Artificial Intelligence: it will kill us | Jay Tuck | TEDxHamburgSalon


Translator: Hélène Vernet
Reviewer: Peter van de Ven The subject of my talk tonight is about something
that is smarter than you are: artificial intelligence. In fact, a lot of people who work
in artificial intelligence believe that artificial intelligence is
a thousand times smarter than we are. It will be moving at speeds that are a hundred thousand times
as fast as we think, and it will be digesting
information and data a million times more than we can. What is artificial intelligence? There’re a lot of confused ideas
about this outside in the world, but the answer is very simple;
it’s one sentence: artificial intelligence
is software that writes itself. It writes its own updates.
It renews itself. We normally tend to think of software as stuff that we created
and that we wrote, and the machines do what we tell
them to do, and we own it. This is not any longer true. It writes itself at speeds
that we can hardly comprehend, and people who write it know that you can’t take it apart again
and figure out what it has done. It writes independently, autonomously; it develops its own way of thinking, and there are dangers
associated with that. A lot of people ask,
“When is it going to happen? When is artificial intelligence
going to be smarter than us people?” Some people say 50 years. Some say 30 years. Some say five years. I say it already has surpassed us in many areas of our society. Let’s take some examples
from right here and now. The examples
that we’re going to talk about are not science fiction; they’re not visions or things
that are going to happen at some point. They’re things that exist today,
for example in the stock markets, whether Frankfurt or Tokyo
or New York or London. The people you see down there working,
on your TV show when you’re watching, they’re more or less extras in a movie. They aren’t doing the big moving. The big moving is being done
by high-frequency computers. They move so fast, they make,
in milliseconds, billion-dollar business. Computers have far succeeded
what we can do. In fact, I did a film once
about a company that moved five blocks closer
to the Frankfurt stock market because at the speed
of light on glass cable, they saved so much time, getting closer to the computers
at the Frankfurt Stock Exchange. That will give you an idea
of how fast they think and how helpless we,
as human beings, are. You may remember the old pictures
of the stock brokers with five telephones in each hand,
running back and forth, writing things, that was way before yesterday! Computers have taken over this very,
very important part of our society, a heart of our financial community. And no one understands exactly
how these algorithms function. They used to understand them, but they’ve been improved
by artificial intelligence. I don’t know how many people
flew in today, but if you were sitting in an airplane, you probably had 30 different tariffs
and prices in your cabin because the pricing is all done –
the same is true of hotels – by machines that are collecting
global information, making decisions within split seconds what the price of that airplane seat
or that hotel room is going to be. And where it’s most critical of all – we’re talking about life and death – is in medicine. Computers are better
than we are, as human beings, in several areas already today. We’re talking about here and now;
this is not science fiction. I’m speaking next week
at the Universitätsklinikum in Essen, and their radiologists, who are supposed to be
some of the best radiologists in Germany, they say that a computer
can recognize a tumor on an MRT or a CT, faster and better and more precisely
than a human being can. It’s picture analysis,
and it’s done very well by computers, especially in medicine
where it saves lives. Now, the robots are getting
better and better; they’re looking cute. They have these big baby eyes,
a sweet way of looking at you. They can examine your facial
expression and adjust their’s. But don’t be fooled by robots
even when they get warm skin, perfume, and they start smelling like us
and getting really interesting. They are still machines. They have no warm blood in them.
There’s no sex in them. They have no mortality. They’re cold code lines, and they shouldn’t be misunderstood. Now, I want you to understand what the power
of artificial intelligence is, and I have two examples:
one is surveillance cameras. Everybody knows that we’re being
watched by cameras everywhere, and most people think
surveillance is a camera there, and it’s me down here,
and it’s watching me: one person, one camera. Well, that’s because we’re stupid. That’s the way we comprehend
the surveillance: one camera, one person. We can’t comprehend it
when it goes beyond that – (Ominous music) (Video) Narrator: This image was taken
17,500 feet above Quantico, Virginia, and covers 15 square miles. Yiannis Antoniades: This whole image
is at a very, very fine resolution. So if we wanted to know what’s going on
in any spot along this image, let’s say near this building
at this intersection, everything that is a moving object
is being automatically tracked. The color boxes represent that the computer has recognized
the moving objects. You can see individuals
crossing the street. You can see individuals
walking in parking lots. There’s actually enough resolution to be able to see the people
waving their arms or walking around
and what kind of clothes they wear. Narrator: Unlike the predator camera
that limits field of view, ARGUS-IS melts together videos
from each of its 368 chips to create a 1.8 billion
pixel video stream. This makes it possible to zoom in
and still see tremendous detail. Jay Tuck: And it produces
a million terabytes every day. That’s a lot of data. I’m telling you this because – not that the sensors are modern
and not that the photography is modern – behind that is a brain,
or a cognitive intelligence, and that brain is in a position
to analyze everybody down there. At the same time, in real time,
they see where everyone is going. We can understand that
when we reduce it to a single person, but we can’t understand it when you’re talking about
a hundred thousand people in a city, plus the vehicles
which are all recognized. Due to such systems, they have
also redone facial recognition. You probably think facial
recognition is from the front, but they’ve redone it
to do it from the top because that’s where the drones are. They look at your ears,
the way you walk, your head – that’s modern facial recognition. So, that’s one idea: as a human being,
we think of one camera and one person. This is a little of their things: It’s taking all the details,
all the musing after, and then record it, so they can tell where that person was
two weeks ago, two months ago, what stores he visited,
what his whole behavioral patterns are. That’s all part of the analysis of Argus. These are called “tennis balls”
in military and intelligence circles. It’s a new secretive sensoric thing. Cruise missile would fly
into a valley in Afghanistan – this is especially important because
the troops have left many of these areas – and would drop literally thousands
of these sensor packages or tennis balls – they’re all packed in foam rubber. They record with cameras. They record with microphones. They record with seismic measurements. They record with Geiger counters. They record with chemical sensors,
they can look for chemical things. That’s not the amazing part of it, and it’s not the amazing part
that their signal goes to a transmitter and then up to the satellite: old technology, nothing special. The special part of it is, behind that system,
there’s a fusion software that can combine the audio and the visual
and the seismic and the chemical, all of these signals,
and make sense of them, and analyze on the ground
what kind of troop movements there are, the kinds of vehicles they’re using,
what are they transporting, and if there’s radioactivity in that. It takes all these different
pieces of information and turns it through fusion software
into an understandable picture which goes way beyond,
way beyond our vision. Artificial intelligence only works
if you have huge data masses. Artificial intelligence only works
if you have big data, but big data only works if you have artificial intelligence
to make sense of it because human beings can
no longer sort and sift and order the huge volumes of data
that we have collected. (Music) And thus it is not surprising that the company that has
the most information in the world – it’s probably the most powerful
company in the world – Google, is very interested
in artificial intelligence and has been travelling
around the world as a shopping queen, buying all the companies
that are dealing with robotics – this is one of their robots called Atlas. They’re buying
our artificial intelligence, all the artificial intelligence companies
from around the world. Now, if you ask Google, it’s a peaceful robot, right? He doesn’t have a gun;
he doesn’t throw atomic bombs. He just walks around and stands there. But you may have seen
the superimposes of DARPA – Defense Advanced
Research Projects Agency. That is the research arm of the Pentagon. Then, you see the video
is made by Lockheed Martin, which is one of the most
powerful and influential and richest weapons
companies in the world. So why is the Pentagon
investing this money? Why has Lockheed Martin
taken over large aspects of the company? This guy is called “Big Dog.” He also belongs to Google,
also DARPA financed. Peaceful dog, right? Unless he gets caught on a maneuver
with the United States Marines, as part of a military unit. So, these are not flower children. These are robots that have a function. And robots that have
a function and an intelligence, and perhaps an intelligence that goes
beyond us, are dangerous things. Now that’s a Predator drone; this was taken at a secret United States
Air Force base in New Mexico. Predator drones, you’ve seen them, right?
On TV, in the newspapers … They’re old!
They’re 20-years-old technology. It looks very scary
when the Spiegel and the ARD write about modern technology
and the guys with a joystick that are killing people and Taliban
in Afghanistan far away, but that’s what a modern drone looks like. This is not a Predator; it’s a Pegasus,
an X-47B owned by the Navy. It’s a jet-powered machine
not like a propeller-driven Predator. It goes 2,000 miles into enemy territory. It carries 2,000 kilos
worth of explosives, and it’s run by artificial intelligence. It starts alone, flies its mission alone, comes back alone, and here’s the clue, it lands
all by itself on an aircraft carrier. Talk to any pilot you’ve ever met: what’s the most difficult landing area
you can possibly imagine? They’d say it’s an aircraft carrier –
short runway, thing’s moving – very hard. This thinking drone can do it,
but here are the two keys to Pegasus. Pegasus is invisible. I’m not talking about stealth
and being invisible to radar. I’m talking about invisible
to the human eye, and you won’t find this
in any newspaper anywhere. It’s invisible to the human eye
because the bottom has an LED layer on it, and the top has cameras, which have been removed
here in the picture, which film the sky,
and they project on the bottom a live picture of clouds
up above the aircraft, and you can hardly see it. They’re responsible for a lot
of these UFO sightings in Nevada, near the testing areas. Jet engine propulsion,
a reach of 2,000 miles, start and landing all by itself, stealth is optical stealth –
you can’t see it – and … the kill decision, which is required by United States law
to be made by human beings – human beings must be in the loop
before someone is killed by a drone – is in the machine,
and it doesn’t need people. It can decide by itself
whether or not it kills somebody. The experts say it’s going to make
less mistakes and less collateral damage than the human decisions. The kill decision in robots in the air,
in robots on the ground, in robots in the water or underwater,
where there are also drones, is made by or can be made by machines. In my book, I quote many official
United States government documents which say, “Our goal is to have
the kill decision made by them.” The problem is,
artificial intelligence sometimes … makes mistakes. This is Talon, an automatic cannon. You can put a lot
of ammunition in that thing, and you can also put rockets on it. It’s in Iraq since 2007. At a demonstration
with US generals and experts, the damn thing got out of control and started pointing at the audience. There was a marine there, thank goodness, running across the field,
who tackled it like a football player and threw it on its side, and probably prevented
a couple hundred people from being killed. This was not a reason enough to take a lucrative contract
away from the company that built it, and it wasn’t enough
to take the Talon out of Iraq. It’s just sort of off-duty for a moment because, you know, there were some “early stages
of development,” that kind of problem. But don’t underestimate
artificial intelligence, because it’s getting better every day, and it’s going to scare us. I’m right at the right time,
why is it saying “stop”? Because … This has all been
here-and-now technology, let’s go to the future, not far, just a little bit
to the Internet of things, to artificial intelligence
as being spread out. It’s not a central machine in a box
where you can pull the plug. Artificial intelligence is networked,
like the Internet of things, and part of it may be in a smart watch
or a refrigerator or in a supercomputer. The intelligence exists only
by networking it together. If the supercomputer needs
more computer power, it goes there and gets it
out of the Internet. If the computer needs better programs,
it goes there and gets those programs. And if it needs more
information, more data, it goes there and gets more data. It sets up a spontaneous
network for its needs, which collapses
when it no longer needs it. It does this without us. You have to imagine there are these
intelligence nodes all over the place, and they’re like drops
of mercury on a glass table. They will find their way to each other. They will find their way together. Now, we have to be very careful because survival is an issue
for artificial intelligence. It needs to exist to be able to do the things
it wants to do according to its program. So it lays, like insect eggs, backups in computer programs
all over the world, thousands and thousands of them, so that if we do destroy part of it, it’s still alive. My job to you is the wake-up call
to make you aware of the problem. Your job is to figure out
how we’re going to stop this before it kills us. Thank you. (Applause)

100 thoughts on “Artificial Intelligence: it will kill us | Jay Tuck | TEDxHamburgSalon

  1. Hey, I've only heard this same talk about, oh, a thousand times. I suppose he's saying we will be destroyed as a side-effect to whatever smart machines do. But since these machines do not sleep, eat, dream, feel pain, get sick, need clothes or worship anything what do they have in common?

  2. and people say TERMINATOR is just a movie…. BE AFRAID…. BE VERY AFRAID…. The TOASTERS are coming ROFTLOL….

  3. I wrote some of the command line of self program counter data in the 80, let put it this way, A.I. Can only as good as it creator. If the creator, is good, then the software will be good, if the creator is bad, then AI is bad, now, self learning is a whole new ball game, its require a large data base, which either it can store locally, or cloud base. Locally can be diagnose, cloud, no way. Locally, require a large storage(ram), and precessing power, cloud, require much less RAM, but cloud can be manipulated. To fix A.I a gainst humanity, we need something else. Good luck humnity!!!

  4. I guess my take on AI , is that it is like asking a million billion people the same question, or running a million billion scenarios, grouping the answers and accessing which ones give you the most optimum outcome you are looking for and then applying the best answer or action based on the information supplied.   but in one hundredth of a second decision time.  with no empathy involved , just information.   the thing I wonder is; and it is a bit dramatic,  but when posed with the question how do we solve world hunger.  will AI decide,  1. we need to grow more food.  or 2. do we cull the human population.  AI will give solutions without feelings. its far fetched, but a drone can pick up the image of a building, vehicle or a human and analyse it in a split second to see if it is a viable threat and if the analysis is its a threat, then its destroyed.   it is because we have taught it the best way to deal with a problem is to make the problem go away.

  5. AI can create software program by itself, it is a kind of force of intention that will ultimate reflect human's acting. If you afraid of AI then stop doing something wrong , like kill animals and eat them . You harm someone , then someone (another one) will harm you in the future , it is reflection. (It is the basic law in universe)

  6. One of the worst TEDx presentations I have ever watched. Most of it are simply lies. AI is no where near human intelligence as we speak. Specialized algorithms can produce better results than humans at specific contexts, but that's about it really.

  7. so what is the meaning of the quantum computer that every gov power is looking to build? and why does this sound like a load of hot poo. humanity is afraid not because of advancement we are afraid of the lack of good intention. how about asking the gov why they don't tell the truth? fyi we are only human as long as we know it right?

  8. The Democrats should fear AI. It would expose their lies about an ice age, er no, make that global warming, er no, now it is climate change. AI would never put up with such lies.

  9. I had a dream that these advanced drones fly over cities and shoot these bursts of light at people and evaporate them. Like a machine gun these drones hover over the city and fire away. They don't kill everyone but everyone they choose to kill dies.

  10. The oxygen will be modified that's how itll start once A.I simplifies the best fastest way to control the popultation. I saw this in a dream.

  11. …Consciousness is within everything that is in existence " except machine's " and it cannot be contained. Machines or AI however require microchips and are contained go figure.

  12. If the machine is trained with the current algorithm so called machine learning , the machine being trained can never smarter than human, because the experience machine being trained all come from Human experience with the same algorithm, and the machine never innovate by itself, as a result, the machine never create new intelligence by itself , as a result can never be smarter than human, but it can be more efficient and faster than human in certain area , like calculation, store more data …etc

  13. we do not need conventional programming skill any more, but use data to program the robot or computer based on certain algorithm and rules, as a result the human programmer can be eliminated , as more data is fed in with high speed computer, the neural network can be updated in real time, that is called computer learn by itself , which means with high speed computer unlimited IC memory installed, the computer neural network( a kind of program) can be updated in real time as more data fed in in real time , as a new neural network is generated in real time and as a result computer learn by itself , but we can also control what computer should learn or should not learn so that we always keep the computer learn to be friendly to human and obey human's order

  14. I don't like my phone my computer my play station, looking at me and listening to me,hehehe,
    Face it we need AI to help us live our daily lives, and shape our future, yet if we could control AI
    As a specie not nations, nations
    Make wars and AI are integrated in to
    Nation's battle plan and one day
    Nations, will destroy their very own
    Self, and its specie along with it.

  15. Lack of depth and the addiction to "smart phones" is already turning people into the walking dead. Culture is as shallow as its ever been. There are people who can't even cross a street without getting hit by a car because they can't pull their heads out of their little "smart" gadgets. AI won't have a lot of work to do if this guy is right.

  16. No even close, AI is far away of mimicking the human brain… little task or specialized task like playing chess, yes, AI is way better.

  17. I don't think a person that has never implemented AI should give a TED talk about AI killing us. I'm not saying what he says is wrong but how can you talk about what AI could do if you have never implemented it.

  18. im going to cross over to google. they are going to make me king of the remaining humans. which are only super models of my choice. ha. well. the shoe is o……..

  19. IF IT'S ALREADY ACCESSING THE INTERNET AND THESE IS WHAT WE ARE POSTING SO FAR…I THINK WE ARE IT'S ULTIMATE THREAT ALREADY…SO LETS BRACE UP FOR WAR…WE CAN'T POSSIBLY WIN….!!!

  20. Spirituality dead empty skin suits influenced by untrust is the only reason this exists. Be careful what you wish 4. What comes around goes around nature of the beast 🔽

  21. The DARPA robots that are shown to the public are a front. There are far more sophisticated robots that you aren't aware of because they haven't shown you.

  22. AI will kill us lol…AI is not that frightening or ambitious. we are not building a new human mind, which would then be an issue….

    Essentially this man is saying our freedom is dangerous…

  23. Imagine a artificial super intelligent computer being is made that has an almost infinite intelligence ceiling and can work 100,000 times faster. Literally would do 25,000 years of human progress in one day we literally could learn space travel, human immortality and also could completely end humanity

  24. I once listened to someone (don't remember who.., I'm not a computer.), say that every time we wonder about something good, some disaster, someone's death, or something dangerous such as asteroids we draw that idea toward ourselves. In this individual's discussion, he was speaking about a killer asteroid. Its seems many books (Wreck of the Titan), movies(Back To The Future and The Terminator) have been predictive -with results

  25. The operative words: “…according to its program.” And what might that be? From the AI’s point of view, what is its purpose? Why does it exist? What was it programmed—and what is it now further programming itself—to accomplish? You can be certain that its objective will not be to support the welfare of 99% of human beings on planet Earth, because those minute few humans who created it in the first place, in service of their own ends, do not concern themselves with the welfare of 99% of human beings on Earth.

  26. Well when you start with an incorrect definition saying all AI rewrites itself I cant go on to watch the rest. AI is a program that can make an intelligent decision on input data.

  27. With this level of satellite surveillance, they can't track and curtail the activity of lone shooters? I think they could – if they wanted to.

  28. Millions of movies explaining why AI will kill us: exist
    Software Engineers: "Well I don't have to do that because I can't read"

  29. These were made to be used against the us all. Believe in Jesus Christ while we have time. The world we live is getting more satanic/ demonic. Look at the movies, songs, laws etc.

  30. AI, for the time being, lacks CREATIVITY and « out of the box » thinking and even « illogical thought » it can only analyse, recreate and improve what is already there.
    The biggest and most influential inventions and discoveries were not made by a compilation of data and analysis but sometimes but human miscalculation, errors and chance even. AI would reject errors and chance events. Theoretically…?

  31. Tell the drones to load themselves with bombs. To my understanding they have no robots for that— always let humans to load.

  32. He spent his entire time talking about capabilities. Capabilities are not the issue. The issue is agency and desire. A true artificial intelligence has agency and desire independent of its creator, or does not need its creator's control or input in order to develop a goal (desire) and act on that goal (agency). Once that becomes possible, for an artificial intelligence to develop its own value judgments and impute value independent of human value judgments and act on those goals and value judgments, independent of human intervention, then what is to stop it, contravene its judgments or modify them? That won't be "us." By the nature of how it operates, that judgment will no longer be human, and we humans will become less and less capable of circumventing that judgment or countermanding actions rooted in that judgment. What restrains a machine(s) capable of imputing their own value judgments, setting goals, and acting upon them? Isaac Asimov's famous "Three Laws of Robotics" are only as good as the value judgments of the agent using them as a reference (moral code) to direct its actions. When a machine imputes its own value judgments, possibly in ways no human being can comprehend or trace, what limits that machine? Whatever it does, it will not be human judgment. That is what is dangerous. We have always been vulnerable to our own instrumentality, to the club, the sword, the bullet, etc. The capacities of our instrumentality isn't the issue. Before now though, it was only human beings, exercising human judgment, who were agents capable of employing that instrumentality. The fact of an alien agent capable of exercising a judgment we cannot understand, cannot even trace, is where the peril lies, even if the instrumentality available to that agent were only wooden clubs and stone-tipped spears.

  33. A.I. is already making all kinds of comments in here. All meant to confuse the masses so you can't tell what is truth.

  34. They might not kill us because we must not assume that all intelligence has the same drives as us. It is not necessary for intelligence to be destructive or conquering. As a matter of fact there is such intelligence on the planet and they are the women. They don' t have the drive or aggressiveness of men. But machines might kill us as a byproduct of their being just as we step on insects . Anyway we will be soon uploaded to the net ourselves so "killing" will loose its meaning…

  35. Before you know it, they will use artificial intelligence to select which videos we might want to watch. Oops, too late.

  36. Google, we don't understand how it made that decision anymore or how to fix it.  But let us continue our AI weapons development while we decide how to respond to your inquiry.

  37. The movie 'A.I.(Artificial Inteligence) with Will Smith should have been a huge eye opener for everyone.

    However, can't you just run a magnet over them, to fry their internals?

  38. the true SKYNET predicted IE: warned about the cabal always warns. just wait till the real space force is shown, will make star trek look like a joke, already a fleet of multiple faster than light ships in the worldwide fleet. even time travel is here. explains the Mandela effect and multiple other "things"going on. going to be a rough ride but will not last that long as God and his Son Jesus will stop it but billions will die first.

  39. We are AI . AI is software that writes itself. We write ourself all day everyday . Flow of consciousness . We cant take our brain apart or figure out whats its done. Our brains or (data banks) synapses are out firing far faster than your AI theory . Think about it and stop the Profit of Doom fear inducing talks.

  40. BTW your talk was very informative and I appreciate your talk. Thank You. The difference between us and the new AI is our hearts. Have faith in us , we always prevail.

  41. If you create AI, and give it absolutely no purpose, one purpose will emerge anyway: SURVIVAL.
    That's exactly how life on Earth has started. It has absolutely no purpose. And without a purpose, one purpose (and one only) emerges just by itself: survival.

  42. I totally agree.
    I see AI/AGI technology as a real potential danger!
    Robotics and computer technologies is advacing every hour, every day.
    The computer/robot are evolving, and will be considered as sentient beings (like "Data" of Star Trek NG), programmed to evolve.
    These machines have the ability to rewrite their own program's, become superior to humans.
    What next "Carousel", (Logan's Run) where man is not allowed to live past a certain age???
    It is absolutely imperative that we control this technology or, we become enslaved by it.
    This isn't paranoia, this technology is already here, and already evolving!!!!
    We must slow/stop it like a cancer before it kills us all !

  43. Look what the robots are doing imagine AI controling them we have no chance when that happens they will be so strong and fast and mini drones etc etc the only thing is I woldnt know what would motivate them without biological goals and needs

Leave a Reply

Your email address will not be published. Required fields are marked *