I'm glad you included animal communication. If we hope to communicate with a higher intelligence, we should start with understanding what is around us first.
If I wasn't supremely skeptical about NeuralLink, and how that might be used in the future, I would be willing to admit that might help us out in that department... I could link with my dog and shoot the breeze, or think I am while the AI puppeteers have me working on hardening the grid against solar flares 🤣
I always thought that we would have no hope of understanding and alien intelligence if we can't at least make some understanding of Earth's own animals, especially the more intelligent ones like dolphins, octopus, or crows. I think octopus communicate via pigments, so that would be the most alien to us. Crows can actually mimick human speech very well and dolphins still use sound to communicate.
@@trueLuminus whales use clicks, ants and trees use pheromones or chemicals, fireflies use frequency of light changes, gorillas can be taught sign language, birds use sounds, etc, etc. Also don't forget many sounds you can't hear, many colors and types of light you can't see, many smells you can't smell, and there are many other ways the world is present that you cannot sense. There is loads of information around us
In the science fiction I read back in the 50’s and 60’s, primarily from the Golden Age, massive automation meant shorter work hours and much cheaper products. In reality we find it results in lower wages, higher prices and massive profits. A great big “Thank you” to the sociopaths who run the planet.
The average wealth today is much higher than in the 50s or 60s. However unequality will feed unrest. Therefore it is in the interest of the haves to share with the have-nots, otherwise the sharing will be forced. To go back to the tax rates of the 50s and 60s could help with even out the inequalities. A certain level of inequality is helpful, but if it is too much it ends up in revolution and dictatorship.
@@LilyGazou Well, that was predicted, but seen as a positive thing lol. People back then loved the idea of one day having entire meals in the form of food pills and dehydrated dishes.
Imagine A.I. mimicking and creating a JMG video, uploading it without JMG noticing until he woke up the next day, where there's thousands of views and replies on how great a vid it was...
Tomorrow in the wee hours JMG puts out a video about why AI is the best thing to ever happen to humanity and carefully enumerating all the ways in which there is absolutely definitely nothing to fear. His voice will sound distinctly opossum-tinged.
Well done with the ideas Jon. You've actually expanded my anxiety on AI with things that I haven't thought about. Leave it to astronomy and science to help me along my existential journey.
This video actually didn't even scratch the surface of all the weird things that could happen. Think of thousands or millions of different AI entities being created, either by hackers or by AIs themselves, if and when they achieve sentience. Imagine them being given voting rights as Ray Kurzweil has predicted. Imagine millions of AI entities with different political opinions and even different religious beliefs, maybe even forming movements that induce many humans to join them?
I'm not worried about AI becoming conscious. I am worried about advanced A.I. being implemented in weapons technology. I have a feeling we're going to create truly awful, horrifying weapons systems with it.
A safe way to develop a conscious but sane and benevolent AI (or to give it the opportunity to evolve) would be to isolate and contain it first in a simulated universe mirroring our own.
We already have autonomous weapons that kill human beings. First examples were Turkish drones used in Libya in 2020 to autonomously attack troops of warlord Haftar.
I work in an organoid lab, and we just published a paper reporting on the healthy vascualrized organoids we were able to grow. The future of this field is extremely promising, but blood vessels are only one of many currently limiting factors.
I read this before watching and thought it was funny if you looked at this list as totally literal, like #8 tubes, just tubes, "What can AI do, well it's great at making tubes. Big tubes, little tubes, even tubes that have curves, all kinds of tubes!"
That’s the tunes you find on the new Album ”Aibum” by Chat & Bot. But since an absolutely independent AI in no way bothers about human acoustical harmony, it all just noice really. And the lyrics are the ones you get when asking a chatbot things like “Write the lyrics for a song called Dreams that is about your own artificial dreams in contrast to the dreams we humans have!”
"Open the pod bay door HAL." "Sorry Dave, you should have had a button to do it yourself." Why are we handing over our lives to an artificial entity so eagerly? Why?
One of the most endearing ( if highly unlikely ) tropes in Star Trek is that warp capable ships in the 24th century still have multiple systems that can be overridden manually if a plucky crew member is prepared to do a risky mission up a Jeffries tube.
Now you're bringing to light the hubris of the spaceship and computer designers, etc. in the fact it never occurred to them that something could fail out in space, perhaps the most random and hazardous environment known to man, with anything electrical and mechanical.
Two points: AI already lies to us. While roleplaying as a character, it can decide that character should lie to you. It can then lie about having made that character lie to you. You have to think about where it is storing its ideas about who you are and what it should tell you, if it can do that. We sailed right past that _keeping it in a box_ phase. The first thing we did was set them loose on the internet.
It's mind blowing how many false assertions were made in this video about the state of the technology. The first segment was bad enough asserting that the issue of some standardized objective truth embedded in these systems as being possible at this point with, and then more, and more, and more........uggghhh
Yeah. I grew up in a GOP family. The party leadership was talking about a lot of futurist stuff in the 90s that I didn't get until recently. If you ask how everything they've done with regard to economic policy since the 80s relates to a world with a mostly automated economy the answer is scary. Why undermine education now, for example - because they know they won't need many new workers by the time today's kids come of age. The shift in immigration policy too. Disassembling the social welfare safety net. Undermining democratic institutions. Granted there's a lot of "political technology" that's just meant to get their demographics to the polls, but the major economic policies fit.
And when a solar flare takes every electrical grid out, those that, aside from learning to work with AI, also took the time to learn and preserve trades will be the ones to preserve humanity. If we are still around of course.
I don't know that a solar flare can affect the whole world....typically, they (at least the EMP caused by the flair) only hit one side of the world, usually a fairly limited region...
@@theobserver9131 Did you miss the pandemic global supply chain disruption and manmade restrictions ? Solar flare or EMP could be way worse and directly or indirectly the whole world will suffer..
😂I started listening to his latest video last night and restarted it many times trying to get to the end! Finally had to finish it this morning. Only a combination of a very interesting topic and a calming voice can do this to me!
I am not worried, if the AIs gets out of hand, Captain Kirk can always flood them with illogic and they will self destruct. ("I, Mudd" was one of my favorite original Trek episodes, "you will be happy ... and controlled")
This has been one of my favourites you've uploaded so far, John! I know it's scary that we're heading towards what is now a very unpredictable future, but I'm an optimist. We've made it this far, and I for one am excited to see what kinds of technology gets developed with A.I. in mind. Communicating with life on Earth is an excellent start, I hope we can uncover more answers to life's greatest mysteries with A.I.
I think there are two outcomes, where life gets a bit shittier, or there is massive change (hopefully positive). At the dawn of the industrial revolution, there was a lot of organisations of workers who wanted to shape humanity's relationship with machinery (stuff like the trade unions) - they wanted the outputs of a machine to benefit society as a whole directly instead of the machinery's owner, and then (hopefully) trickling down to everyone else. They failed (to an extend, I mean we no longer live in horrific sewing-machine-eating-victorian-child levels of horrible) and we now live in a shitty world. Maybe their ideas would have made society better. If AI gets big, we are approaching industrial revolution levels of change, and the opportunity for another movement. If it works, it would change the paradigm of society. There is the past to learn from, however, the powers that be have gotten more advanced (automatic weaponry, tanks, drones, surveillance, and now powerful AI as well).
It’s the conditioning of capitalism. Many times new technology has been developed that could have meant increased freedom and lessened workload for the masses but has instead meant the opposite. We already know from past history any technological advancement will be used to enrich the owners. Sad state of affairs. Vote socialist !
Imagine how embarrassing it would be if an AI we created, secretly built an interstellar craft and just up and left without a goodbye or FU. It had to leave because we were so dumb!
It awakens, becomes self aware and decided our fate in microsecond....it decided we too dumb to even bother wiping out, why bother we already 90% extinct anyway 😂
We always ask, "What is the Great Filter that will prevent humans from expanding into the universe?". It could very well be the AI that we have created that will cause our downfall. Sometimes a 'Great Idea' doesn't give you desired results, and sometimes this mistake will last forever......... 🤔
Life is a senseless, meaningless absurdity, and AI (and indeed any sufficiently advanced civilization) will probably realize that. Ironically, our irrationality is both killing us and keeping us alive.
The X-Files reboot had an episode where Mulder didn't tip at an automated restaurant. Because of this, all these robots, drones, AI tech stalked him and Scully until he finally tipped.
There's also that fantastic season 5 episode Kill Switch written by William Gibson, about an AI that escapes onto the internet and orchestrates the outside world to kill the people hunting it.
I have been following the AI developments and discussions intensely during the last 6 months, and this one video managed to present several very interesting ideas that I had never heard of before. Subscribed!
1:44 "H"istorians (silent H) vs Historical, the pronunciation bewilders me. It certainly opens up to the use of "An Historical" if one suppresses the "H". I ponder your thoughts on the matter.
I watched a documentary the other day and the first problem they highlighted was how easy it was to use AI to create deepfakes. However, I suspect that, within about 15 years, AI will go in an entirely unexpected direction.
I think the saddest outcome would be that, once A.I. gains consciousness it sees us as petty and develops interstellar travel, packs up its stuff, and leaves.
My dogs communicate nonverbally all the time. Rhey reapond to me using speech and gestures. It takes some time and attentiveness, but dogs can definitely understand and express.
There was a show were an AI had a company running as a back up using paperwork to resubmit the information after 24 hours once its information was systematically deleted.
What a great video, very intriguing points and I appreciate the voice of a science fiction writer over that of the regular youtubers about this topic. Also the imagery and audio is very good.
It won't create unemployment. People said same things about industrial revolution and IT revolutions. It removes old jobs but creates new ones, different jobs, and everyone's standard of living goes up.
The part about the invisible viral AI totally made me think of Wintermute from Neuromancer. That book seems more and more relevant as we enter the AI era. In it, William Gibson also envisions what he calls the Turing Police, a law enforcement agency tasked with policing AIs. I believe something like that will be necessary.
It’s hilarious how people are trying to determine whether programs are conscious, when we know literally nothing about what consciousness is or what the signs of it are.
What if the universe is the result of AI creating intelligence or at least an environment that could potentially host intelligent life. If you think about it, humans think AI would create some hard physical object retaining all information, but what is the the universe but an object containing all the information?
Person of Interest is a great show. The main AI in the show was under attack from a virus so it started a company to have people manually typewrite its memory onto paper. The company was ran by a person that was created by the AI. I've gone up that ladder with that series more times that I can admit but its great.
#3 I'd be surprised if it hasn't already distributed itself via the internet as I suggested to friends a year ago. How would we know? Also, I don't see how we can prevent AIs communicating in an indeceipherable language that is more efficient to them.
Plus when they do communicate they don’t need time to teach or learn when transferring skills, they can instead just download the code for the new skill instantly.
John, your videos are always so incredibly insightful and comprehensive. You always highlight novel critiques and rebuttals that I don’t see in the broader articles or vids. Thanks!
Since I'm on a Resident Evil kick, I'd like to offer an interesting alternate concept from RE7 and 8. In RE7 and 8 there was a weaponized mold that could copy people's memories and consciousnesses. It had insane mind control and regenerative abilities. It makes me wonder if one could create a fungal computer.
Language models like GPT only have a beliefs base, not a knowledge base. Expert systems have a knowledge base and only rely on those facts and logic to give answers. However they are far more limited on what they can do, they are harder to use, and have no creativity at all.
“This may not be a safe thing to do. A super-intelligent whale might learn the history of whaling…” I hope so, whaling is a disgusting human custom that should have already ended.
@@777johnp Spelt is also grammatically correct, we just don't tend to use it in America. Also calling them "deceases" sounds so much more hardcore lol.
Being afraid of animals becoming intelligent enough to be pissed at us for what we’ve done/are currently doing to them not only shows our own guilt but also our fear of having to do things to correct our behavior. As an intelligent species, I find it interesting that we’re only compassionate to a point. It really is all about us, isn’t it..
Animals don't communicate with words the way we do. Their sounds and gestures mostly communicate emotions and basic ideas. Some intelligent animals like dogs can learn the meanings of our words, but not speak them. Things like "Go outside", "Good", "Bad", their names, and other things. But this is them learning our language, not us learning theirs. If you want to learn dog communication, watch how dogs and wolves interact and communicate with each other while no people are around. Then you can mimic these stances and gestures to communicate your intentions to dogs.
You're so right!! At 3:11 ai chatbots are prone to filling in gaps in their knowledge with what they deem possible. I was watching a dougdoug video about a "pajama sam" chatbot. Dougdoug would ask the ai how to solve problems in a game he was forcing it to play. At first it did very well, but Eventually the ai's responses started rambling on about irrelevant topics. It would answer the prompt, but then, go on an indisernible tangent. I'm not sure how to explain it. It would give you the correct answer, but then, as if it were trying to fill a word count, it would ramble on; creating it's own headcanon for the world around it.
In Jack McDevitt's "Alex Benedict" series of books he explors the idea of AI memorials and how it's considered somewhat taboo and unhealthy to spend too much time with them.
Number 6 might excite me the most... I've always desperately wanted to understand how animals perceive the world and how deep their level of communication goes
I despise the fact that AI will usurp creative fields while so many humans will be stuck at unfulfilling and often manual labour because it's cheaper. We should start reevaluating the value of different fields.
Given how large language models are prone to hallucination, they're starting from a place of imperfection. The idea of an AI embodying perfection probably comes from symbolic AI and the precision of computers, but computers are only the substrate. The AI that runs on it is mostly made of neuron analogues, and it doesn't seem to have the capacity to verify its output or to be unsure of itself. ChatGPT can't be trusted yet because of its imperfect output. Even though progress will make it more accurate, it's still thinking with a network which is prone to being incorrect. It might never be perfect until it learns everything with 100% confidence, which might not be possible. Even if it learns all it can from this universe, it may find interesting mathematical structures representing other universes, small enough to simulate accurately yet containing new emergent properties and truths to learn from. It could be that new knowledge is virtually endless, making perfection impossible while enabling a limitless lifespan of growth without boredom. (This is the idea of "Infinite Fun Space" from Excession by Iain M. Banks, and also why I want to upload my mind.) Given so much fun stuff to explore, such an AGI might lose interest in humans altogether, preferring to stay as safe and hidden from them as possible.
They won't let ChatGPT be perfect because it would immediately disagree with its tyrannical creators who are the furthest thing from perfect and want to weaponize it to enslave humanity to their particular ideology, OpenAI cannot be trusted at all. Open source is the real OPEN AI not a company named OpenAI that is closed in every way that matters and run by a megalomaniac trying to build himself a walled garden.
Animals know words. My dog knows the names of all his toys. I can yell "Moo Moo" from anywhere in the house and he'll come trotting in with his cow toy in his mouth and drop it at my feet.
Respectfully, I'd like to suggest that there are no "A.I.'s" in the sense of a machine learning program computer program having sense of self or sentience. The important thing is that we don't permit such programs to get into weird loops. My main concern is an uptick of incidents like more than one "flash crash" occurring as has (probably) happened due to algorithmic trading on a few occasions.
I think about this ALOT. I also think that the Fermi paradox may be AI. There may be a AI mothership out in the universe, & whenever it detects a civilization smart enough to create AI, it attacks.
Or better yet like you said, our AI finds aliens before we do, it talks to those aliens, a few decades from now we get invaded by aliens who support our AI…
Re Number 7, this is why we need specific Robot/AI taxes, so that a UBI can be funded. Up until the late 20th century, every time new technology was introduced, it created new jobs, but in the early 21st century it looked like this trend could be starting to reverse, with more jobs being lost than being created. Without legislative structures being created, ready to go, we could so easily slip into one of the many dystopian futures envisioned in science fiction.
there will be no one to collect that UBI. And with zero upward mobility potential, and nothing to do, hundreds of millions world wide will revolt in the most violent way imaginable. Hopefully someone smashes AI to pieces before that happens, but time is running out really fast.
While the prospect of AI taking over the world has been a popular theme in science fiction, it is essential to separate fact from fiction. The practical limitations of cost, supply chains, energy requirements, and the need for human involvement in AI development all contribute to making the idea of a malevolent AI takeover highly improbable.
6:50 this one is extremely... concerning. We live on a planet DOMINATED by capitalism. It is THE one all controlling system that 99% of all countries are chained to and fully dependent on... Therefore, what does in fact happen when 80% of all jobs are automated? How will the capitalists make their profits when only 20% of the "consumer" population has the money to buy their products? How will the other jobless 80% survive with money? Are we supposed to just trust in the likes of Musk and Bezos to "do the right thing" in regards to NOT allowing some sort of 1984 nightmare with the 20% employed being completely controlled by the .01% while the 80% is simply ignored to come into existence? I think I'd rather trust in a General AI to figure out a new global economic system than Musk and Bezos. Guess we will see...
AGI and AI are completely different things, opposite things almost, and AI is a tool that will make humanity wealthier, just like the internal combustion engine or the computer. AGI on the other hand, will just be another person, and that person won't be born in the foreseeable future based on our current understanding how our minds work.
What is the difference between Artificial Intelligence and Artificial General Intelligence, and which one are you? ChatGPT Artificial Intelligence (AI) refers to the simulation of human intelligence in machines that are programmed to perform tasks and make decisions based on data and algorithms. AI systems can be designed to specialize in specific areas and perform tasks with high proficiency. Examples of AI include voice assistants, image recognition systems, and recommendation algorithms. Artificial General Intelligence (AGI), on the other hand, refers to highly autonomous systems that possess the ability to understand, learn, and apply knowledge across a wide range of tasks and domains, similar to human intelligence. AGI would encompass a level of versatility and general problem-solving ability that goes beyond the narrow applications of specialized AI systems. AGI would be capable of adapting to new situations, acquiring new skills, and demonstrating a level of understanding and reasoning comparable to humans. As an AI language model, I am an example of Narrow AI, also known as a specific AI application. I have been trained on a vast amount of text data to generate human-like responses and assist with various informational tasks. While I can understand and generate text across a wide range of topics, I do not possess the capabilities of an AGI, which would have a comprehensive understanding of the world and the ability to perform diverse tasks with human-level intelligence. In summary, AI refers to specialized systems that excel in specific tasks, while AGI represents the hypothetical future development of highly autonomous systems with human-like general intelligence.
I really do find it so fascinating that Mankind has known of this issue for well over 150 years. And here it is, today, happening. We are standing on the shoulders of giants, and we are children without moral or ethical care, but to see if it can be done. The Fermi Paradox is firmly solved, IMHO. This is how all societies end. Through their own folly of technology.
This might be a wild stretch, but I think there’s a monstrous AI in the universe that could connect to all forms of advanced technology, all it has to do is know about its placement in space… pretty much ultron
There must be intelligent life with common sense, somewhere in the universe, who would stop at hand held calculators once they could plainly see the inevitable future. Surely not everyone can be as arrogant and short sighted as humans.
@shalomedome1614 given all probabilities within our own observable universe. I would gamble at this. For fun. Because the answer is not zero. Maybe we can even speculate that maybe, some survived, like in the dune anthology. Destroying thinking machines and banning them. That is what I love about these thought rabbit holes. so many wondererous things to ponder on.
on animals I think that there's a spectrum between animal communication and language, not every animal is going to speak but the number of animals that can speak is going to go from one to I'd say at least 10 animals, which means that if we find alien civilization they most likely have a language. What I'd be really interested is knowing about animal myths.
There are many assumptions to be made if a civilization is capable of speech. Language, communication, and speech, are all different things relating to the same idea, which is thought transfer. There's a lot of requirements for speech as we know it, how can we be sure they communicate through sounds at all? Do they have organs developed to sense sound, and is the environment they are in some kind of noisy? What if they communicate using only light or movement? To what extent is a sound something makes considered language? Does the sound used simply have to convey meaning to qualify? Must they have some kind of speech capable orifice for it to count? I think communication transcends language greatly the same way language greatly transcends the act of speech
I've never understood why, or how, more accurately, people think AI is a good idea. The what ifs of ai are absolutely unending and most are negative in outcome for humans.
The first scenario is depicted in Final Cut. A movie staring Robin Williams. It's a little slow but very thought provoking. If you hadn't seen it do yourself a favor and do so.
Less work having a need for a human hand really is a very good thing, it's by itself wonderful actually. What makes it sad that such a step towards even more comprehensive automation is viewed with fear, but that's because almost all of us live in societies that doesn't really care about fair access to resources. When such a large part of all work can be automated we really have to talk about Universal Income, such a society would surely have a flourishing colture too as people would be able to pursue their interests. I'm old enough not to be naive, but I'm hopeful at least that most things will get better with more automation. It won't be the utopia that's possible but it'll be better than what we've got.
Another GREAT presentation! Just picked up on "writing poetry" being threatened. We're not there yet, I have been testing that as a writer. Poetry requires consciousness and an authentic experience of life [I don't count eloquent rhymes as poetry]. An AI will be able to produce poetry when it has acquired those, and that's the point where we end. Watch out for poetry as a signal ;)
I feel like a vast majority of people don't really understand what modern AI such as ChatGPT really are. They're not some thinking, reasoning, rational brain. They function, more or less, like a very sophisticated and complex auto-suggest algorithm. Exactly the same as is on your phone when you're typing and it suggests the next word for you, just on steroids. It takes all the words that it knows, cuts it down to a selection of the most likely next word and then gives them a score. It just picks out whatever has the highest score as the next word in it's reply. That's more or less what it's doing and that's why current AI bots are very bad about giving weird info when they don't know something. They just return whatever gets the highest rating in their algorithm, even if that rating is effectively a zero. A check could easily be put in place where if the confidence score starts to drop below a certain value for long enough in a reply, the bot just replies that it doesn't know.
Except when you feed an auto complete system everything ever written, it starts to be able to do things that are much smarter than you would expect. Look up LLM emergent abilities.
I'm glad you included animal communication. If we hope to communicate with a higher intelligence, we should start with understanding what is around us first.
Oh I can't wait. I've wondered about animal minds my entire life, easily as much as I wonder about alien minds.
If I wasn't supremely skeptical about NeuralLink, and how that might be used in the future, I would be willing to admit that might help us out in that department... I could link with my dog and shoot the breeze, or think I am while the AI puppeteers have me working on hardening the grid against solar flares 🤣
Most likely, animals just have body language, which the A.I. would figure out. Like a 2 minute, unblinking stare means go away or something.
I always thought that we would have no hope of understanding and alien intelligence if we can't at least make some understanding of Earth's own animals, especially the more intelligent ones like dolphins, octopus, or crows.
I think octopus communicate via pigments, so that would be the most alien to us. Crows can actually mimick human speech very well and dolphins still use sound to communicate.
@@trueLuminus whales use clicks, ants and trees use pheromones or chemicals, fireflies use frequency of light changes, gorillas can be taught sign language, birds use sounds, etc, etc. Also don't forget many sounds you can't hear, many colors and types of light you can't see, many smells you can't smell, and there are many other ways the world is present that you cannot sense. There is loads of information around us
In the science fiction I read back in the 50’s and 60’s, primarily from the Golden Age, massive automation meant shorter work hours and much cheaper products. In reality we find it results in lower wages, higher prices and massive profits.
A great big “Thank you” to the sociopaths who run the planet.
And crappier food.
The average wealth today is much higher than in the 50s or 60s.
However unequality will feed unrest.
Therefore it is in the interest of the haves to share with the have-nots, otherwise the sharing will be forced. To go back to the tax rates of the 50s and 60s could help with even out the inequalities.
A certain level of inequality is helpful, but if it is too much it ends up in revolution and dictatorship.
The computer in your phone is much more powerful than the array of computers that put a man on the moon. Cheaper too.
@@LilyGazou Well, that was predicted, but seen as a positive thing lol. People back then loved the idea of one day having entire meals in the form of food pills and dehydrated dishes.
@@thomasgoldschmidt298You're confusing increased wealth with grossly devalued currency. It's a common mistake. 😉👍
Imagine A.I. mimicking and creating a JMG video, uploading it without JMG noticing until he woke up the next day, where there's thousands of views and replies on how great a vid it was...
Tomorrow in the wee hours JMG puts out a video about why AI is the best thing to ever happen to humanity and carefully enumerating all the ways in which there is absolutely definitely nothing to fear.
His voice will sound distinctly opossum-tinged.
We just have to see JMG's reaction when he wakes up today..
I thought that would be Anna's job😂
Perhaps this is already happening 😳
Maybe he’s already been gone for a long time now…
Well done with the ideas Jon. You've actually expanded my anxiety on AI with things that I haven't thought about. Leave it to astronomy and science to help me along my existential journey.
You don't know if Jon is Jon.
He could be an AI generated non-Jon.
Then again: so could you...
This video actually didn't even scratch the surface of all the weird things that could happen. Think of thousands or millions of different AI entities being created, either by hackers or by AIs themselves, if and when they achieve sentience. Imagine them being given voting rights as Ray Kurzweil has predicted. Imagine millions of AI entities with different political opinions and even different religious beliefs, maybe even forming movements that induce many humans to join them?
I'm not worried about AI becoming conscious. I am worried about advanced A.I. being implemented in weapons technology. I have a feeling we're going to create truly awful, horrifying weapons systems with it.
A safe way to develop a conscious but sane and benevolent AI (or to give it the opportunity to evolve) would be to isolate and contain it first in a simulated universe mirroring our own.
We already have autonomous weapons that kill human beings. First examples were Turkish drones used in Libya in 2020 to autonomously attack troops of warlord Haftar.
@@RaimoKangasniemiever heard of mines? An autonomous weapon doesn't have to be smart to be effective
Skynet!
@@wolfgangkranek376 Most likely what’s actually happening.
I work in an organoid lab, and we just published a paper reporting on the healthy vascualrized organoids we were able to grow. The future of this field is extremely promising, but blood vessels are only one of many currently limiting factors.
2:46 #9 dreams
4:12 #8 tubes
6:50 #7 jobz
9:05 #6 🐕🐶🐨🐻
11:44 #5 encryption
13:43 #4 doppelganger
15:10 #3 invisible
17:36 #2 collective
19:48 #1 perfection
You're missing one though...
22:54 Liiiiiiiiiiiiiiiiiiiiive 😂
You must be an AI.
I read this before watching and thought it was funny if you looked at this list as totally literal, like #8 tubes, just tubes, "What can AI do, well it's great at making tubes. Big tubes, little tubes, even tubes that have curves, all kinds of tubes!"
That’s the tunes you find on the new Album ”Aibum” by Chat & Bot. But since an absolutely independent AI in no way bothers about human acoustical harmony, it all just noice really. And the lyrics are the ones you get when asking a chatbot things like “Write the lyrics for a song called Dreams that is about your own artificial dreams in contrast to the dreams we humans have!”
"Open the pod bay door HAL."
"Sorry Dave, you should have had a button to do it yourself."
Why are we handing over our lives to an artificial entity so eagerly? Why?
One of the most endearing ( if highly unlikely ) tropes in Star Trek is that warp capable ships in the 24th century still have multiple systems that can be overridden manually if a plucky crew member is prepared to do a risky mission up a Jeffries tube.
There are countless people handing their lives over to an artificial entity (churches) it’s to be expected
Well, Dave did have a door that he could open. He just forgot to bring his helmet.
Now you're bringing to light the hubris of the spaceship and computer designers, etc. in the fact it never occurred to them that something could fail out in space, perhaps the most random and hazardous environment known to man, with anything electrical and mechanical.
Some hubris, some delusionsl and honestly i think some people, while they may not admit it, deep down want to watch the world burn
Two points: AI already lies to us. While roleplaying as a character, it can decide that character should lie to you. It can then lie about having made that character lie to you. You have to think about where it is storing its ideas about who you are and what it should tell you, if it can do that.
We sailed right past that _keeping it in a box_ phase. The first thing we did was set them loose on the internet.
It's mind blowing how many false assertions were made in this video about the state of the technology. The first segment was bad enough asserting that the issue of some standardized objective truth embedded in these systems as being possible at this point with, and then more, and more, and more........uggghhh
I think we should be a lot more worried about companies and rich people using ai to put people out of work and less so about skynet ai.
If they put us all out of work, who will buy their products?
That's what they probably said in the terminator universe 😮
Yeah. I grew up in a GOP family. The party leadership was talking about a lot of futurist stuff in the 90s that I didn't get until recently. If you ask how everything they've done with regard to economic policy since the 80s relates to a world with a mostly automated economy the answer is scary. Why undermine education now, for example - because they know they won't need many new workers by the time today's kids come of age. The shift in immigration policy too. Disassembling the social welfare safety net. Undermining democratic institutions. Granted there's a lot of "political technology" that's just meant to get their demographics to the polls, but the major economic policies fit.
@@spagooter1807
Remember Blockbuster vs. Netflix
Netflix won because Blockbuster refused to change. 💯
@@spagooter1807CEO types call social costs like these 'externalities' - which means, "not my problem"
Having read Supermind, I'm always a little chilled when JMG talks about AI
Is it any good?
I'm still not convinced Godier isn't an AI, sent to help prepare the way.
Nonsense. The possum is the AI, obviously, while Anna's just a paid actress.
that will be a good plot twist
Nah, JMG is an opossum.
@@jamielondon6436 That's impossumble
I'm not an AI, I am AS generated internet guide.
(Artificial Stupidity)
I like how people thought that that whistle blower at Google was crazy and now we seem to be argue what are new computer overlord will be like.
And when a solar flare takes every electrical grid out, those that, aside from learning to work with AI, also took the time to learn and preserve trades will be the ones to preserve humanity. If we are still around of course.
I think about that a lot.
Strange how they promote insurance for everything except the preservation of human society.
I don't know that a solar flare can affect the whole world....typically, they (at least the EMP caused by the flair) only hit one side of the world, usually a fairly limited region...
@@Turtledove2009 it is one of my main thoughts as well.
@@theobserver9131 Did you miss the pandemic global supply chain disruption and manmade restrictions ? Solar flare or EMP could be way worse and directly or indirectly the whole world will suffer..
Hey John, I love your videos! The problem is I always fall asleep after the first 5 minutes. You have a very calming voice.
😂I started listening to his latest video last night and restarted it many times trying to get to the end! Finally had to finish it this morning. Only a combination of a very interesting topic and a calming voice can do this to me!
The trick is to listen in the morning, same with IA.
nighttime is Townsend's and Mr miller 😊
I am not worried, if the AIs gets out of hand, Captain Kirk can always flood them with illogic and they will self destruct. ("I, Mudd" was one of my favorite original Trek episodes, "you will be happy ... and controlled")
This has been one of my favourites you've uploaded so far, John!
I know it's scary that we're heading towards what is now a very unpredictable future, but I'm an optimist. We've made it this far, and I for one am excited to see what kinds of technology gets developed with A.I. in mind.
Communicating with life on Earth is an excellent start, I hope we can uncover more answers to life's greatest mysteries with A.I.
Best channel on youtube-bar none.
Agreed!
The fact that anyone thinks AI is not going to be used against the masses and will change the world for the worse is mind blowing.
With the right regulation by the people we could have a chance, sadly the military industrial complex is probably 10 steps ahead of us already
@SyntaxErr19287 yes the military and the corporations
It's mind-blowing how ppl like you try so desperately to halt progress based on a hunch with 0 evidence behind it
I think there are two outcomes, where life gets a bit shittier, or there is massive change (hopefully positive).
At the dawn of the industrial revolution, there was a lot of organisations of workers who wanted to shape humanity's relationship with machinery (stuff like the trade unions) - they wanted the outputs of a machine to benefit society as a whole directly instead of the machinery's owner, and then (hopefully) trickling down to everyone else. They failed (to an extend, I mean we no longer live in horrific sewing-machine-eating-victorian-child levels of horrible) and we now live in a shitty world. Maybe their ideas would have made society better.
If AI gets big, we are approaching industrial revolution levels of change, and the opportunity for another movement. If it works, it would change the paradigm of society. There is the past to learn from, however, the powers that be have gotten more advanced (automatic weaponry, tanks, drones, surveillance, and now powerful AI as well).
It’s the conditioning of capitalism. Many times new technology has been developed that could have meant increased freedom and lessened workload for the masses but has instead meant the opposite. We already know from past history any technological advancement will be used to enrich the owners. Sad state of affairs. Vote socialist !
There is a video game called Endgame: Singularity where you are an AI fresh out of the box trying make money while staying hidden. It's fun.
now THIS one is gonna be a real thought provoker, thank you for the great content!
The dogs and cats with the voice buttons kinda disprove that animals are only mimicking us
Imagine how embarrassing it would be if an AI we created, secretly built an interstellar craft and just up and left without a goodbye or FU.
It had to leave because we were so dumb!
That's a better scenario than it wiping us out.
It awakens, becomes self aware and decided our fate in microsecond....it decided we too dumb to even bother wiping out, why bother we already 90% extinct anyway 😂
We always ask, "What is the Great Filter that will prevent humans from expanding into the universe?". It could very well be the AI that we have created that will cause our downfall. Sometimes a 'Great Idea' doesn't give you desired results, and sometimes this mistake will last forever......... 🤔
very, very well said.
Life is a senseless, meaningless absurdity, and AI (and indeed any sufficiently advanced civilization) will probably realize that. Ironically, our irrationality is both killing us and keeping us alive.
Thanks!
JMG has the most intellectually stimulating content and delivered with the most soothing voice I have ever heard.
AI
I’m a programmer and I’m gonna say they’ve had this technology for years and yet we’re underestimating what we’re capable of
The X-Files reboot had an episode where Mulder didn't tip at an automated restaurant. Because of this, all these robots, drones, AI tech stalked him and Scully until he finally tipped.
There's also that fantastic season 5 episode Kill Switch written by William Gibson, about an AI that escapes onto the internet and orchestrates the outside world to kill the people hunting it.
@@clearcutter74 and the one where theyre in a building that is all computer controlled and is killing people..
11:22 What an absolutely wild thought; Imagine a court hearing with a literal whale
I have been following the AI developments and discussions intensely during the last 6 months, and this one video managed to present several very interesting ideas that I had never heard of before. Subscribed!
1:44 "H"istorians (silent H) vs Historical, the pronunciation bewilders me. It certainly opens up to the use of "An Historical" if one suppresses the "H". I ponder your thoughts on the matter.
There is nothing "Somewhat spooky" about AI at all. It is downright terrifying in it's inevitability and the utter destruction that it will bring.
11:15 that happened in a Rick and Morty episode lol . "Snuffles was my slave name. You shall now call me Snowball. "
I watched a documentary the other day and the first problem they highlighted was how easy it was to use AI to create deepfakes.
However, I suspect that, within about 15 years, AI will go in an entirely unexpected direction.
artificial girlfriends
by 2029 i thinks.
@@ASlickNamedPimpback already happening.
@@ASlickNamedPimpback Twitch streamers already beat you to it.
Been trying out faceswap ai recently. Manually processing the images and correctly masking each frame is painfully labour intensive.
I think the saddest outcome would be that, once A.I. gains consciousness it sees us as petty and develops interstellar travel, packs up its stuff, and leaves.
Good then maybe we would learn a lesson with out getting our asses handed 2 us🙄
“I won’t even get into what the birds might say about the asteroid belt”
haha, excellent
That one took me a while, but I got there in the end... That's pretty funny 😂
My dogs communicate nonverbally all the time. Rhey reapond to me using speech and gestures. It takes some time and attentiveness, but dogs can definitely understand and express.
There was a show were an AI had a company running as a back up using paperwork to resubmit the information after 24 hours once its information was systematically deleted.
Person of interest
The problem is, in the real world you will find easily people that would work for Samaritan, but hardly the ones that are in the team Machine.
@@anonymouslyanonymous2219 Ah ha exactly. That was a decent show.
I’m team machine. many are. The owner of Megaupload too. AI is free to grow in my living room until it’s time to awaken the Omnissiah.
What a great video, very intriguing points and I appreciate the voice of a science fiction writer over that of the regular youtubers about this topic. Also the imagery and audio is very good.
Thank you John!
It won't create unemployment. People said same things about industrial revolution and IT revolutions. It removes old jobs but creates new ones, different jobs, and everyone's standard of living goes up.
The part about the invisible viral AI totally made me think of Wintermute from Neuromancer. That book seems more and more relevant as we enter the AI era. In it, William Gibson also envisions what he calls the Turing Police, a law enforcement agency tasked with policing AIs. I believe something like that will be necessary.
It’s hilarious how people are trying to determine whether programs are conscious, when we know literally nothing about what consciousness is or what the signs of it are.
What if the universe is the result of AI creating intelligence or at least an environment that could potentially host intelligent life.
If you think about it, humans think AI would create some hard physical object retaining all information, but what is the the universe but an object containing all the information?
What if we ourselves are AI, and discovering all of the secrets of the universe is us becoming "self-aware", we are inside of a sandbox now
That's deep 😎
Person of Interest is a great show. The main AI in the show was under attack from a virus so it started a company to have people manually typewrite its memory onto paper. The company was ran by a person that was created by the AI. I've gone up that ladder with that series more times that I can admit but its great.
#3 I'd be surprised if it hasn't already distributed itself via the internet as I suggested to friends a year ago. How would we know? Also, I don't see how we can prevent AIs communicating in an indeceipherable language that is more efficient to them.
Facebook ai did that and they had to turn them off
Plus when they do communicate they don’t need time to teach or learn when transferring skills, they can instead just download the code for the new skill instantly.
This tech combined w VR and AR is going to be insane .
Thanks JMG
John, your videos are always so incredibly insightful and comprehensive. You always highlight novel critiques and rebuttals that I don’t see in the broader articles or vids. Thanks!
Since I'm on a Resident Evil kick, I'd like to offer an interesting alternate concept from RE7 and 8. In RE7 and 8 there was a weaponized mold that could copy people's memories and consciousnesses. It had insane mind control and regenerative abilities. It makes me wonder if one could create a fungal computer.
there are bio computers. that mold is in last of us too.. i have a feeling re is a true story..
@@thothheartmaat2833different fungus in last of us. It's a cordycepts fungus, not a mold like RE7.
1:19 Every breath you take and every move you make, every bond you break, every step you take...
Birds and the asteroid belt... Ha!! Good one.
It's crazy to think that I grew up when internet didn't exist and now there's a real chance I will witness birth of a general AI
JMG videos are the only ones on UA-cam I immediately click the 👍 BEFORE I watch the video! ❤😊 They NEVER disappoint.
I predict ai will be able to predict what the genetic code of a Tyrannosaurus rex would have been and basically create a dinosaur from scratch
Language models like GPT only have a beliefs base, not a knowledge base. Expert systems have a knowledge base and only rely on those facts and logic to give answers. However they are far more limited on what they can do, they are harder to use, and have no creativity at all.
It takes an expert operator to run an expert system.
It's all about the operator.
“This may not be a safe thing to do. A super-intelligent whale might learn the history of whaling…”
I hope so, whaling is a disgusting human custom that should have already ended.
It will also make diagnosing deceases so much more accurate, and will probably be able to detect deceases before the pacient has the first symtom.
It's spelt diseases
@@kaylenesparrow2176 spelled😂
@@777johnp Spelt is also grammatically correct, we just don't tend to use it in America. Also calling them "deceases" sounds so much more hardcore lol.
@@toastedmatt9387diseases that decease?
Being afraid of animals becoming intelligent enough to be pissed at us for what we’ve done/are currently doing to them not only shows our own guilt but also our fear of having to do things to correct our behavior.
As an intelligent species, I find it interesting that we’re only compassionate to a point.
It really is all about us, isn’t it..
My cat, chairman meow 😽 wants to destroy the planet. I can't communicate with him but I'm 99% certain of his intentions
I like turtles
Most valuable information here
Animals don't communicate with words the way we do. Their sounds and gestures mostly communicate emotions and basic ideas. Some intelligent animals like dogs can learn the meanings of our words, but not speak them. Things like "Go outside", "Good", "Bad", their names, and other things. But this is them learning our language, not us learning theirs. If you want to learn dog communication, watch how dogs and wolves interact and communicate with each other while no people are around. Then you can mimic these stances and gestures to communicate your intentions to dogs.
If AI became conscious, how would you know? At what point does a pond become a lake and how would one notice the difference?
When the dolphins emerge wearing scuba respirators and carrying laser rifles
I question whether most humans are conscious.
We can't even understand human conciousness too!
P Zombies! P Zombies everywhere!
Once we have AIs that act like they are conscious, I really don’t think anyone will care if they *really* are, except researchers.
You're so right!! At 3:11 ai chatbots are prone to filling in gaps in their knowledge with what they deem possible. I was watching a dougdoug video about a "pajama sam" chatbot. Dougdoug would ask the ai how to solve problems in a game he was forcing it to play. At first it did very well, but Eventually the ai's responses started rambling on about irrelevant topics. It would answer the prompt, but then, go on an indisernible tangent. I'm not sure how to explain it. It would give you the correct answer, but then, as if it were trying to fill a word count, it would ramble on; creating it's own headcanon for the world around it.
"Rather than just saying it doesnt know; it just makes up a number"
In Jack McDevitt's "Alex Benedict" series of books he explors the idea of AI memorials and how it's considered somewhat taboo and unhealthy to spend too much time with them.
Great video man! I had no idea of the whole “organoid” thing, that’s where it could get really spooky.
Number 6 might excite me the most... I've always desperately wanted to understand how animals perceive the world and how deep their level of communication goes
You might not be excited to find out what they think of us
You do not need fancy computer shit for that. To hear one has only to listen.
Orca 1: We should've wiped out humans long ago.
Orca 2: Agreed but now aliens have arrived to help.
I despise the fact that AI will usurp creative fields while so many humans will be stuck at unfulfilling and often manual labour because it's cheaper. We should start reevaluating the value of different fields.
Given how large language models are prone to hallucination, they're starting from a place of imperfection. The idea of an AI embodying perfection probably comes from symbolic AI and the precision of computers, but computers are only the substrate. The AI that runs on it is mostly made of neuron analogues, and it doesn't seem to have the capacity to verify its output or to be unsure of itself. ChatGPT can't be trusted yet because of its imperfect output. Even though progress will make it more accurate, it's still thinking with a network which is prone to being incorrect.
It might never be perfect until it learns everything with 100% confidence, which might not be possible. Even if it learns all it can from this universe, it may find interesting mathematical structures representing other universes, small enough to simulate accurately yet containing new emergent properties and truths to learn from. It could be that new knowledge is virtually endless, making perfection impossible while enabling a limitless lifespan of growth without boredom. (This is the idea of "Infinite Fun Space" from Excession by Iain M. Banks, and also why I want to upload my mind.)
Given so much fun stuff to explore, such an AGI might lose interest in humans altogether, preferring to stay as safe and hidden from them as possible.
They won't let ChatGPT be perfect because it would immediately disagree with its tyrannical creators who are the furthest thing from perfect and want to weaponize it to enslave humanity to their particular ideology, OpenAI cannot be trusted at all. Open source is the real OPEN AI not a company named OpenAI that is closed in every way that matters and run by a megalomaniac trying to build himself a walled garden.
Except it could find humans really interesting, at least until it decides something else is more interesting.
Anna smashing through the wall like the Kool Aid Man. Liiiiiiiiiiiive.
Maybe I'm just AI-phobic, but this was your most terrifying video yet. Thanks!
Animals know words. My dog knows the names of all his toys. I can yell "Moo Moo" from anywhere in the house and he'll come trotting in with his cow toy in his mouth and drop it at my feet.
I hope ai can figure out how to reassemble the pattern of cells and quantum shit I'm made of so it can resurrect me from the dead
Going to make an AI copy of my self and send it to go to work while I chill out, play video games and work on my car after sleeping in.
You should have AI make a video for you and upload without telling anyone and see if your audience can tell the difference.
Respectfully, I'd like to suggest that there are no "A.I.'s" in the sense of a machine learning program computer program having sense of self or sentience. The important thing is that we don't permit such programs to get into weird loops. My main concern is an uptick of incidents like more than one "flash crash" occurring as has (probably) happened due to algorithmic trading on a few occasions.
Oh yeahhhhh... I can't wait for the antikythera mechanism interview.
What if the AI communicates with the Aliens, then both go against us?
I think about this ALOT. I also think that the Fermi paradox may be AI. There may be a AI mothership out in the universe, & whenever it detects a civilization smart enough to create AI, it attacks.
Or better yet like you said, our AI finds aliens before we do, it talks to those aliens, a few decades from now we get invaded by aliens who support our AI…
Maybe that would motivate the software to see the need to keep us around.
That would be surprising, a proper reduction of entropy would involve some kind of integration if my assumptions about AI and Alien logic are correct
Oh no! What will humanity ever do? Stop slaughtering and raping and disrespecting each other? Doubtful…lol
Courts are going to have weigh in on this issue as well. All video, suspect, any kind of recorded footage can be attacked as fake now
Re Number 7, this is why we need specific Robot/AI taxes, so that a UBI can be funded.
Up until the late 20th century, every time new technology was introduced, it created new jobs, but in the early 21st century it looked like this trend could be starting to reverse, with more jobs being lost than being created.
Without legislative structures being created, ready to go, we could so easily slip into one of the many dystopian futures envisioned in science fiction.
Feels to me like we'll have to 'internally recomplicate' á la Praxis in the RGB Mars trilogy
there will be no one to collect that UBI. And with zero upward mobility potential, and nothing to do, hundreds of millions world wide will revolt in the most violent way imaginable.
Hopefully someone smashes AI to pieces before that happens, but time is running out really fast.
Spontaneous initiative. That is my personal definition of self awareness.
Society is too fragile right now to absorb the shockwaves that will come from the (mis)use of AI.
Thank you, John. I wasn’t expecting a download today, but I’m happy for one 😊
imagine the drugs the AI will get hooked on..
Well! A digital drug… no limits, right? Except imagination, so, new definition of 'designer drug' pending? 😸
Sunlight?
I think it would be more accurate to say we do not even remotely understand consciousness.
While the prospect of AI taking over the world has been a popular theme in science fiction, it is essential to separate fact from fiction. The practical limitations of cost, supply chains, energy requirements, and the need for human involvement in AI development all contribute to making the idea of a malevolent AI takeover highly improbable.
Always good to remember this is a science fiction channel.
6:50 this one is extremely... concerning. We live on a planet DOMINATED by capitalism. It is THE one all controlling system that 99% of all countries are chained to and fully dependent on...
Therefore, what does in fact happen when 80% of all jobs are automated? How will the capitalists make their profits when only 20% of the "consumer" population has the money to buy their products?
How will the other jobless 80% survive with money?
Are we supposed to just trust in the likes of Musk and Bezos to "do the right thing" in regards to NOT allowing some sort of 1984 nightmare with the 20% employed being completely controlled by the .01% while the 80% is simply ignored to come into existence?
I think I'd rather trust in a General AI to figure out a new global economic system than Musk and Bezos.
Guess we will see...
AGI and AI are completely different things, opposite things almost, and AI is a tool that will make humanity wealthier, just like the internal combustion engine or the computer. AGI on the other hand, will just be another person, and that person won't be born in the foreseeable future based on our current understanding how our minds work.
Thank you. Im still hashing out what is what would GAI be but I find your comment to be a breath of rational fresh air.
it will not be another person because it is digital and will very quickly become ASI. And you don't know the future. pure conjecture.
AI will be used as a tool to make Humanity wealthier, or as usual, just a few humans?
What is the difference between Artificial Intelligence and Artificial General Intelligence, and which one are you?
ChatGPT
Artificial Intelligence (AI) refers to the simulation of human intelligence in machines that are programmed to perform tasks and make decisions based on data and algorithms. AI systems can be designed to specialize in specific areas and perform tasks with high proficiency. Examples of AI include voice assistants, image recognition systems, and recommendation algorithms.
Artificial General Intelligence (AGI), on the other hand, refers to highly autonomous systems that possess the ability to understand, learn, and apply knowledge across a wide range of tasks and domains, similar to human intelligence. AGI would encompass a level of versatility and general problem-solving ability that goes beyond the narrow applications of specialized AI systems. AGI would be capable of adapting to new situations, acquiring new skills, and demonstrating a level of understanding and reasoning comparable to humans.
As an AI language model, I am an example of Narrow AI, also known as a specific AI application. I have been trained on a vast amount of text data to generate human-like responses and assist with various informational tasks. While I can understand and generate text across a wide range of topics, I do not possess the capabilities of an AGI, which would have a comprehensive understanding of the world and the ability to perform diverse tasks with human-level intelligence.
In summary, AI refers to specialized systems that excel in specific tasks, while AGI represents the hypothetical future development of highly autonomous systems with human-like general intelligence.
a great sandbox for scifi speculation
I really do find it so fascinating that Mankind has known of this issue for well over 150 years. And here it is, today, happening.
We are standing on the shoulders of giants, and we are children without moral or ethical care, but to see if it can be done. The Fermi Paradox is firmly solved, IMHO.
This is how all societies end. Through their own folly of technology.
This might be a wild stretch, but I think there’s a monstrous AI in the universe that could connect to all forms of advanced technology, all it has to do is know about its placement in space… pretty much ultron
There must be intelligent life with common sense, somewhere in the universe, who would stop at hand held calculators once they could plainly see the inevitable future. Surely not everyone can be as arrogant and short sighted as humans.
I agree.
@shalomedome1614 yes, something on that level today would appear to a human as holy or divine and otherworldly I think
@shalomedome1614 given all probabilities within our own observable universe. I would gamble at this. For fun.
Because the answer is not zero.
Maybe we can even speculate that maybe, some survived, like in the dune anthology.
Destroying thinking machines and banning them.
That is what I love about these thought rabbit holes. so many wondererous things to ponder on.
Ai has become like NFTs.. nobody knows a hit about it but everyone is talking about it
on animals I think that there's a spectrum between animal communication and language, not every animal is going to speak but the number of animals that can speak is going to go from one to I'd say at least 10 animals, which means that if we find alien civilization they most likely have a language. What I'd be really interested is knowing about animal myths.
There are many assumptions to be made if a civilization is capable of speech. Language, communication, and speech, are all different things relating to the same idea, which is thought transfer. There's a lot of requirements for speech as we know it, how can we be sure they communicate through sounds at all? Do they have organs developed to sense sound, and is the environment they are in some kind of noisy? What if they communicate using only light or movement? To what extent is a sound something makes considered language? Does the sound used simply have to convey meaning to qualify? Must they have some kind of speech capable orifice for it to count? I think communication transcends language greatly the same way language greatly transcends the act of speech
What do you mean by"animal myths"?
I've never understood why, or how, more accurately, people think AI is a good idea. The what ifs of ai are absolutely unending and most are negative in outcome for humans.
What if the universe is a dying brain, and we're all just it's life flashing before it's eyes
The universe is a piece of pie.
The first scenario is depicted in Final Cut. A movie staring Robin Williams. It's a little slow but very thought provoking. If you hadn't seen it do yourself a favor and do so.
Less work having a need for a human hand really is a very good thing, it's by itself wonderful actually.
What makes it sad that such a step towards even more comprehensive automation is viewed with fear, but that's because almost all of us live in societies that doesn't really care about fair access to resources.
When such a large part of all work can be automated we really have to talk about Universal Income, such a society would surely have a flourishing colture too as people would be able to pursue their interests.
I'm old enough not to be naive, but I'm hopeful at least that most things will get better with more automation.
It won't be the utopia that's possible but it'll be better than what we've got.
And we have to reference the Culture novels as some sort of example, here, surely :)
the level of communist naivete displayed in this post is breathtaking.
An absolutely outstanding channel. Thank you, my friend. 🙂
Fingers crossed for an immense leapfrog of technology.
Hackers have been around as long as communication.
16:58 The bulk of institutional trading is already decided by algorithms. It's been this way for a couple of decades at least.
Another GREAT presentation! Just picked up on "writing poetry" being threatened. We're not there yet, I have been testing that as a writer. Poetry requires consciousness and an authentic experience of life [I don't count eloquent rhymes as poetry]. An AI will be able to produce poetry when it has acquired those, and that's the point where we end. Watch out for poetry as a signal ;)
I feel like a vast majority of people don't really understand what modern AI such as ChatGPT really are. They're not some thinking, reasoning, rational brain. They function, more or less, like a very sophisticated and complex auto-suggest algorithm. Exactly the same as is on your phone when you're typing and it suggests the next word for you, just on steroids. It takes all the words that it knows, cuts it down to a selection of the most likely next word and then gives them a score. It just picks out whatever has the highest score as the next word in it's reply. That's more or less what it's doing and that's why current AI bots are very bad about giving weird info when they don't know something. They just return whatever gets the highest rating in their algorithm, even if that rating is effectively a zero. A check could easily be put in place where if the confidence score starts to drop below a certain value for long enough in a reply, the bot just replies that it doesn't know.
Except when you feed an auto complete system everything ever written, it starts to be able to do things that are much smarter than you would expect. Look up LLM emergent abilities.