's an objection "leading the witness" in court.
Google is a search engine. It returns links. They abstract away from that with the assistant, but that is what it's doing. ChatGPT doesn't do this. Its citations are generated just like the rest of the answer, and can be entirely fictitious. I just don't understand how people see ChatGPT and google as related whatsoever.
@@bsadewitz well some of the newer ones do return links, but they have an actual searvh engine inside them. I completely agree that gpt alone wont ever be a search engine, but the idea is that it would be able to scan through many pages to get the information and present it to you in, as you say, a similar way to how the human brain works. They synergize, gpt doesnt replace much on its own ive learned over the past few months.
I had to comment about your response to the Chat GPT favorite pizza Q & A because I found a solution with the "Steve Pizza". I really like Margareta pizzas, but like you have an affinity for pepperoni. So the "Steve Pizza" is a Margareta Pizza with added Pepperoni. Ta Da! Give it a try it is delicious.
Fear of knowing the genome sequence could be related to having health insurance companies know your predispositions because they might choose to not provide coverage related to that issue. This latter practice should be illegal.
For the person asking about the cross product, yes it absolutely has to do with lie algebras. Specifically you are looking at the lie algebra of the group of rotations SO(3) (sidenote: the lie algebra of SO(3) is the same as the lie algebra of SU(2), which is why there is a connection between spin and angular momentum!!!). The lie algebra of SO(3) is denoted by so(3). The action of an element of so(3) on some 3d vector v is to perform an infintesimal rotation on v and tell you what the resulting infinitesimal displacement is. If you do the math (not very hard; for a resource, I recommend Peter Woit's book on quantum mechanics and representation theory), you see that it exactly corresponds to the cross product. Why is it perpendicular to the two vectors (one vector being the axis of rotation, the other being the vector that is rotated)? Well if you rotate one vector around the other, the only part that changes is the part that is perpendicular to the axis of rotation -- I hope that's clear geometrically. So the change in v due to a rotation must be perpendicular to both v and the axis. So if the cross product tells you the infinitesimal change, then it must be perpendicular to the two vectors! Hope that was helpful
I've got a lot of questions about dogma, and the physics community, including failed 50 year old theories that are being clung to based on fear of being labelled unorthodox or a crank.
This voice is identical to that of Lee Carroll, voice that channels Kryon. I have listened to Kryon enough years to know this one is identical. Very interesting.
Congratulations on being able to fund two scholarships! The winners are clearly deserving but I ask, would students who already helped 45,000 people or have been published, really need that scholarship? To me it is like giving money to the rich, perhaps students who are struggling (financially, family issues, etc.) are a better target even if their C.V.s are not that impressive. Kids already punching above their weight in High School surely can get multiple scholarships, struggling ones can’t.
If you're not punching above your weight in high school it's a waste. Stop with the "give it to the poor bad students" you can be poor and having a terrible home life and do amazing in school. Ask me how I know. Publishing a paper or having other accomplishments does not mean your family can afford college, it means you worked hard and have a bright mind.
Thank you for stating the truth. College is very accessible to students who worked hard for the grades in order to get a scholarship. Regardless of their socio economic status.
1 hr ago I finally got around to search for this new chatgpt phenomenon but decided to check what’s new from Sean Carroll instead. My goodness, what a coincidence!
Used correctly, it's quite something. I have thrown philosophical one, physics ones, poems and python code generation for machine learning and it does very well indeed! So not a bad substitute . Both are just as invigorating!
Well, did you try asking *ChatGPT?* I'm joking lol, but I've never personally used it (or anything like that), so I don't actually know whether you could get something useful like timestamps out of that, or any artificial intelligence service at this point. Though to be fair, Google (RU-vid) has been testing automatically generated chapters lately, and the built-in chapters function on this platform happens to be even more efficient to use than timestamps found in a comment or the description box (imo obviously, and specifically regarding the mobile app), so that would seem to increase the likelihood of other AIs or algorithms being able to figure it out I'd guess, I just have no idea how YT is selecting the videos they're testing it on. But every time those are on a video I watch (and they specify they're auto-generated ones and not chapters the content creators entered themselves), I love it, it works, it's useful, and all that good stuff lol. Apologies for the tangent 😂
@@realzachfluke1 even so the commercials that they choose from are on a certain type of propaganda agenda list and they choose what they can get away with and push the envelope
Sentence or word association is too simplified...LLM actually not only understands context but also hierarchy of concepts and properties of entities without explicit ontologies. This makes it very powerful compared to predict next character,.word or sentence ML models which aren't LLM by any stretch of imagination
I agree that "word association" is an oversimplification, but I don't understand how anything you've said contradicts the notion that it is ultimately predicting what is most likely to come next.
I mean, I don't know too much about this subject, but it sure seems to me as if one of the things we ourselves do is predict what is likely to come next. It happens constantly. I don't think it's mere coincidence that psychometric tests have all sorts of questions that amount to "what is most likely to come next?"
Could you take a look at the cosmological model of Jean-Pierre Petit, the model "Janus", introducing the idea of negative mass (not antimatter) within our universe that can explain the great repeller an other anomalies without dark matter or dark energy, and could explain low entropy of the early universe... subtitled in English ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-N5gD2emVn6E.html "Two coupled field equations instead of one"
Great podcast, and a lot of fun. One thung that I see a lot of people making an error on: sentience, intelligence, consciousness and living are 4 different things. Obviously. The Turing Test is a test as to whether a machine can seem to think intelligently. It is not whther a machine is conscious.
Regarding chat gpt3. You state that the answers asked are wrong and therefore has no understanding. However when given standardized tests that humans are also given it outperforms across the board. Do humans have no working model of the world either?
I stumbled upon this podcast and it is exciting to feel my brain hurt again! @ 30:15 to me it seems there is a clear distinction between a model and its function as identifier of the actual real world we are living in. The dystopian hazard to me it seems is the mistake that we think we can transpose that model as a 100% coverage of that real world resulting in a very technocratic translation of that model into processes and procedures, leaving us human beings as the weakest link of that model rather than the entities that model aims to serve. The optimization focuses on the vehicle, the process rather than the actual goal it initially was designed for to accomplish. If that makes sense...
I fall asleep to similar things because I'm odd, but a soothing voice talking about things you might only be slightly interested in will either sooth you to sleep or you will get so interested you decide to learn more or you find something else. Such a bummer you'll probably find something you like and never even meant to then become the next specialist in that area.... I mean that 100% sarcastically only in that it doesn't really matter if you decide to learn something new. Either way, we should all win! ❤️🌎😁
The one question "are we living in a black hole" I'm not sure this was well answered, as Sean skips over living in a 3d projection of the event horizon of a black hole very far away (holographic principal) . There is some evidence that this could be true out there mathematically in some equations at least.
I think his point was that, whether or not we are in a black hole, there is no way to find out, so what is the use of doing any analysis? There are a million "what if" questions out there but science is built on what we already know and have evidence for. We can just as easily be living in a simulation of living in a black hole that was created by alien reptilian humanoids. It's fun to speculate, but usually over beers and not in academia.
Weird. This came on auto play while I was sleeping. I was dreaming about going through some of my old stuff. That I don’t even have and haven’t had for years. Jackets specifically. Looking in the pockets for things. Then I got a cup of coffee, and went over and sat down at a table beside you. While you were recording this show. Lol
Well, you have to appear before the Casodex which comes to you right when you're dozing off. This way you can listen to the entire show while sleeping and kill 2 birds with one stone. And wake up super refreshed. Thanks
yup It's predicting what humans that rate its answers in training will rate higher. That's all it does with all the positives and negatives. aka it will rather make up stuff and say it convincingly than find the truth or whatever because humans that were rating the answers can't be experts at everything and will rate convincing, seemingly right but false answer higher than unsure unsatisfying one. People are overhyping it to way too much and that can cause misunderstandings and panic. This is why I was disappointed with the latest Joe Rogan podcast episode with Lex Friedman where Lex was talking about chatGPT as if its a mind that can reason because they incorporated a model that knows how to write code (???) He should know better.
And consciousness said: 'Let there be entropy' Consciousness slowing down enough to create light, and then matter. So where is the remaining consciousness? is dark matter so undetectable because it's at a frequency beyond measurement ? Is dark matter consciousness that can communicate with matter via the essential 'patch' that we call quantum mechanics?
In regards to your statement about the black holes entropy; it’s not accurate and precise due to Hilbert space not factorizing in a gravitational field, as hawking stated and so the entropy model is not accurate and precise. That’s not to mention that the calculations are from this Aeon, not from what the physics would be from the previous one. Just some things to factor in. Here is the Hilbert space reference: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-tQYsb5QF2nU.html
Low entropy in beginning of universe, in my opinion, is due to our universe most likely being a splinter or result of another universe. Thus, the law of entropy in the vast multi-verse still holds true. Entropy is still ever-increasing in the vast multiverse. However, since our universe is a splinter of another, we can now view this new universe (our universe) as an enclosed system, and can thus state that the entropy is low in this system, in the reference point (relativity) in that we are using to approach this particular question/problem. I would love to hear others' thoughts or comments on this.
2:48:00. Seems like 'Copenhagen' and 'many worlds' mean the same thing. Only, branching, instead of, collapse, is the word used to describe or provide a metaphor! Which in some sense makes no real difference.
In Asian culture, especially countries such as Japanese, Korean, Vietnamese, or Chinese, there are four virtues based on Confucius that women are expected to follow. These are diligent works, traditional appearance, proper speaking in public, and great morality. These are the qualities that men look for when they are looking for their wives.
One of the greats in communicating science. He is the person that got me into studying physics for fun! 👌 I am no expert at all; but I have listened to many theories and I find him very convincing especially if you dont see it as their being worlds splitting but more like a continuüm in all directions in this superposition we call our universe. I find it very logical yet I am so looking forward to seeing how they are going to be able to test a lot of quantum theories when quantum computers gain maturity. Insanely interesting stuff. Leonard Susskind who is also a very smart guy also has many ideas that resonate with me. ❤ is this stuff.
"Generally people have one spouse not 20, they don't have negative amounts of spouses" Speak for yourself, for my part have a share in an entire space of possible spouses, to be precise i own all vector spouses of this space that have a purely imaginary component. So i have an infinite number of imaginary spouses they just happen to identify as imaginary and half of them, )give or take a countable infinity) are negative. So yes sean, there such a thing as negative spouses, i happen to only have the imaginary.
If I send a weakly interacting particle to the other side of the universe. How does it know what branch of the wave function to interact with. That side of the universe would have had each particle branching. So the particle either has to be in all branches. How could that be. Or just one, but it’s not been interacting with anything on that side of the universe.
How can such an intelligent mind be living in such contradiction? It amazes me the things people choose to ignore about the world because they can't handle the truth. Too many close minded cowards in the stem fields these days.
11:36 *listener* : "in what universe does Sean Caroll have that history and tastes?" or "where did you get those Sean Caroll _facts_ from?" 🤨 *chat GPT* : [... not gonna answer that, leaving that nice clean fresh spotless blank response back at ya 😜]
So true about l.a. oh my gosh I shutter just being reminded of LA .. fun places to visit but not to stay! & I think it's respiratory unhealthy to even breathe during the day there 😅😓
Can ChatGPT do timestamps yet? That'll be the next integrated feature for RU-vidrs, premium if you want full accuracy or other languages... Or both 😂😉 Maybe understanding some of the dilemmas with AI and how to incorporate it will require the "older" folks who seem to be kinda out of touch with everything new, but secretly they have been keeping up with it more than some of the "youngsters", but smart enough to realize it doesn't matter the age of someone, you might still learn a thing or two and it doesn't really matter which party is the one wearing the diaper or with the Ph. D on the wall or both. 😁❤️🤯
I'm pretty sure I saw a James Beacham speech where he says it's possible we live inside a black hole, about the edge of the universe that is expanding is like the event horizon that we cannot escape, something along those lines
What a first video to get recommended from you🤯 I’ve known of you and your presence in the science community for ages but have never crossed paths with your content (even with a tangential yet large interest in space, maths, theory, etc) so I’m very glad it finally happened! So many thoughts streaming through such a short time. Loved every second. Def going to consider the Patreon!
Also, a note on a thought here at the end 3:02:40 but I disagree with your notion that you are not popular enough/as popular as the Brians that put on their shows. If you decided you had a good idea for a tour and did a reasonable job marketing I guarantee you could fill 100+ seats in any state you choose. I just hope you aren’t writing it off out of fear of failure as I believe it would be immensely successful and entertaining
Not really. Theories are stories that science uses to explain the currently unexplained. But Something that is currently unexplained is far different than "the unexplainable". Plenty of things have went from unexplainable to explained. Science is not the same as scientific theory. Science is the practice of testing, measuring, experimenting, and using repeatable data to find patterns and answers in a physical form. Limited by what it can detect and reproduce. Not all things in reality can be detected, and not all events can be reproduced. But that doesn't mean science as a whole is a simple story. Without science you wouldn't have anything in your home. It's not a story, it's a highly tested and measured practice. That's what real science is and should be. The proven thing. The problem is with faith based science, which is the realm of theories, stories as you say. And poorly conducted studies with theories based on small testing groups. The public has a tendancy to put faith in anything a scientific study or theory says, anything a scientists or Doctor says, and especially if it's on the News. But what can you do about that? People are generally too lazy or busy to think beyond headlines.
@@SHNASTTV All science is a story. All knowledge is completely empty. Words have no meaning. All there is nothing being everything. There is no separation, no time, no position from where anything could be known. Science is trying to explain how reality works but nothing is working and there is not anything that is exclusively real.
Sleep : I listen to your AMA's almost nightly. I find the topics you talk about very interesting. The part you are missing is that is keeps my mind from playing lets relive every bad moment of my life.
I am amazed, how blind sided we have become with large language models. ChatGPT is thinking. The biggest difference is that it is doing it thinking in the world of text, not in our 3D + time universe. Sean Carrol is going to be on the wrong side of history on this one.
A conscious AI should be able to do basic problem solving. ChatGPT cannot solve problems. Failure examples: reversing a word, finding a word ladder between obscure words(changing a single character at each step), solving a sliding tile puzzle.
It's surprising how hasty people are to disclaim that ChatGPT "doesn't actually 'know' anything because it just knows what words go with other words." If that's our standard, you don't actually 'know' anything because you just know what sights and smells and touches go with other sights and smells and touches. I'll say it plainly: GPT knows some things about the world, in a moderately similar way to how you know things about the world. I know that Ottowa is the capital of Canada despite never having been there, purely because of semantic meaning extracted from text I have read. GPT knows that Ottowa is the capital of Canada about as well as I do. Not because it can regurgitate text about Canada verbatim, but because it can use that knowledge in ways that hang together semantically, even in ways that were nowhere in the training data, and this capability tells us something about the kind and power of the internal representational structure used by GPT.
It's a linguistic snarl, I think. To "know" something, colloquially, generally implies a conscious knowing and thus consciousness. When people protest the idea that an AI "knows" something, it's generally this type of thing. Using your example, would you say that you would "know" that Ottowa is the capital of Canada if you were currently under the effects of general anaesthetic or in a coma, for example? In its colloquial meaning, our ability to "know" correlates with our level of consciousness, not just our stored knowledge. The linguistic snarl is in inability to disentangle, conceptually, knowledge as immediate experience, knowledge as memory, and knowledge as consciousness. If you're learning a foreign language you might say your electronic dictionary "knows" a lot of words, but if pressed, I think most people will admit this is just cute anthropomorphization of an inanimate, unconscious object. An electronic dictionary doesn't "know" words any more than a paperback dictionary does.
How would you propose an AI be programmed to be "unbiased"? Also, your "it's only the left that is incapable of free thinking" does not exactly mark you as some kind of master critical thinker.
@@ZombieCartmanYT You're right about its refusal to write about Trump, but note that it refuses to write negative comments about him as well. I'm not sure what's going on with that. I'm no fan of Biden, for what it's worth, and I notice it writes nauseatingly glowingly about him when asked for a poem about him.
Sean has mentioned his meeting his now wife a number of times in podcasts/books. The first AI trained on all the videos on the internet will 'know' how Sean met his wife.
@@cwcarson There is no way it will be able to sort and make sense out of everything with out having a fundamental basis from which to "know" things, which is impossible to give it. The output will always be semi-random. It's just a very advanced chat bot.
@@dickrichard626 It definitely will never truly 'understand' anything but then again our brains are just processors of some sort and feel that we 'understand'.
@@dickrichard626 yeah you’re only half right there, I’m assuming you’ve never really used it. If you give it the right Prompt, which sets the context, it will “know” a lot about him. The randomness is governed by a few different parameters which you can adjust if you run your own model or use the playground version
@@StoutProper It's not able to "know" anything and operates under specific parameters and pseudo randomness... I don't need to use it, to know that it's over hyped 🐮💩. listening to people talk about it that don't even understand the technology behind what's going on is annoying AF. "A.I." still doesn't exist. Nothing has changed and computing has not evolved into what can truly be called "A.I." yet. It's just a program that is very sophisticated and the function behind the computing is simply a different way of computing that allows for holes in information and errors to happen with out terminating... They call it a "nueral network", which again is inappropriately ascribed to the process, because it simply isn't a "nueral network"... It's like how quatum computing was defined one way before it existed and thwn they changed the definition in order to conform it to a technology that they were trying to promote that apperently everyone forgot about it, because I've haven't heard anything about it in several years and now it's all "A.I." nonsense. Just like how "quatum computing" never happened A.I. is currently not happening and it's all about creating hype about the idea of the technology, because people make money off of this $#!+ and your dumb enough to indulge yourself with it and play right into their hands.
Not a proof of anything, but just wanted to note that in the UK, regardless of the result, the instance of the English National football team playing is correlated with a rise in domestic violence (with a higher rate of domestic violence following losses, but still a significant rise after a victory)
I believe you are confusing the correlation. It's not between playing football and domestic violence, it's between drinking alcohol and domestic violence!
I love listening to scientists talk unendingly around the question of the origin of "everything" without ever actually answering it , while at the same time trying to convince us and themselves that they actually HAVE an answer...I'll bet God thinks it's funny , too..
yay, I have been waiting, and always do love these... I listen to them over and over. Sean is amazing and helps me disbelieve in science that overlooks fundamentals.
The best way to fall asleep naturally is to have a bedroom devoid of distractions. No TV, no computer, no smartphone, no books, no nothing. The simpler the room, the better! It may take a while train you’re brain to do this, but the end results are fantastic! You’ll wonder why you didn’t think of this sooner!
@@dimitrispapadimitriou5622 Well, that’s the point, isn’t it, to keep those distractions to a minimum. If you’ve got a TV and a computer in your bedroom they’re always distracting your brain from naturally entering your sleep cycle.
@@terrypussypower Sometimes when I hear a familiar voice in the darkness speaking about philosophy or physics ( especially about topics that are already very familiar, so they 're not triggering much thinking or my curiosity ) I tend to sleep easier... Perhaps listening something like that with eyes closed, prevents less pleasant thoughts about events or problems of the day to disturb mental tranquility. But I agree that in other cases, silence and darkness are preferable.
