Elon Musk on Artificial Intelligence (and the Basics of AI) – Documentary

elon musk has a big brain but even he can't keep up with artificial intelligence which is why he started neuralink saying that instead of trying to beat ai we should join it by connecting our brains to computers elon has already put artificial intelligence into cars letting tesla cars think and drive for themselves but what other sci-fi ways could artificial intelligence be used in the future from curing diseases and predicting when someone will pass away to creating worldwide crises this video takes a look at what elon musk has to say about artificial intelligence and we will also go over the basics of ai explaining what it actually is why it is important and why elon musk thinks it's dangerous and his book recommendations for understanding artificial intelligence and we will even have an ai create a small clip for this video asking the ai how will elon take over the world [Music] elon musk has a lot of fears when it comes to artificial intelligence even having nightmares about it the reason why artificial intelligence can be such a force is that it creates powerful computing that can work on its own this can help solve problems that would take us humans years to solve for example you can feed in worldwide data about diseases and the artificial intelligence could spot patterns to help find cures artificial intelligence is also what is making self-driving cars possible but it is what the future of artificial intelligence has in store for us as humans that elon musk is focusing on elon has said that ai will be the best or worst thing for humanity and that we are heading for a point in time when ai will be vastly smarter than humans which could be in the next few years and at this point things could become unstable and weird one scary problem for elon is that robots will be able to do everything better than humans and he is not sure exactly what to do about this he says that he is not generally for oversight and regulations but with ai we need to make sure that people are developing it safely saying that if weapons are regulated so should artificial intelligence vladimir putin has said that the nation that leads an a.i will be the ruler of the world elon musk talks about how the biggest risk with artificial intelligence is not that it will start to think for itself and try to take over but that the artificial intelligence will follow the will of the people that programmed it and the people that programmed the ai they don't even have to have bad intentions but if they have not fully thought out the functions of the ai there will be bad outcomes elon's example for this is a hedge fund who wants to use next-gen artificial intelligence to maximize profits based on this goal the ai then decides that the best way to do this is to invest in defense stocks the ai could then find ways to boost the stock prices by instigating battles and protests just by joining social media groups it's just like on an episode of silicon valley when the ai son of anton was asked to debug and fix code and the artificial intelligence thought that the code was so bad that the most efficient way to fix it was to just delete it all son of anton was also asked to find cheap hamburgers for lunch it then calculated the best option and went ahead and ordered 4 000 pounds of meat to the office the crazy speed of improvement is also what scares elon he refers to an artificial intelligence program which is now owned by google called alphago which plays the board game go the artificial intelligence went from not being able to beat a good go player to beating the european world champion in six to nine months and then beating the world champion soon after alpha zero came along which learned how to play by playing itself alpha zero crushed alphago a hundred games to zero elon says that you can basically put any rules for a game in front of it and that it will read the rules and play the game at a super human level artificial intelligence needs to be built in union with humanity this is why he started neurolink so that we don't get left behind connecting our brains with computers will allow for a wide range of applications from prosthetics to treating diseases you could also stream music directly into the brain or even play video games and you wouldn't even need controllers as your brain controls what is happening in the game and your senses will be triggered in your brain while you play elon tweeted saying that if you can't beat them join them let's take a step back and break down artificial intelligence into its basic parts ai is a branch of computer science it is a computer program that can work by itself a basic way of explaining artificial intelligence is that it tries to recreate the human brain our brains get fed information from all around us and it processes this information and makes sense of it there are a number of ways that our brains get fed information for example through our eyes this is a branch of artificial intelligence called computer vision where ai's use cameras and sensors to see our world and then works on identifying objects and reacting in ways that the ai has been programmed to just like a tesla car that uses sensors and cameras to steer the car while on autopilot or a robot arm fitted with a camera and has been programmed to only pick up red jelly beans another way our brains take in information is through speaking and listening this is a branch of artificial intelligence called speech recognition the ai works on understanding what humans are saying and is able to respond or perform tasks just like alexa siri or google's assistant we also feed our brains information by reading text this branch of ai is called nlp or natural language processing where ais are used to understand text which it can then translate or figure out what the topic is about one example of this is amazon analyzing what you're searching for on its website as well as your past purchases and then give future recommendations with all of the information being taken in our brains are able to see patterns in our world this is a branch of artificial intelligence called pattern recognition and machine learning this is where ais are given large amounts of data and it can then see patterns such as patterns and diseases or even stock prices and investments and the artificial intelligence can even act on this information much faster and much more accurately than humans artificial intelligence systems of today are narrow in function they can only perform a singular task or several similar tasks for example the autopilot ai is good at driving cars but it is not able to play the board game go as it has not been trained to these kinds of ai are low risk to humanity but artificial general intelligence is an ai system that will learn many tasks and is capable of learning many more this kind of ai is often self-aware and is the kind of ai that could be a risk for humanity [Music] since the future of artificial intelligence is always on elon musk's mind he has a few books that he recommends people to read the links are in the description there is the book called life 3.0 being human in the age of artificial intelligence this book looks at how ai will affect the different areas of our lives elon says that it is a compelling guide to the challenges and choices in our quest for a great future of life intelligence and consciousness on earth and beyond another book that elon recommends is called our final invention artificial intelligence and the end of the human era this book looks at the risks of rushing the development of ai without taking any precautions lastly there is super intelligence which is about understanding the future of humanity when machines pass humans when it comes to intelligence links are in the description below for those books let's move on to examples of artificial intelligence working in the real world today how are tesla cars able to drive by themselves let's break down the basics of the artificial intelligence that is working inside of a tesla car us humans have our eyes and ears feeding information to our brains when driving which figures out what needs to be done next when it comes to steering accelerating and braking a tesla autopilot gets its data from eight cameras that cover the car and can see up to 250 meters away these cameras can spot other cars read road signs and see road markings there are also 12 ultrasonic sensors which let the car know how far away objects are and also helps with parking and then there is the forward-facing radar as an added layer of security and is ideal for rain snow and mist conditions all of the data that gets collected is inputted into a computer chip that is in the car which then outputs steering acceleration and braking commands elon has said that in the future the autopilot will let tesla cars be 100 to 200 times safer than human drivers all of the data collected by the car also gets sent back to tesla hq where there is a central computer that is receiving data from hundreds of thousands of tesla vehicles it is training and learning from the data to make a safer and better version of itself that will then get uploaded back into the tesla cars we can take a look outside of elon's world and see artificial intelligence being used in the likes of facebook and instagram facebook directs people towards adverts using the language ai this ai takes the text that a user creates and posts and analyzes the meaning and trends to provide personalized product recommendations facebook also uses an ai to recognize people in photos and can sometimes be more effective than humans in recognizing whether two different images are of the same person or not when it comes to instagram ai is used to show you the content it thinks you want to see the most the ai looks at what kind of posts you have commented on or liked and who you follow and what they like it then uses this data to find the content you are most likely to enjoy keeping you on the app longer speaking to a group of students on an ask me anything video elon said that our mindset when it comes to social media is quite weak if ais are programmed to maximize advertising clicks an example being facebook then this can create all sorts of crises if the goal is to drive clicks could the ai then send people down dark paths elon warned that we should be on the lookout and monitoring social media ai to see if it's being used to manipulate people's minds saying that we should use an ai to detect if an ai is being used on social media elon wants to use ai against ai openai is another company that elon musk has started this one is an artificial intelligence research lab one of the projects that they've been working on is called gpt you can ask it questions or ask it to write articles or stories for you for example you can write the first line of a story and it will complete it for you the second version of the program called gpt2 was deemed too powerful so they took it down a hoax article that sounded just like a human could be written just by feeding it one line it is able to do this because it has studied hundreds of billions of texts one of the people testing gpt-3 had it write an article about itself and what it could be used for the ai said such things as i was thinking of how cool it would be to build a twitter-like service where the only posts are gpt-3 outputs the ai is also able to study comments and predict replies and write responses that are able to fool humans for this video we asked an ai to take our kindergarten-like sketches and turn them into semi-realistic images that we then pieced together into a video clip we also asked an ai the question how will elon take over the world this is what it said elon will take over the world by the snap of his fingers [Music] but it's a mystery and you won't tell me oh okay sorry no reason to be scared this is just a party [Music] what other sci-fi and freaky things could an artificial intelligence do in the future how about predicting death something that is being worked on at the university of nottingham where an ai is being trained to predict the risk of early death due to a chronic disease what about merging advanced robots with a.i once a robot can be made to look just like a human then artificial intelligence can be used to learn how humans speak move and interact with their environments this would lead to a future where humans cannot tell the difference between robots from other humans in our other video on the future of gaming we talked about how video games will use sensors to monitor you while you play artificial intelligence will be used to build the game as you play designing it based on what you like and don't like the game could even build new worlds within the game itself taking you deeper into the simulation the question remains how will artificial intelligence change the course of humanity will we need to join it merging our brains to computers to survive and will we be able to create artificial intelligence that protects humans by keeping other artificial intelligences under control on the next episode of venture city we take a look at amazon and their growing robotics empire hit the subscribe and thumbs up button to not miss a video

Channel: Venture City

Views: 144,259

Rating: 4.8431835 out of 5

Keywords: Elon Musk, Artificial Intelligence, ai, Neuralink, Tesla, Machine Learning, Neural Networks, Documentary, Basics of Artificial Intelligence, Fundamentals, For Beginners, Self Driving, Future of AI, Future of artificial intelligence

Id: 1Pql2nNR80k

Length: 14min 7sec (847 seconds)

Published: Sat Oct 17 2020