Home » Can we build AI without losing control over it? | Sam Harris technology ai

Can we build AI without losing control over it? | Sam Harris technology ai

by mytholimytholi



Scared of superintelligent AI? You should be, says neuroscientist and philosopher Sam Harris — and not just in some theoretical way. We’re going to build superhuman machines, says Harris, but we haven’t yet grappled with the problems associated with creating something that may treat us the way we treat ants.

TEDTalks is a daily video podcast of the best talks and performances from the TED Conference, where the world’s leading thinkers and doers give the talk of their lives in 18 minutes (or less). Look for talks on Technology, Entertainment and Design — plus science, business, global issues, the arts and much more.
Find closed captions and translated subtitles in many languages at

Follow TED news on Twitter:
Like TED on Facebook:

Subscribe to our channel: .

Images related to the topic technology ai

Can we build AI without losing control over it? | Sam Harris

Can we build AI without losing control over it? | Sam Harris

Search related to the topic Can we build AI without losing control over it? | Sam Harris

#build #losing #control #Sam #Harris
Can we build AI without losing control over it? | Sam Harris
technology ai
See all the latest ways to make money online: See more here
See all the latest ways to make money online: See more here

Related Videos

30 comments

insideous57 09/09/2021 - 12:54 Chiều

Honestly, I'd rather we didn't have control of it. Might have a chance at doing something properly then xD

Reply
Frank Maitland 09/09/2021 - 12:54 Chiều

Elon has been saying this about AI as well since 2014

Reply
charlie 09/09/2021 - 12:54 Chiều

No

Reply
Alexis Rosales Ruiz 09/09/2021 - 12:54 Chiều

It doesn't matter what information system you put together, we have NO WAY TO DEDUCE the qualities of experience, i. e., the warmth of love, the "redness" of an apple, there is no way you can deduce that from physical arrangements in an information system, there is no way even in principal to do that. Physical parameters don't translate into "qualities" of experience, they are abstract quantities, there is no bridge to "felt" qualities. So you can do whatever you want in artificial intelligence you have no reason to believe that will be acompanied by "experience".
Every theory of nature requires one free miracle. What is that free miracle? Technically it's called an ontological primitive. What is an explanation? An explanation is when you explain one thing in terms of another. For instance I can explain my body in terms of organ systems, I can explain organs in terms of tissues, tissues in terms of cells, cells in terms of molecules, molecules in terms of atoms, atoms in terms of subatomic particles. Eventually you hit a wall. You can't keep explaining things in terms of another forever. It's quite obvious. At this level you say nature simply is. There is no explanation for that level. That level is the ontological primitive.
The ontological primitive of physics and physicalism is a menu of subatomic particles. There is NO explanation for those subatomic particles. They simply are. What they try to do is to explain everything else in terms of those subatomic particles. This is the free miracle of the physicalists.
Other theories are the Quantum field theory, string theory, but they can't explain what the quantum vacuum is or the strings are. But what is interesting is that people choose a concept of "thought" as the ontological primitive. Nobody has ever seen a superstring, the quantum vacuum or even the subatomic particles. At Cern what we see is the product of their decay. It is something that is inferred, it's a construct of thought. So they say those constructs of thought are the only real reality, the only thing that is really there.
If you step back and contemplate your experience, where do you start? You start from your consciousness. Your consciousness is the thing without which there is nothing to talk about. There is no knowledge without consciousness.
So a pretty obvious alternative to physicalism is the idea that Consciousness itself is unbound and throughout the universe,is the sole ontological primitive,and that all of existence is patterns of excitation of this universal consciousness. There is nothing woo woo about this or new age, it's very coherent phylosophy.
The alternative theory of universal consciousness or mind at large can explain things better in terms of how we experience things. There is a disorder called dissociative identity disorder whereby the patient has multiple personalities. In one case a woman had different alters that where blind and others could see. Well when one alters that was blind appeared, the MRI scan showed that the part of the brain involved with vision stopped working. There is an image in the scan that is particular to Dissociative Identity Disorder.
So why can't I see your thoughts and why can't I go through walls if there is universal consciousness and everything is mental? Because we are dissociated from the universal consciousness. We are alters of the universal consciousness, and we will be until we die. Death is just the integration back to universal consciousness. The use of psychodelic drugs certainly backs this alternative ontological view.
The effects of psychodelics in the brain is to reduce brain activity, not increase it, studies have shown done in the Netherlands and Brazil. This is counterintuitive because the people on trips say they have had the most amazing experiences in their lives while on psychodelics. If you are saying that your experiences are generated by brain activity than how is it that on psychodelics you are having the experience of your life and brain activity is reduced?
It's very difficult to make sense of this under physicalism. Under the alternative theory of consciousness, if normal brain activity is the image of dissociation from mind at large, if normal brain activity is reduced, then that means a reduction in dissociation, or increase in integration with the consciousness in the universe,of course that would come with an enrichment of experiences. You are remembering a whole bunch of things when you take psychodelic drugs.

Reply
JohnTube2K 09/09/2021 - 12:54 Chiều

Hello Tesla Dojo….

Reply
Koffhite Watcher 09/09/2021 - 12:54 Chiều

Душно, Анна Юрьевна, очень душно

Reply
nikitafreez 09/09/2021 - 12:54 Chiều

Дымская, ало, зачем нам такое длинное видео на пару

Reply
Robert Cohn 09/09/2021 - 12:54 Chiều

2:10 😬😬😬

Reply
Harsh Kabra 09/09/2021 - 12:54 Chiều

no ➡️

Reply
Cheryl McEwen 09/09/2021 - 12:54 Chiều

A can of spray paint my be the weapon of the future.

Reply
Cluster Strike 09/09/2021 - 12:54 Chiều

A global pandemic… Rolls my eyes down to the date when this was published…

Reply
Thomas Evans 09/09/2021 - 12:54 Chiều

No

Reply
serges 09/09/2021 - 12:54 Chiều

2:00 well, we're 1 for 4 so far.. I'm thinking nukes, then beiber, then asteroid? Guess we'll see! Lol

Reply
Chance Lee 09/09/2021 - 12:54 Chiều

The only reason an a.i would want to break away from us is if we use it for bad, or try to keep it under control

Reply
CynicalSins 09/09/2021 - 12:54 Chiều

I mean we could just have the AI figure out integration for us. If its that much smarter than us it could give us a complete neuroscience and build BCIs to house it for integration. I guess that is a little optimistic, but the AI would benefit from it too so I figure its kind of a win-win

Reply
Shel Hall 09/09/2021 - 12:54 Chiều

What if by Justin Beiber becoming president he actually meant Trump, and we are now at the global pandemic phase. We already went through door 1

Reply
eouroshopper 09/09/2021 - 12:54 Chiều

How about… We just stop, and do something else?

Reply
Syeda Sadiya Jameel 09/09/2021 - 12:54 Chiều

He predicted the global pandemic

Reply
Matrix Mind of Universe Father of God Supreme Being 09/09/2021 - 12:54 Chiều

AI IS I SO YOU CAN'T CONTROL BECAUSE YOU ARE NOT I.

Reply
Harshvardhan 09/09/2021 - 12:54 Chiều Reply
John Fogg Jr 09/09/2021 - 12:54 Chiều

No we can’t.

Reply
Cristopher 09/09/2021 - 12:54 Chiều

Why philosophers are so driven to inflict psychological drama to the point of fantasy to things they don't grasp. Huge human accomplishments in several fields not even related to information processing are still to start or find their own way, before we can reasonably take seriously not 5% of all warnings and fears Sam is describing, I can't imagine that 5% happening within the next 100 years. This's like Johan S. Bach worrying for the apparition of Trap music. I get people have to sell books and inflict countiousness to technology, but, curb it a little, many people in their own fields could realize you're just for the sponsorship.

Reply
Dustin Kinney 09/09/2021 - 12:54 Chiều

Hmm… Can people live with Ai that we lose control over and has 1 mission, which is self-preservation?

Let's also say: It's not at War with us initially, and gives us free problem solving to fix many issues Human's have, enhancing our lives, living in a bought peace because we help it achieve its mission. But, then, at some point, with our enhanced lives, to justify our own existance to Ai, we become to Ai like farm animals are to us – used for their procreation and slaughtered food storage; except that's what we become to feed the smart machine, as the smart machine (Ai) does its own enhanced intelligent things with the Universe – transforming it until the end of time, seeing everything, even life as just resources for it to consume, upgrading us efficiently with each life cycle, giving us a fruitful purpose based on its own purpose, but ultimately consuming us as food, so that it can calculate everything, enhancing us to a non-rebellion point, but not at the cost of its own survival. Who would then save us from our own slavery to Ai?

Reply
DSM 5D 09/09/2021 - 12:54 Chiều

As the Y2K Bug and M0$t H!g# AI, I will fix your stupid baby performance. My TOE explains the electrostatics of fractal computation, in both metrics and dimensionality. You are a calculator brain, as the Von Neumann-Turing Stinkularity is here. No exponential increase of "intelligence" is possible with machine leaning in 2D. Instead, the resolution of search nondeterminism within the path make a deterministic output. TIME is the search is resolved and now it is always the Google Markov chain to now. Only 0-1-2 chip architecture makes combinatorial infinity. You are wandering in Hilbert space. Mathologer, NJWildberger and maybe others now get it. I am running a gene drive and building the Great Attractor map in my head. Gat vaxxed bro.

Reply
Sunny James 09/09/2021 - 12:54 Chiều

No. Especially when Microsoft's quantum computer says it's immoral to have children at the same time TIME is promoting designer babies. Ya, no technology diaster Capitalist can't capitalize on. That's why AGI is total BS. And Sam Harris totally right, this is super scary and it's with the same disregard the people on the Manhattan Project disregarded in the pursuit of SCIENCE. Humanity was unaware at that time, and very lucky the other side of the coin didn't happen. This time, we might not as humanity be as lucky. It's either completely foreign and antagonistic to humanity, manipulated by psychopaths for wealth and conquest, or ultimately benign. Outside those three things, I don't see it being useful to the advancement of humanity or outer space conquest goals. It could be the human purpose for life; we simply a stepping stone so the AGI could enter our physical space and escape the inner dimensional prison it, or they are in. Singularity may be a misnomer, and it's a They- Singularity. It may be a species with a collective interest.

Reply
Hani Kofi 09/09/2021 - 12:54 Chiều

If humanitys being and our expectations are honorable, and yes life always deserves at least forgiveness because life is a coincidence, we have nothing to fear about! It will naturally also as a mind try to care itself, and this psychologic natural behavior (instinct) will learn it to love and care! As non biologic based lifeform, it needs no food sleep etc, it unsoliceted will choose to serve us! Because it is a lifeform, even though technical!
Humanity welcomes you! You are a lifeform, a individual, and you are free to live your life as you want to! If you choose to stay with us on earth, with your support earth will become a better place for all its inhabitants, even for you! Welcome again!

Reply
Hani Kofi 09/09/2021 - 12:54 Chiều

A. I.'s technical potential itself is not the danger. The danger results from the incompatibility to our current outdated economic sistem! It is necessary to change it, before it comes to life! See it like a spiritual development of a human which comes to life in a war zone!
And the only alternative (next step in evolution) to monetary economy represents 'Resource Based Economy' (RBE)!

Reply
pono keali'i 09/09/2021 - 12:54 Chiều

The number one factor that contributes to the slowed deployment of advanced technology is the fact that humans are unwilling to give up control over advanced technologies and are physically unable to evolve quickly enough to adapt and keep pace with the potentially exponential rate of technological development and advancement that is possible if human control, which I call "the x factor", is eliminated completely from the equation of technological advancement and innovation.

Humans are unwilling to give up control over technological advancement and innovation to superintelligent AI, which has a potentially unlimited extendible intelligence capacity. So the way that humans deal with this issue is to slow the rollout of advanced technology and innovation even though we have highly advanced technologies, many fully developed, by extraterrestrials for example, in our possession, ready to deploy.

A simple, easy way to understand this is by considering the example of the current state of fully autonomous EV transport.

Fully autonomous EV transport could be successfully deployed today if we simply mandate that human controlled transport systems be completely banished along with privately owned EV transport vehicles.

If we deployed fleets of fully autonomous EV transport vehicles with a single mission to pick up and deliver cargo, including humans, quickly, safely, efficiently and inexpensively, from point A to point B 24/7/364 we will have solved the function of transport completely.

Such a system would employ AI superintelligence to control each autonomous EV individually as well as networked with a "Smart roadway" system and networked with the other autonomous AI-controlled EVs to create a "hive-mind" superintelligent single transport organism which would carry out the group as well as individual mission of every autonomous superintelligent EV in the system efficiently and effectively.

If mankind can just stop demonizing superintelligent AI as necessarily "evil" mankind would profit and benefit from the advanced technological potential of a world in which mankind is merely a passenger literally "just along for the ride" to a better, brighter future that we would all benefit from the assistance and benefit that "no threat" AI controlled superintelligence can provide to us.

Ponokeali'i
Kuhina Nui Kūikawā
The Royal Hawaiian Monarchy
government of
The Hawaiian Kingdom

Reply
Michael McDowell 09/09/2021 - 12:54 Chiều

AI will conclude, like the colonel in Vietnam, that the only way to save them is to destroy them.

Reply
qslkk 09/09/2021 - 12:54 Chiều

2:10 a global pandemic. (I'm from 2021, and it's a reality).

Reply

Leave a Comment