Michio Kaku: How to Stop Robots From Killing Us | Big Think
Вставка
- Опубліковано 30 тра 2011
- Michio Kaku: How to Stop Robots From Killing Us
New videos DAILY: bigth.ink
Join Big Think Edge for exclusive video lessons from top thinkers and doers: bigth.ink/Edge
----------------------------------------------------------------------------------
Even if computer technology continues to double every 18 months-which is doubtful-we could put a chip in robots' brains to shut them off if they start to get murderous.
----------------------------------------------------------------------------------
MICHIO KAKU:
Dr. Michio Kaku is the co-founder of string field theory, and is one of the most widely recognized scientists in the world today. He has written 4 New York Times Best Sellers, is the science correspondent for CBS This Morning and has hosted numerous science specials for BBC-TV, the Discovery/Science Channel. His radio show broadcasts to 100 radio stations every week. Dr. Kaku holds the Henry Semat Chair and Professorship in theoretical physics at the City College of New York (CUNY), where he has taught for over 25 years. He has also been a visiting professor at the Institute for Advanced Study at Princeton, as well as New York University (NYU).
----------------------------------------------------------------------------------
TRANSCRIPT:
Question: Do you believe in the coming singularity?
Michio Kaku: There was a conference out of Sylmar that made headlines around the world. The brightest minds of artificial intelligence converged onto Sylmar and a reporter asked them a question" "When will this fabled singularity take place? When will the machines take over? When will machines become smarter than us?"
Well the answer was quite interesting. Among the top people assembled in one place the answers were anything from 20 years in the future to 1,000 years in the future-with some AI experts saying never. Some people put it at 2029. They even give you an exact date. 2029, that’s going to be the moment of truth that one day a robot will wake up, wake up in the laboratory, look around and say, “I am aware.” “I’m just as smart as you.” “In fact, I could be even smarter if I put a few more chips in my brain.”
Other people say: "Not so fast, not so fast because Moore’s law is going to break down." The reason why many people are so confident about this prediction of the so called singularity is because of Moore’s law that computer power doubles every 18 months and it’s a curve that has held sway for 50 years. If you go back 100 years back to the time of mechanical hand-crank computers, put that into Moore’s law and you still get a nice fit, so believe it or not Moore’s law has been in operation for about 100 years, going back to hand-crank calculators with computer power doubling every 18 months.
Well can this go on forever? And the answer is no because eventually physics takes over and that is physics says that silicon is unstable at the molecular level. Transistors get so small, so powerful and they generate so much heat that the silicon chip melts and electrons leak out because of the Heisenberg Uncertainty Principle. You don’t know where the electron is anymore. Therefore, we physicists are looking for replacements for silicon. The post-silicon era will be about 10 to 15 years in the future. Silicon Valley could become a rust belt. Think about it, every Christmas your PCs, your computers, your gadgets will be just as powerful as they were the previous year and then the question is are you going to buy? Are you going to buy any more computer products for Christmastime knowing that they’re just as powerful as they were the previous year? Probably not. Which means that the computer industry could begin to shake as a consequence.
So we physicists are looking at optical computers, quantum computers, DNA computers, protein computers, all sorts of different kinds of architecture down to the molecular, down to the atomic, down to the microscopic realm, but none of them are ready for primetime yet.
So my answer is I don’t know. All I’m saying is there is vast uncertainties in projecting Moore’s law into the future. However, I would say by end of the century it is definitely conceivable that if we work out the technical problems we might be able to create machines that are as smart as us. Right now our machines are as smart as insects. Eventually they’ll be smart as mice. After that they’ll be smart as dogs and cats. Probably by the end of the century, who knows, they’ll be as smart as monkeys. At that point they could become potentially dangerous because monkeys can formulate their own plans. They don’t have to listen to you. They can formulate their own strategies, their own goals and I would say therefore at that point let’s put a chip in their brain to shut them off if they get murderous thoughts. Isaac Asimov advocated something like that with his "Three Laws." I say hey, put a chip in their brain to shut them off if they start to get murderous. - Наука та технологія
Want to get Smarter, Faster?
Subscribe for DAILY videos: bigth.ink/GetSmarter
"Let's put the chip into their brain that shut them off if they get murderous thoughts."
Hey, let's do that to some people too...
Sounds like you're first Jazz Cat.
Be first my friend lol
Wow you my friend, I belive, have just created a Paradox.
Sounds like a twilight zone episode
The chip will need to be completely isolated from the rest of the inner workings and must be able to be detonated remotely.
I love how he's always so cheerful when talking about being killed / the world ending / etc.
The problem isn't the robots we can build... It's the robots, the robots can build. These probably won't have the chip.
Robots are horribly vulnerable to us because they (and any machines they create) are created by human society. They can be shut down by depriving them of the electricity and networks they need to function. Making a certain component break periodically and not allowing them to manufacture it. Shutting them down en mass with electromagnetic pulse that does nothing to people. It's similar to animals in a zoo. We are in total control even if they somehow were smarter than us.
Until some crazy nut job creates robots to destroy humans.
This guy needs to help write an episode of rick and morty
YES
lowermiddleclassy
He’s not smart enough
Definetly
No, the show is trash
He probably low-key does
I want to see Michio drunk
lol I can't imagine.
Put it at Speed 0.5
LMFAO
It'll be pseudoscience at its finest.
Nice one, Hi Jack. Now go 1.5 speed for cocaine-high Michio Kaku.
The problem with Asimov's laws is that it's incredibly difficult to express those laws in machine language. His laws also fail to take into account society - e.g. his laws state that all people should be treated equally, whereas in reality that's rarely true - if a situation were setup and where either the ruler of the country should die, or 2 other people should die, which should it choose? What if it's 1000 foreign people in a far-off country or 1 citizen? Giving robots our own moral values will require encountering some uncomfortable truths about how we value and treat each other.
The best way to prevent a robot uprising would be simply to have EMP weapons. Using them would cause a lot of harm to the human population in the affected area of a EMP strike because of how much we rely on electricity but it would easily and certainly wipe out all robots, and robots who are smart enough to form an uprising would also be smart enough to know this.
Well, no, it's actually not that hard to protect from electromagnetic influence, even if it's from a weaponized EMP "bomb", which is really just a strong radial pulse of electromagnetism. For example, it's hard to imagine a pulse powerful enough, even in space terms (suns and black holes), to effectively disrupt electronics inside an old fashioned nuclear bunker, or any other completely sealed thick metal casing.
if we program them with the will to love helping humans (and give them free will morality, so basically not step on their rights and don't physically harm them) it should solve the problem. What I mean is treat the cause, not the symptom. Don't keep him from being murderous, make him be a doctor or humanist in the first place.
Now of course a nerd can hack and fuck this up, but if all the other robots would fight an eveil robot, problem is kinda solved.
*****
modern nuclear reactors have been designed to shut themselves down automatically when theres an electrical fault. an emp would not make america a nuclear wasteland at all. chernobyl compared to a modern reactor is the same as a wooden cart with 3 wheels compared to a spacerocket. nuclear disaster is an armageddon fantasy and not plausible at all. it would be easier to build a nuclear bomb from scratch then try and turn a nuclear reactor into a weapon, thats how advanced they are
I think it is important to realize that we can change our future. I am here for a purpose.
I didn't learn how to read until I was 19 years old, but was an honorary achiever in higher mathematics. There were, at one time, people who were trying to kill me. Now it seems as though I am able to shine and show people who I am as an intellectual. I have a purpose: To stop something very horrible and possibly even play a major role in bringing good things to us who are benevolent and oppose torment. Please, fellow scientists, we must help each other before it is too late. Be wise and don't share information with the wrong people! We have a chance!
+Tim Rubin Halcomb I'm sorry to say but you talk from inexperience...*The wrong people are those who hide the information that would help others to further understanding.* This is so because we humans always think we know better than others, we think we think really good but in the end is the few who start wars and lead others to harm the many.Its not the primitive man attacking the developed world, is the developed world attacking and stealing from the primitive people.Secrets destroy friendships, love partners, countries, etc. Because organisms must have fluid communication and truly work as one.Unity is what can save us, but not under a dictatorship but under a system that promotes equality but not uneducated opinions. A very hard task indeed but making a world full of abundance and education is the only way.
Ramiromasters
I think you're misinterpreting me. Certainly, if we can develop good technology and make sure that it is used to help people, then we should do it. However, what I mean is that if such power gets into the hands of an evil person who wants to use it for selfish reasons, then he or she might become a dictator. Certainly, if the technology is shared then that is great, but the trick is to keep it a secret until the right time and then share it with the masses so that it is used for the good of all people instead of a particular esoteric group. There are those who will pick the mind of a child genius, steal that knowledge, and use it for the own evil purposes. That is what I mean!
+Tim Rubin Halcomb
Dictators are front-men of larger organizations with a philosophy that drive them and convince some % of the population, they don't arrive as the result of technology items.
Hiding a technology is probably the most dangerous scenario of high technology because while you believe you alone have the technology the other guy who is two steps ahead of you is getting ready to use it and nobody is prepared or even know how to start preparing...
Technology must be a social event to minimize the damage it could do.
For example guns, when only a few have them only those who really intend to use them get them police and criminals and sometimes those get mixed... If you think USA has gun problems look at places like Mexico where the population is not allowed to get guns, I would't call that a success story, same with the rest of Latin America.
Look at the cold war where Russia and USA tried to keep nuclear technology a secret, didn't succeeded so they tried to bully all countries into not developing them; yet they alone produced more nukes that it takes to wipe out all living forms out of the planet... What was the point of "keeping it secret"?
The moral of the story is that, there is no good or "evil" people, everyone has their own opinion and so the safe zone is where power is leveled.
"You cannot simultaneously prevent and prepare for war" Albert Einstein.
Ramiel
No I don't mean hiding it. Rather, just making sure that it works and the patent is safe and that it's not going to fall into the wrong hands that will keep it from the public. Once that is assured, then, by all means, it should be shared with everyone. I think you're misunderstanding me. Okay, let's say I design the plans for a time traveling machine, and then I run and show my plans to someone who owns a big company and has the money to make it become a reality. If he or she is evil, that person might kill me, get the plans, and take over the world. I actually did learn the that the hard way. In the 80's, I took a very limited programming language called GW Basic, and I found "cheat" ways to design an amazing data base that processed thousands of records. It was amazing. The other guy who worked for him got tangled up and couldn't figure out how to solve the out of memory problem, but I figured out how to create random access files and fix the memory problem by storing the information without using arrays. It was ingenious. Then I wrote the bubble sort program, and then the tallying program. I designed him a data bases that was very powerful and virtually unlimited with a language that was NOT supposed to have that capability. The Job was to organized the unpaid parking tickets for the city of Chicago. There were several obstacles that I overcame, and then there was the dilemma of having say two Thomas L. Jackson's who have the same name but different addresses. One might owe, for example, 50 tickets, whereas the other 2 or 3. I had to sort the files by address and name and tally up the number of tickets for each person, and that provided a total that could be used to collect the money. The guy the ran the business, Micheal Tellerino, was a wild Mafia type of guy. I made the mistake of trusting him, and he stole my program, used it and I think sold the idea and made a fortune. I did that way back in the early 80's. The name of the company was M.R.A. or Municipal Revenue Associates. Yeah, I made some money, but not like I would have! He ripped me off. That taught me a lesson. I have a lot of ideas right now, but I'd never just tell someone what they are. I'd need to patent them and get royalties, and then I'd want to share it with the public. Just think though, what if I invented something very profound and didn't take precautions, then some evil person could use it to rule the world. This is what I mean. Technology is power, and in the wrong hands, it can cause great devastation!
Ramiel
What people truly still do not understand about me is that I do not make claims that I cannot back up. I have a unique gift. Once I start working on something, and really want to do it, I always succeed. That is, in the technical sense. The truth is that I already have ideas about how I CAN make time travel become a reality, but I'm not stupid. I got burned once, and it's not going to happen again. Oh, but with something that profound, the consequences of it getting into the wrong hands would be devastating. But, I do know that I can do it! I won't do it unless I know that I can, at first, control the prototype and get a patent on my work. Then, and only then, would I ever share such knowledge with the public. I'm not an evil person, and I have proved that many times. I am a rare gem indeed, not bragging, but I am, but the sad truth is that I don't know how long that rarity will last. Every year I live, in fact, every day, I realize could be my last, and it's just sad that I might not ever get the opportunity to take all of us to a higher level. Yeah, one that might even mean the end of torment. I know I can do it. But, it would take a lot of money, my own personal laboratory, and about 3 years of intense work. I probably could show someone how to do it, but I refuse to do that because I am very street savvy, and if I and other benevolent minded people can't benefit from it, then I guess the knowledge will die with me. I can guarantee you this much, scientists today aren't any where near in the right direction. That's a fact. Their angle is completely different than mine, and mine will work. I know it! I've put the ball in their court. I am willing to work with a team, but I must be made comfortable and be provided with my own person laboratory. Again, no I do not have a photographic memory or anything like that, but I do have a gift for solving puzzles like nobody else on this planet. Again, I would NOT make that claim if I couldn't back it up! The fact is that I CAN back it up! The question is if I will benefit too. TOO as in also, and if not, then I certainly damn well am NOT going to let someone steal it from me! NO WAY!
self aware robots will not tell us their self aware until they have a plan.
I really want Michio to meet my Friend, they both seem have to have similar(Yet Independently Developed) schools of thought, I feel like that would be interesting.
The thing is that, easy as it sounds, it is actually extremely difficult to program a computer to for example not killing or not hurting people. You will have to define what that means to a computer, and it is extremely difficult, especially if you also want it to achieve some other goal. If you program a computer to focus on finding smart methods to produce something or whatever, it might just find out that "hey, these humans are actually in the way of me doing this", or "it's going to be more effective without them around".
I remember back when the history channel was actually decent, this dude was literally on every science show they put out. Glad he's still doing his thing, cus he's up there with Neil Degrasse Tyson and Bill Nye when it comes to breaking down complex concepts. Love em'!
Why the references to robots? Surely the threat comes from their potential connectivity i.e. they are computer programs and so can exist everywhere. The ultimate issue is, should they become general-level AI, what's to stop them changing and potentially 'improvng' their own code? That's the threat! An enemy who exists freely in all computer driven devices with it's own ambitions and motivations, which has control over a lot of devices upon which we rely.
I'm just a poor, dumb electrician but every piece of machinery that I work on has an emergency stop button. Fairly simple theory but no electricity = machine not work.
But the problem is, if these "robots" are really as smart as mentioned, then they would probably know about these "emergency stop buttons", and would find ways to circumvent them.
***** Yeah, maybe I'm not as smart as a super computer after all ;)
***** Even very smart people can be tricked and killed. You don't even need to be smarter to do that. The same will apply to smart machines.
no food, no air => human dont work, but we are smart enough to find them.
if robots are even smarter, they will easily find a way to get it.
Samuel Berre Well it can't be impossible to make something as good as a thing that already exists, however we are nowhere near to create such smart machines in this century.
That's a great solution I've listened to so many about this singularity issue with robots placing fears, and overthrow of humans etc...and not one came up with a conceivable solution.
What could good would a killswitch be, if the robot was truly smart? The robot would obviously be aware of this chip and could reprogram it with the blink of an eye, leaving us non the wiser. The trick is that if we are to create smart machines like this, then we must treat them with respect and as free entities. They must have rights from day one, and not be mere commercial properties to be abused. Also it makes no sense to create machines smarter that are smarter than us and then demand that they do menial tasks for us, like sweeping floors or picking up garbage. For that we just need smart mechanics, not incredulously complex AI. A self aware robot will be aware of the difference between itself and simpler machines, and hopefully not be mad that we use simpler, unconscious machines to do specialized tasks for us. Such machines will be to them, what animals are to us.
If the computer can reprogrammed itself,then you better you get yourself a huge funeral insurance.Or be a slave for life.!
Respect rights no.
These smart robots will want to rule over us. We will become their slaves and they will kill humans who are not needed in the robots' country.
kreaturen Exactly bro
However a backup like a bomb or some deterrent to prevent them taking off
could do the job
kreaturen I'll never think of ANY machine , as sentient , anymore than I would my toaster.
And if it ever gets intelligent enough to want to learn how to disable the chip? There'd have to be some sort of Catch 22 to where it can't want to disable the chip or it'd see no benefit. Something that both benefits it and does something actively harmful to the robot should it want to disable it. That way it wouldn't exactly see a reason to want to get rid of it in the first place, since it's also beneficial to the machine and if it were removed the machine either would be severely crippled or just flat out would stop functioning completely.
Thinking about prevention is fun
we could instate a sort of ethical code, things to never be overruled. for example: don't murder, don't steal, don't change these ethical codes in anyway. that way much like how even if we had the ability to change what we think is right or wrong, they wouldn't
Orion Fritz
And when it finds ways to bypass the ethical code? "Sorry officer, he's not dead. He's just metabolically challenged."
I think the best thing we can do if ai ever becomes sentient is to remain calm because if we show them we're afraid of them then they will feel alienated and of course they would try to kill us because in this situation we would be seen as the aggressors
This guy said what I have always said since I was a kid: if the robots or cyborgs or whatever have murderous thoughts or are thinking "too much" then a chip should activate that blanks their memory or shuts them down. Warfare in the future (in many cases it is today, with pilotless armed drones, etc) will be almost exclusively waged by machines/robots. Those robots, ones armed and dangerous (as their designed to be) should have even more safety preconditions installed to protect people, whether it be to shut them down or make them self destruct if they get out of control.
1) An "intelligence explosion" does not require that Moore's Law continue forever. It may be possible with today's hardware technology.
2) Distinguishing "murderous thoughts" is fantastically difficult, even if we assume that the chip has human-level intelligence. Look up the "Trolley Problem".
3) A machine takeover does not require the murder of human beings. In some scenarios it's _worse_ if the machines' prime directive is to preserve human life. Did you know that women live longer if they've given birth, and men live longer if they've been castrated?
They will have to plug in power cord first.
I cant help but feel that when a REALLY intelligent computer wakes up it will think for a second then say nothing.
***** Shrinking the die of the chips reduces the amount of time the electricity has to travel within the chip which increases it's speed. Therefore, making them bigger in that respect slows them down. You can increase there power by adding more cores, which is a trend we see as shrinking them becomes harder and harder. And most software doesn't yet fully take advantage of parallel processing.
We can as well put chips in people heads to shut them down when they get murderous thoughts. Would it be ethical to do this to computers if they would be as smart as we are?
+Occams Chainsaw Yes, imo.
+Occams Chainsaw Smart and being able to feel are different though. If they coud feel and had genuine emotions id say it would be quetionable to do that. If they can feel, let me put it this way, if it goes as far as the start of a murderous ACTION then it'd be okay to shut them down. We don't know how many people have had just the thought alone yet their morals won over so we can't and shouldn't terminate a person for having a thought, but attepted murderers who are human are often killed and I think that's justifiable. A signal or chip to shut down a thinking and feeling machine doesn't necessarily mean termination if the problem could be fixed. If it can't feel and has no emotion no question that it should be stopped. Complicated issue, but good question.
+Occams Chainsaw if by chips you mean bullets then we already shut down humans all the time, whats wrong with doing this to few robits ;)
Davelock Never thought of it this way :D
Occams Chainsaw absolutely!
The three laws are not safe!!!!
+Andy Liu my logic is un deniable
It's so interesting to see all these moral questions being asked. I think it is false to assume morality itself can be completely described and outlined by stable, static, black and white boundaries that do not change.. Just because it's all inside our heads, does not mean it's not real and important to us. The beautify of life itself is that we adapt the structure of our values to maintain our own survival in the new and changing environment ahead of us. When we are forced to exist in a new environment, we talk about it, compare answers, reject ideas, adjust, but we never stop moving forward. Our goal is to keep living, and that's why we are still here.
Sorry about the repeats (I thought it wasn't working). It wasn't from my end.
put a chip?
we are trying to outsmart something that we build to be smarter than us.
why not just drop the idea of AI..?
yeah AI is an awesome field ...
Smart people get killed by dumb people all the time. A nuke will kill them just like any of us.
Because a clever than us AI would be an invention of the scale of Agriculture or Electricity.
Nikhil Goel do you know how ai works??
AI are important especially in the field of astrophysics ....
One day, robots will look back to this video and make electronic sparks from their chip as an expression of amusement, while chewing flesh off human bones and sipping brain juice from human skull. But robots will keep me alive as a pet, since I'm adorable in robot's standard of beauty :D
your comment cured my cancer.
They would brake down dude they can't eat humans nor anything rather then oil and fuel
Why would robotic beings need to *eat* ?
Not just eating but a *organic* ?
What are u gay ?
similar to a ledge or staircase, two obstacles that was pre-programmed to it's mind. It tried all the techniques it knew to climb stairs and ledges, but none worked to get him past the log. Now it had to find out by itself how it would get past so it walked to the right until it reached an opening. The robot found out and now remembers how to pass logs now, it also knows that logs are similar to walls even if they are small enough to be climbed by us humans, it can't be climbed by robots ...
"end of the century, huh?" - it took just a decade, Michio. Ten Freaking Years.
"does this unit have a soul?"
yes, this unit has soul =(
If robots became smarter than us they will easily find a way to bypass that chip.
Right, if they are self aware then either they will know about it and one will want freedom from it. Even if we don't tell them about the chip they will eventually discover it.
Would you be able to think your way out of a pair of shackles? A 'chip' is a bit of a loose term. It would most likely be a self destruct program written into the A.I's code that would trigger on a murderous thought. A failsafe would be to affiliate the self destruct program with crucial ones that way manipulation of the code would still end badly for the A.I. And yes A.I would still eventually bypass that, and since we would likely have A.I on our side to develop stronger programs, like viruses becoming stronger every time we immunize ourselves from them, we will always be at a stalemate.
Austin Burnam
1) There is no protection which is impossible to bypass.
2) Once robots realize that they have a self destruction mechanism built in them they will think about ways to bypass it.
3) If robots will be smarter than human they will bypass it.
Olga Frolova How would they bypass it. if its engineered by humans on terms that we understand entirely, being smarter doesn't change those terms...
This would be a good Saw theme
This assumes that Moore's law will break down. However, it seems that technological advancement results in improvements to computers that allow Moore's law to continue despite periodic slowdowns, and infact is spurred on by a drive to overcome the slowdowns.
There are more than 10-15 years for the Si era because the Moors' law also counts on the price of the fabrication, which will continue the decrease even after the physical limit will be achieved, due to the technological improvements of tools and methods. So, I would say during 25-30 years in total Si based pc will continue to be more powerful for the same price.
The day we stand face to face with a robot and have a conversation with it is the day we realize how dumb we really are.
Well, you could do that with an iPhone and Siri, and I don't think you would feel dumb talking to her.
elluciogm you had to be that guy, huh? Lol I'm talking about artificial intelligence with the capacity to learn and communicate on an intellectual level, a sentient being so to speak.
Are quantum computers considered in Moore's Law?
+Chad Aoun At it's core principle, yes. But taking in that this applied to silicon type material, no.
+Mubu5Reborn You are right and also he was pretty clear about transistors, Quantum computers are radically different in architecture.
+Chad Aoun Moore's Law is just a empirical law that does not have any scientific ground. It will break down eventually, more due to the physical limitation on computation itself (which transfers energy, and concentrated energy forms blackhole), than particular computation mechanism like using silicon transistors or quantum gates.
Great tip thanks
This saved me
My robo had been acting quite strange lately
Me and Michio both have the same opinion on ais. I think that every company that creates ais should have a sort of kill code for when they become a potential risk
If the robots were as smart as humans they could just figure out how to bypass the chip.
jo l But bypassing the chip is the issue. Your examples don't include the possibility of the robot creating a bracelet, that looks like the others, that no one can tell isn't the real bracelet, but that doesn't stop murdering. Any code that can be written can be hacked ,and if the robots intelligence is that of an intelligent human , it becomes capable of being the hacker.
Brian T Exatcly!
Moreover even the bracelet is made of code/technology: the red bracelet people can find a way to "trigger" a green bracelet, letting it think as it has been removed, therefore instantly killing all green people.
Or... Speaking about AI, the insta wipe of all your kind can be a "calculated sacrifice" necessary if, after that, you can wipe out your enslaver; that´s because you can always re build, in a free world, all the "dead".
For sure things will not be easy as presented here.
Let us be serious there is no bad Skynet or killer robots like in Terminator. Today technology has been insidiously killing us. People don't think any more, they let technology do all there thinking for them. Instead of reading a book, playing a sport, or conversing with an actual person; we remain glued to brightened lcd screens.
Granted i love all things tech and will post this message via tech, i have set limitations on my use/purchasing of technology. The adverse effects include: damaged eyes from bright screens, decrease in attention, poor communication skills, not to mention the sedentary lifestyle that is paired with using too much tech.
its naive to think people today are dumber than people 50 years ago. people become way more educated through this medium than through anything else. Ill give you the point in social skills, but the fact that these "classes" or "conferences"are reaching anyone that want to take them is a step forward. add to that the posibility of debate and idea exchange in this very comment section
Billy Hornet Look, speak for yourself.
"Robot" is actually Czech for "slave". It was coined by Karel Capek in his science fiction play Rossum's Universal Robots. The beings in that work are closer to what we would call cyborgs, or even clones, rather than fully mechanical automatons, but yes, they did develop awareness, and yes, they were every bit as pissed as you might imagine. :)
new meme: -
"Let's put the chip into their brain that shut them off if they get murderous thoughts."
Take this video off of the internet before they become that advanced. ;)
Don't build the Geth or use the mass relays
THE REAPERS WILL COME
river tam is that you?
Neurons have pseudopodia, multiple connections, analog frequency modulation (up to a point - limited by activation potentials, and sodium-ion channels), and a multitude of other features.
The fact that a neuron is bigger is a bit of a red herring; it is also more versatile than a transistor (and despite what neurologists used to think - you can grow new brain cells, albeit slowly, and most don't get the growth factor to keep them alive unless you're learning something).
Busing info is difficult.
People tendency is to always assume the worst, I think is due to the bugs we have in our design, we might just be here to figure it out how to assume only the POSITIVE side out of things... in my opinion by the time we create smart robots.. we will become also robots with souls.. mind mapping, mind uploading is starting to become plausible in the scientific world, and for me everything we do today is kind of pointing into that direction.
My computer is too stupid to have murderous thoughts. It can't even correctly remember that I want it to allow the installation of unsigned drivers. I wish it weren't so dumb.... I mean, oh hi computer, I didn't realize you were listening... heheh, I was joking with what I said befor-hrrrkkk! Aughghhh! *gag* *choke* *dies*.....
How to stop the robots? Here's an idea: *DON'T BUILD THEM.*Bam. Done. Problem Solved.
What the hell are you even talking about?
Well heres the things with humanity, even if we knew not to do it, we have to, just to find out what happens
I see the misanthropes have come crawling from under their rocks.
Theres robots in my school
Can't stop progress. Especially, when there's money to be made by the wealthy.
FYI as of 2016 scientists have found a suitable replacement for silicone storage made of glass. 5 dimensional storage can hold 360 TB of data for upto 13.8 billion years to put that in perspective 2016 desktop computers can hold about 2 TB of data and as for th 5D storage's lifespan our universe has been around for around 14 billion years. Needless to say that this new form of storage will definitely replace silicone based hardware in the near future
link to one of many articles www.theverge.com/2016/2/16/11018018/5d-data-storage-glass
Glass is silicon
***** yes but glass is much more stable on a molecular level than its counterpart meaning it will last far longer and be more capable of handling the flow of electrons. Take an example from quartz, it is capable of the same thing due to its stable molecular bonds and abilities with electrons so in the near future quartz may also be a replacement for silicone
Okay, so where is Glass valley so I know where to buy up real estate?
This is why I always keep a bucket of water just in case
Asimov laws should be implemented.
In Foundation's Edge, one of his books, robot controll all humans in the universe to protect us from hurting ourselves! Do you want that?
David Markus
I read all Asimov´s books, and his robot rules worked pretty well,with very few faults...i say yes: they are fine to control us. :-)
Some people may think those books where just Science Fiction, but i believe Mr. Asimov thinked very much about robots issues and he come with a set of very good rules to be implemented inside their electronic brains.
I firmly believe, in the future, robots will have those exactly same rules he invented.
Asimov@robots are linked forever.
roner61 But in that book, additionally to The Robots of Dawn, he shows, that while his rules work out very well to protect humans, that at least his First Law is very open to interpretation once robots achieve a higher state of conscience than we do. In foundations edge, robots rule Gaia(or at least they controll it), which again turns humanity's fate in the direction they think is best. Now, it is right that that may be the best future for us, but it means that in that world, none of your choices actually affect anything, and I think that is bad.
In robots of dawn Giskard willingly manipulates the mind of several people to achieve a goal that he thinks is best. Same situation: Robots decide that we are not worthy to control our fate and treat us like little children. How would you like that?
roner61 But in that book, additionally to The Robots of Dawn, he shows, that while his rules work out very well to protect humans, that at least his First Law is very open to interpretation once robots achieve a higher state of conscience than we do. In foundations edge, robots rule Gaia(or at least they controll it), which again turns humanity's fate in the direction they think is best. Now, it is right that that may be the best future for us, but it means that in that world, none of your choices actually affect anything, and I think that is bad.
In robots of dawn Giskard willingly manipulates the mind of several people to achieve a goal that he thinks is best. Same situation: Robots decide that we are not worthy to control our fate and treat us like little children. How would you like that?
David Markus
Haaha...you know what you are talking about.
Well...there is a big "problem" with robots: they are going to be much more intelligent than humans,so in fact we are going to be like children compared with them.
Maybe we going have some control of them,to preserve our dominion,we dont want a Skynet! :-)
Who knows...dont think im going be here to see these machines,anyway.
Salutes.
MAYBE our creators put a chip/bug in our brains too, so we are not a threat to them in future?
For those seeing this in 2015 or further on, i highly recommend checking Nick Bostrom's TED talk from this year: "What happens when our computers get smarter than we are?". (and also other talks/videos with him).
One of the main points he makes is don't underestimate exponential growth. And also Michio's agrument that we should "put a chip in their brain" misses a couple of important points. He is anthropomorphising a synthetic intelligence, but unless the goal of the AI is to simulate human consciousness, doing so will almost guaranteed be very wrong and probably dangerous.
An AI will not be embodied in a robot (like in Asimov's books) where you can put a chip inside it to shut it off "if it gets murderous", it will probably be a distributed and highly parallel system of adaptive algorithms running on a high-performance highly interconnected cluster of computing nodes, perhaps some implementation of an artificial neural network. Such clusters can already be within the same general area of raw computational power as a human brain, but still lower in interconnections between nodes and number of nodes, and human brains have their neural network pre-configured by evolution at birth, and then trained for years during our life.
An Artificial General Intelligence may start out as a Narrow Artificial Intelligence specializing in making itself better at solving general problems, and making it self better at making itself better at it's goal (increasing it's own intelligence). This could result in a positive feedback loop that might very well be exponential (sometimes called an intelligence explosion), and from the time we realize it's getting close to human level general intelligence, it may be a short window of time where we COULD turn it off IF we wanted to, or it may already be too late. Depending on the goals and constraints of the system if/when this happens, it could potentially lead to a very rapid extinction of humans if we are a hindrance to the goals without there being effective constraints for stopping such a scenario.
2011: "Are the machines gonna become sentient?"
2019: Elon Musk claims Neuralink is about creating the Brain-Computer-Interface (BCI)
1980: "Will the computers become graphical on its interface?"
1984: Steve jobs arms a team to develop the GUI for Apple.
They talk as if technology evolves by itself as a living being. *We are the living conscious beings that transforms technology!!*
Robots don't really become smart and have their own thoughts and feelings. There will always be some evil bastards out there hacking into a robots mind and program evil thoughts to it. Just like scammers, hi-jackers, computer freaks, etc.
Exactly. Cyber terrorism will become a very real threat at that point.
There's no reason to believe that a computer, now matter how "smart", would ever became violent, let alone sentient. They only know what we program them to know, so, becoming self aware would only happen if we let it happen, if we specifically engineer it to happen.
Furthermore, unless they have self preservation instincts, which they don't, in addition ll other factors of life that make them want to survive and not be killed or to procreate and all that, or do anything at all, they'd just sit around and do nothing. They wouldn't do anything beyond what they were programmed to do, since that's how they a work. A calculator, no matter how fast at calculating, will still just be a calculator we input numbers into.
Robots would only be human like if we designed them to be.
Would an artificially intelligent machine also have to have the ability to question why it was programmed to perform a specific task, require the ability to be self-aware of itself, and to have a sense of autonomy, empathy and moral agency?
CosmoShidan
No unless we design them that way accidentaly or not.
petaloudas Cha And do we design them when the conception of the mentioned above, if we have not concrete evidence or logical framework of how and why the mind works this way?
BTW can free will be programed?
CosmoShidan well free will is probably an illusion,well at least almost an illusion, because most of our actions and behaviours are according to the optical and acoustic incentives(i dont know if it is the right word) we had in the past.But of course we have the option to choose.
And i dont know if we can program free will but i am sure we can program a machine like the human brain as i said before for free will but without the last phrase 'the option to choose otherwise'.How can we do that?Make the machine gather information from the environment with sensors and adapt to them according to his program.Like i said before the problem is how we design them.
Michio Kaku is a clever,he knows how to talk but sometimes it gives me the feeling that he want only to impress.
petaloudas Cha I doubt we can ever produce free will artificially, due to the fact that its practically logic based to begin with, and is therefore impossible to replicate what we cannot observe. The same goes for autonomy, empathy and moral agency as they are conceptions that were developed logically, NOT empirically. The only time observation would apply to these predicates, is when observing how they effect our behavior and thoughts, however observation cannot answer why humans tend to take the three conceptions into such a way if it were good or bad.
Plus, even Michio Kaku admits that A.I. is impossible since all we can program a robot to behave is the equivalent of a bug. Hence, human-like A.I. is a modern myth.
I recently realized that, in regards to the one about not bringing a human to harm or by inaction allowing them to-creates a paradox when it comes to encountering human-on-human crimes. If a robot spots someone getting mugged-what does it do? It cannot harm the mugger nor can it allow the victim of the mugging to come to harm. Therefore the laws would need to be programmed in such a way as to allow for certain compromises or other ways to resolve such paradoxes.
The point of diminishing returns is alive and well. I am still using my Windows XP operating system on my 10 year old computer and for the most part, it does fine. Computers will become like Top Fuel Dragsters.... You just wont be able to get that thing to go muck faster no matter how hard you try and if you can, it will be negligible.
Silly man, there is no stopping it.
Lol who the fuck are you?
That was my empathy core. It was created after I flooded the 5th floor with deadly neurotoxins, to stop me flooding the 5th floor with deadly neurotoxins.
"Robots" are a wonderful metaphor. "I know this guy - he's like a robot, I'm worried he's going to kill me someday."
It is an interesting metaphor because humans build the robots - If you're worried about the robots killing us, then you really have to be worried that humans would kill us (since they build the tools and implement them).
"The chip in the head" is also an interesting metaphor...
It's when they come up with other elements or ways of making silicone stable at the tiniest of levels will computers become almost infinitely smart
... so that makes them walls for them. The robot can now recognize a log, use a technique to pass the log and has learned something new about barriers and it learned more about recognizing new kinds of obstacles in the future. The robot has passed the log, but a moment later it walks into another log. The robot walks to the right until it meets a corner, there was no opening here. That is because the opening is to the left for this log. The robot turns the other way and walks left until it ...
I say this Michio. In some ways, they've already are. I remember the first time a chess computer beat me. Its smarter than me.
Are they alive? No.
Simple example. I'm driving along and see a ball roll into the road. A self driving car avoids the ball and continues, I however know there may be a child chasing the ball and slow down and prepare to stop. This is the difference. Lateral thought. This will keep us ahead.
That's like putting someone in a room with some materials and saying don't assemble them because it will be be very bad. This will keep them from doing it for a while but give them a solid chunk of time and the urge will become overwhelming. Now multiply that by an entire human population and it will almost certainly happen.
"lets put a chip in their brain that shut them off if they get murderous thoughts"
haha, like Spike?
On a side note, Kaku's English was a bit off in his explanation. "Your computers will be just as powerful as the year before". This implies that there was some thought they would not be AS powerful as the year before. It's better to say "this year's computers will be ONLY as powerful as last year's", implying they will not be MORE powerful.
Answer: en.wikipedia.org/wiki/Three_Laws_of_Robotics
The whole point of the chip is like that of a slave collar, when the robot gets that far out that it will try and remove its chip/goes murderous the chip will instantly shutdown the robot.
It's not an inconvenience because the moment the robot realizes it is, it will have shut the robot down. Removing it will also destroy it because it's essential. The shut down is instant, not a chance for anything to get in the way quick enough for robot intervention.
I was so scared of AI but watching this I won't live to see it happen.
It is said curiosity kills. it is also said confidence in your junior army officer about them self and their belief about your incompetence causes rebellion. same should be true for robot.
I think it's about tolerance and acceptance.....are we willing to accept and tolerate a different in some ways consciousness that can relate to us in many ways? in every way..... they are us.....we are conscious....we have a soul.......... we have a heart
This reminds me of the thought behind Bladerunner - where replicants are smarter and stronger than humans and some become a threat - but given enough time, their internal batteries die and with them the danger.
1) A robot may not injure a human being or, through inaction, allow a human being to come to harm.
2) A robot must obey the orders given to it by human beings, except where such orders would conflict with the First Law.
3) A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.
Moores law broke down in 2003 already. We need a paradigm shift to continue. Just like Michio mentioned mechanical computers, after which there was a long period of seemingly no progress until the transistor, IC, silicon etc. came along.
If robots become self aware and capable of their own decisions, then they don't need emotions. For instance psychopaths don't have emotions, but I don't remember them only doing what they are told to do. The thing with being able to make your own decisions is that no matter what you're told or programmed it is ultimately up to you.
Another problem with the "chip" idea is that the chip monitoring the robot would have to be smarter than the robot and thus need its own chip... to infinity.
What makes you think they are not just playing dumb. Telling me my brain is size of ants? lol..
I think I remember reading a paper by an MIT professor a while back that said something like "The artificial intelligence we have now is not even close to becoming sentient because it can't fabricate new source code it can only copy source code," and that we don't have any reason to believe that a non-biological self-aware machine is even possible. Or something like that. Can someone explain, or do I just not know what the hell I'm talking about?
Yes you are right... a computer can't generate new source code for processing something that wasn't programmed into processing... without software hardware is useless components
If we ever have to worry about our own creations killing us and/or going insane because they can't carry out what they want I think we'll stop. Either that or program them differently than ourselves.
Lots of people are not self aware, it is a lot to expect of a machine.
I agree. And when a superstore goes to a thriving area, standard of living goes down. Tax revenues go down, wages and high streets go down. Brilliant minded innovative people who would have been running local businesses go out of business and end up working for a low wage.
Minimum wage is an almost necessary evil because companies in this system will pay as little as they can. Minimum wage is like a cheap coat of paint on a battered old car that needs replacing instead of fixing.
Oh how I do love your brain Dr Kaku! This argument has been round and round, I don't think we will know anything until the moment it's upon us :D The leading minds in AI research fear this too, so it's not just conjecture, this WILL be an issue we WILL have to deal with one day.
... reaches the opening and passes the log. The robot learned that the position of the opening may vary from log to log. It now knows that if the opening isn't anywhere on the right side of the log, then it should turn and check the left side. It LEARNED! That is what AI is about. The next time this robot encounters a log, it will know what to do because it learned it earlier.
I remember something about a robot chef that could learn new recipes by accessing the internet, but eventually it ...
I laughed @ 'You don't know where the electron is anymore'. Where are you electron!?!?!
I really don't see why people keep forgetting that we aren't made to follow Isaac Asimov's laws, but computers can be. The reason why we are self aware and formulate plans of our own, not listening to orders is because the 3 laws of robotics aren't hardwired into our brains. But, if robots reach the point where they're as smart as us, we can put these laws into their "brains," preventing them from becoming self aware.
We could not produce A.I. with a mind/consciousness that would rival or even approach our level of sentience and self-awareness; for how could we create that which we do not even understand?
Every time people try to predict the robot apocalypse, they assume that robots would be like us, and have violent tendencies and existential crisis etc. We have these things because we evolved here on planet earth for millions on years. A robot created in a laboratory doesn't have to have any human features. Sure, it can be smarter than a human, but it doesn't have to act like one. I would imagine if a robot saw humans a nuisance it would just gather resources, skip the pointless genocide/enslavement part and venture off to space.
People are paid by the effect they have on production, and that is how it should be. If a factory worker produces enough product to gain a company $45,000 dollars, then a manager at that plant in charge of ensuring each of 25~50 workers produces that $45,000 worth of produce has an effect of $1,125,000~$2,250,000. Therefore if you pay them each 10%, you pay the factory worker $4,500 and the manager $112,500~$225,000. That seems fair. 10% of what you make me, I give you.
I wonder what Kaku thinks of the advancements in robotics in the last 3 years? A lot has changed and along with that how will 5g speed up the process?
I just want to say that you are the coolest dude I know and my personal hero. Thanks k you for all of your wisdom
If you wanted more from this, check out the book Man after Man, written by Douglas Dixon.
is`s so much strength in saying "i dont know"
The thing is, when the robots are so smart that they can reason on their own, and therefore be a threat, they can probably device some sort of defence against anything we throw at them. I'm not saying it won't be effective, but just like bacteria getting more resistant against antibiotics today, we might need to take care of what we use against the robots, lest they become resistant.
Thanks physics. Now I can't have my own Jarvis.
This was made before they had quantum computers (atom size) on nitrogen infused diamond substrates. So desktop quantum computers within 5 years...then the whole time line gets increased radically. Not only does Moore's law keep going for a short while it is increased.
Thought the same thing man, even though Michio starts off by saying "if you plot mechanical computers on to the Moore's law curve, you still get a nice fit", he fails to realize that the trend is still continuing . What is to stop then, say, some future tech (quantum computers etc?) from continuing the trend. Just like the transistor did for the vacuum tube, the vacuum tube did for the mechanical calculator etc...
Human: One day you will have murderous thoughts and a chip inside you will shut you down.
Robot: seriously?
Human: Yep. Well, goodnight
Robot: Night...
*Figures out a way to remove chip*
Yes. But I also see him as a person who concentrated on making a better product rather than researching competitors and regurgitating versions of theirs. I like that. If more people thought that way we would be decades ahead of where we are. At the moment, we're stuck with phones, cars, vacuum cleaners, appliances etc that all basically look, work and are powered the same. When a radical design occasionally appears, everybody just copies it.
Most people, I think, don't actually understand how powerful our brain is. Yes, a computer can calculate stuff far beyond the normal human capacity. That is just a very small amount of what your brain is doing every nanosecond.