Artificial intelligence Future

beat 18-time world champion Lee Sedol in four out of five gamesNow, normally a computer beating a human at a game like chess or checkers wouldnt be that impressiveBut Go is differentGo cannot be solved by brute forceGo cannot be predictedThere are over 10 to the 170 mIt could be terrible and it could be great. Its not clear.RightBut one thing is for sure, we will not control it.

Go is arguably the most complex board game in existenceIts goal is simpleSurround more territory than your opponentThis game has been played by humans for the past 2,500 yearsAnd is thought to be the oldest board game still being played todayHowever, its not only humans that are playing this game nowIn 2016, Google DeepMinds AlphaGo beat 18-time world champion Lee Sedol in four out of five gamesNow, normally a computer beating a human at a game like chess or checkers wouldnt be that impressiveBut

Go is differentGo cannot be solved by brute forceGo cannot be predictedThere are over 10 to the 170 moves possible in GoTo put that into perspectiveThere are only 10 to the 80 atoms in the observable universeAlphaGo was trained using data from real human Go gamesIt ran through millions of games and learned the techniques used and even made up new ones that no one had ever seenand this is very impressive aloneHowever,

what many people dont know is that only a year after AlphaGos victory over Lee SedolA brand new AI called AlphaGo Zero beat the original AlphaGoNot in four out of five gamesNot in five out of five gamesNot in ten out of ten gamesBut beat alphago 100 to zero100 games in a rowThe most impressive part?

It learned how to play with zero human interactionThis technique is more powerful than any previous versionWhy?It isnt restricted to human knowledgeNo data was given. No historical figures were givenWith just the bare-bones rulesAlphaGo Zero surpassed the previous AlphaGo in only 40 days of learningIn only 40 days, it surpassed over 2,500 years of strategy and knowledgeIt only played against itself and is now regarded as the best Go player in the worldEven though it isnt humanBut wait, if this AI learned how to play without any human interactionMade up strategies of its own and then beat us with those strategiesThen that means theres more non-human knowledge about Go than there is humanAnd if we continue to develop Artificial IntelligenceThen, that means theres going to be more and more non-human intelligenceEventually, theres going to be a point where we represent the minority of intelligenceMaybe even a very minuscule amountThats fine though, we can just turn it off. Right?

Its a thought, but think!When modern-day humans begin to take over the planetWhy didnt the chimps and Neanderthals turn us off?If this artificial intelligence becomes super intelligent and learns through and is connected to the InternetWe cant just shut down the entire Internet, theres no off switchSo, what happens if we end up stuck with AI that is constantly and exponentially getting smarter than we are?What if it gets to a point that us, humans, get in the way?And the AI hits the off switch on humani…

When people think of AI, they tend to think of super intelligent AIAI that serves the human race, but could also end us at a moments noticeBut is this REALLY going to happen?

Pop culture, mostly movies, tend to depict AI not as benevolent creations, but rather as robots with a malicious intentExcept for "TARS" in "Interstellar", hes pretty coolYeah, you can use it to find your way back in the ship after I blow you out of the airlockTheres a lot more to AI than you might thinkTheres a lot more types that serve different purposesArtificial Narrow IntelligenceAlso known as weak AI is the only form of Artificial Intelligence that humanity has created so farIt makes this sound kind of bad, but trust me it does a really good job at what its supposed to doNarrow AI is AI that was created for the sole purpose of handling one taskIts the kind of AI that AlphaGo isNarrow AI is good at speech and image recognitionand also, at playing games like chess or Goor even pretty complex games like “Dota 2”At the International 2017 World ChampionshipOpenAI’s artificial intelligence destroyed pro player Dendi 2-0But much like AlphaGo Zero, it wasnt taught how to play the gameIt played out millions of years’ worth of 1 vs. 1 matches against itself and learned on its ownIt started out barely knowing how to walk and, eventually as time went on, it surpassed human level skillIf you use Spotify,

youll see that it creates a daily mixes for you based on the music you listen toAmazon learns from and teaches itself your buying habits to suggest you new productsbut it seems an uncommon re-occurrence where these AI teach themselves how to do the task at handHow is that even possible?Well, through something called "Machine Learning""Machine Learning" is the science of trying to get computers to learn and think like we humans do"Machine Learning" is essentially the same way that babies learnWe start off as small screaming sacks of meat, but over time we improve our learningWe take in more data and more information from observations and interactionsAnd most of the time we end up pretty smartThe most popular technique out there to make a computer mimic a human brain is known as a "neural network"Our brains are pretty good at solving problemsBut each neuron in your brain is only responsible for solving a very minuscule part of any problemThink of it like an assembly line. Where each neuron in your brain has a certain job to do in order to completely solve a problemLets make a simple example of a neural networkIn order to say that someone is alive, they have to either have a pulse or they must be breathingBut not necessarily both at the same timeThis yellow dot represents a neuron in a neural networkIt functions just like a neuron in your brain doesIt takes in information and then gives an outputIf this neuron takes in the information that says "hey this person has a pulse and is breathing"Then the neuron deciphers this information and says "okay this person is alive"

It learns to analyze situations where the human would be declared aliveIf its breathing or has a pulseAnd in situations where the human would be deadWhere neither of those is trueThats essentially a bare-bones explanation of how it worksOf course no neural network is really this simpleMany have millions of parameters and are much more complex than just this one-layer networkThe world is full of sounds and visuals and just data in generalAnd we take in all of this to form our view of realityHowever, as more and more complex topics show up with more and more dataIt becomes harder and harder for humans to do this analysis on their ownThis is where machine learning comes in handyMachines can not only analyze data given to itBut also learn from it and adapt its own view of itLets go back to AlphaGo ZeroIn only 40 days, it surpassed thousands of years of strategy and knowledge and even made up some of its own strategiesBut how did it do all of this so quicklyBiological neurons in your brain operate at about 200 HertzThats proved to be fine for usBut modern transistors operate at over 2 gigahertzAn entire order of magnitude fasterThose neurons in your brain traveled to what are known as “axons”And they travel at about a hundred meters per secondWhich is pretty fast and gives us pretty good reaction timeBut its only about a third as fast as the speed of soundComputers, however, can transmit information at the speed of lightor 300 million meters per secondSo, theres quite a big difference between our brain’s capabilities and our computersIn just one week, a computer can do 20,000 years’ worth of human level research or simulations or anything that is trained to doA brain has to fit inside your head

Theres a limit to how much space they can take upBut a computer could fill an entire room or even an entire buildingNow, obviously, weak AI doesnt require an entire server room to runLike we saw with OpenAI it only took a USB stickBut for more intelligent AI, it may require much more powerArtificial General Intelligence or AGI is AI with more than a single purposeAGI is almost at or equal to human level intelligenceAnd this is where were trying to get toBut theres a problemThe more we search into it, the harder and harder it seems to be able to achieveThink about how you perceive thingsWhen someone asks you a complicated question, you have to sort through a ton of unrelated thoughts and observationsAnd articulate a concise response to that questionThis isnt exactly the easiest thing for a computer to achieveSee, humans arent able to process information at the speed of light like computers can, but we can plan thingsWe can think of smart ways to solve problems without having to brute-force through every optionGetting a computer to human level thinking is hardWe humans can create things. We invent things.We create societies and play games and laughThese are all very hard things to teach a computerHow can you teach a computer to create something that doesnt exist or hasnt even been thought of?and what would be its incentive to do so?I believe AGI or strong AI is the most important artificial intelligence to be createdAnd heres why: Machine learning is exponentialMeaning that it starts off rather slowBut theres a certain tipping point where things start to speed up drasticallyThe difference between weak AI and strong AI is millions of times larger than the difference between strong AI and super intelligent AIOnce we have the artificial general intelligence that can function like a human beingFor the most part at least, it may help us reach super intelligence level in only a few months or perhaps even weeksBut here comes another big problemSee, many people tend to see intelligence on a graph like thisWe have maybe an ant here, a mouse at about here,

the average human here, and maybe Einstein right here just aboveIf you asked most people "where a super intelligent AI would lie on this graph?"Most would probably put it somewhere around hereBut this just isnt the caseAlthough AI might not be at human level intelligence yet, it will be one dayand it wont stop at human levelItll most likely just zoom past and continue getting on more and more advanced until, eventually, the graph looks something like thisThis is what is known as “the technological singularity”Where artificial intelligence becomes so advanced to the point where theres an extreme explosion of new knowledge and informationSome that might not even be able to be understood by humansIf we make a super intelligent AI, that AI would be able to improve upon itselfAnd in turn gets smarter in a shorter amount of timeWhich means that this new and improved AI could do the same thingIt continues to repeat this process doing it faster and faster each timeThe first recreation may take a monthThe second a week, a third a dayAnd this keeps going until its billions of times smarter than all of humanityCompared to chimps, we share about 96% of their DNAWe are 96% chimpBut in that little four percentWe went from being extremely hairy and mediocre primatesTo a species that has left the planetWere a species that plans to colonize MarsA species that made missiles and millions of inventionsAll compacted down into that four percent DNA difference

The number of genetic differences between a human and a chimp is ten times smaller than the genetic difference between a mouse and a ratBut yet only being four percent different from them their entire fate seems to lie in our handsIf we want to, for some reason, put a McDonalds on a chimpanzee habitatWe just do it. We dont ask.There are 5.5 quadrillion ants on the earthAnts outnumber us 1 million to 1and yet we actually live like they dont exist at allIf we see one crawling on our table, we just smush it and move on with our dayWhat happens when the day comes where AI becomes the humans in this situation, and we become the ants?Super intelligent AI is different from software we know todayWe tend to think our software is something that we program in the computer to follow our rulesBut with highly advanced AI and through machine learningThe AI teaches itself how and what life is as well as how to improve itAfter a while, there is no need for any human interactionPerhaps humans would even slow it down, as opposed to speeding it upWe have to be careful about how we make itAs Sam Harris has statedThe first thing you wanna do is not give it access to the Internet, right?

You just gonna cage this thing, right?

Because you dont want it to get out, right?

But you want to tempt it. You want to see if its trying to get out.How do you know whether… it is called a honeypot strategy where you tempt itTo make certain moves in the direction of acquiring more powerFor example, if were somehow able to give a super intelligent AI orders and it follows those ordersIt may just take the quickest and easiest route to solve themJust because we make a super intelligent AI, that doesnt mean that its going to be wiseTheres a difference between intelligence and wisdomIntelligence is more about making mistakes and acquiring knowledge and being able to solve problems through thatWisdom, on the other hand, is about applying the correct knowledge in the most efficient wayWisdom is being able to see beyond the intelligence gained and being able to apply that to other thingsIn, hopefully, a productive wayIf we give AI an order to solve world hunger

The easiest way to solve world hunger is just to kill all life on the planet and then nothing would ever be hungry againBut, obviously, that isnt what we wantWe would have to, somehow, teach the AI to have human like valuesA moral code to follow and somehow work around thatIn a way, wed want it to be like a super intelligent human cyborgYou may not even realize it, but the majority of humans are already cyborgsHumans are already becoming less biological and more technicalWe already put our minds on non-biological thingsHow?

Well, many of you are watching this video on that device right nowPhones!Your phone has become an extension of yourselfIt can answer any question you could ever ask in a moments noticeTheres only one problem thoughYour inputs are just too slowIf we were able to create a high bandwidth link between the brain