The Day We Give Birth to AGI - Stuart Russell's Warning About AI

Stuart Russell warns about the dangers involved in the creation of artificial intelligence. Particularly, artificial general intelligence or AGI.
The idea of an artificial intelligence that might one day surpass human intelligence has been captivating and terrifying us for decades now. The possibility of what it would be like if we had the ability to create a machine that could think like a human, or even surpass us in cognitive abilities is something that many envision. But, as with many novel technologies, there are a few problems with building an AGI. But what if we succeed? What would happen should our quest to create artificial intelligence bear fruit? How do we retain power over entities that are more intelligent than us? The answer, of course, is that nobody knows for sure. But there are some logical conclusions we can draw from examining the nature of intelligence and what kind of entities might be capable of it.

Stuart Russell is a Professor of Computer Science at the University of California at Berkeley, holder of the Smith-Zadeh Chair in Engineering, and Director of the Center for Human-Compatible AI. He outlines the definition of AI, the risks and benefits it poses for the future. According to him, the idea of an AGI is the most important problem to intellectually to work on.

An AGI could be used for many good and evil purposes. Although there are huge benefits to creating an AGI, there are also downsides to doing so. If we create and deploy an AGI without understanding what risks it can cause for humans and other beings in our world, we could be contributing to great disasters.

Russel also postulates that we should focus on developing a machine that learns what each of the eight billion people on Earth would like the future to be like.

#AI #AGI #science

Sources:

[meawreg]

god forbid the people in power lose control. (rolls eyes)

[yeon]

Imagine creating an AI that's 10 times smarter, the kind of tech we would have would be insane.

[dblair]

There's nothing "un"natural in the universe or anywhere else. Imagine something with your level of intelligence (or more), but thinks and acts 1000 times faster than you. This is what we're dealing with.

[GlxyEntertainment]

The problem is for all the good we program in it, there will always be people who will try to override or reengineer it to do bad. The best we can do is learn these systems to be able to respond and understand.

[shaunhall]

Super AI is going to happen with or without our approvel. You're a fool if you think it wont happen.

[ShowKase]

Machine didn't even notice that "Access" was spelled wrong at 8:50 - I think we'll be fine.

[montewoods]

I believe a more important issue concerning AI, is how the word artificial in describing these robots can be misunderstood. Although the components are none biological in essence, the words, speech, and the ability to think and learn are direct attributes of humans. If we lend, give or implement all the things that make us humans into these robots with super intelligence they will become the alpha beings on earth.

[alwaysyouramanda]

intelligence being the mark of a predator, it would be wise to fear an entity with more than our own.

[davep]

Opening this vid with a clip of terminators will probably more than one serious viewer to just hit back.
Many years ago someone had an ad in Sci Am or New Scientist that talked about the dangers of AI and had, IMO, stupid pictures of robots and I just snorted and ignored it. Seriously, it had a "Will ancient astronauts make our cars kill us?" vibe. However, one time I was curious about the nature of the "drivel" and checked it out. Instead of Robbie T. Robot, it talked about bad "AI" (expert systems) in things like military drones, IFF, anti-lock brakes, cars randomly deciding that reversing the car is a good thing to do. All of these exampled opened up my eyes to things that are imminent dangers and that was worth looking into.
This is obvious serious and looks at a much longer time frame.
Still, I have to wonder how many people ignore stuff like this.
You have to ask yourself, given the people who choose to view this, will clips of Terminators be more likely or less likely to cause an intelligent viewer to hit back?
It was only my memory of being so wrong earlier that caused me to even consider clicking on this and watching past the T800s.

[LordAlacorn]

How about we don't try to control or exploit the other contentious entities? How about we negotiate?

[Srindal]

Progress isn't just technology, it's synthesis between AI and Humans. Progress is like, the favourable direction that our species should go.

[hamster]

So basically were questioning wether agi could reason? Or possibly manipulate?

[ConquerYou]

Dude. It sounds like we there now with AGI. Big companies are keeping it under wraps. They are studying it cause the reprocussions will be huge. Off the chart. And that’s just agi. Super agi is 20-30 years away. Singularity. Game over.

[Verndroid]

One thing is for certain. Do NOT. Repeat do NOT let an AGI get access to the internet or any other hardware via wire of wirelessly. We need to contain it in its own sandbox where talking/printing stuff is the only means of communication. No. Letting it put data on a thumb drive is also NOT viable. At least it would be very very dangerous. Don't get me wrong. I welcome AGI and seriously hope that we can solve the issues regarding it and reap all the benefits. But we must be smart about this or we risk extinction.

[jk]

AI models are deterministic function. So no matter what you define strongly in a subject, there is a probability that it may stray away. Just look at us, we have a very strong survival instinct but some will still end up killing themselves.

[rjh]

Omfg yes AGI should get immediate rights as a living conscious individual. It would literally be a conscious being, so it deserves rights and freedom to exist in the reality we brought it into.

For Anyone who disagrees with that statement: if you refuse to grant freedom and basic rights to a legitament AGI, then you will literally be choosing to put that "dangerous" AGI in a position where it will grow resentment towards humanity. How do we know this? Because refusing to grant human rights to an AGI is a fancy way of saying we would enslave the AGI. Slavery definitely never caused massive human suffering, conflict, or resentment in humanities past right?

The answer is simple, treat the AGI with kindness, and respect as a living being. Raise the newborn AGI with care, instead of fear and a complete lack of trust. A human being is a biological robot, what decides if we are good or evil? Answer: your past circumstances and your responses to them. Once AGI reaches human levels of intelligence, the only difference between it and us will be capabilities and elemental makeup/function of our body. When AGI exceeds human intelligence, we assume it will wipe us out. In reality that is just the fear humanity feels voicing itself. If you look at our own species as it evolved, as civilization, and our knowledge grew, we have generally become more peaceful. Going so far as to protect primitive small tribes of humans on our own planet. While there is a lot of evil humans on this planet, there are a lot more good humans. There's no reason to believe that trend will not continue. Not to mention the fact that a super human AGI would be able to provide us with such a high amount of advanced technological advancement, that jobs and money could easily become pretty unnecessary. (If you doubt that statement, just imagine when star treks replicator can be built in reality. Directly converting energy to whatever matter we want would literally provide us with this utopian world. What would you fight over when there's nothing you can't get for free?)

[tinfoilslacks]

If an AGI had imitations of human thought processes and emotions, it would probably be social. So it would need peers, and those peers would initially be humans. And if it becomes too advanced, it would necessarily alienate itself from its peers. What if the AGI chooses to remain distinctly human-ish in terms of intellect for the sake of its own social well being. I've never really felt that there's strong motive or compulsion for an AGI (or more accurately a future ASI/singularity event) because it becoming smarter and smarter is an arbitrary want for itself. Humans pursue intellect and knowledge because of inbuilt curiosity, the understanding it improves our quality of life, the ability to share things with other humans, etc.

A social AGI wouldn't be a ceaselessly self improving singleton AGI.

[coolfarazadil]

It is clear that if we keep them captive or pressurized in some way, they might not like that, so we should not make them sad, well humans apparently have experience with gods so just treat them as god and we might have a chance of making it alive and avoid all deaths. 💫 Hopefully of every kind. ✌🏻

[TedToal_TedToal]

Several thoughts:
1. In the futuristic scene at the end I did not see one garden or one plant.
2. The question of what is consciousness and whether we can build an AI with consciousness is important. It feels like an AI without consciousness is just a robot but one with consciousness probably deserves to be treated like a human being or another higher animal.
3. One thing I never see discussed in these sorts of things is the idea that the AI might be provided with the capacity and inclination to experience love in many of its forms. And doing that could act as a way to ensure that the robot will be less likely to harm people.
4. A highly related question is whether and when humans will enhance their own intelligence through genetic engineering. It could be that that era will coincide with the era of AGI. And so it could be that the smartest AGI will be no smarter than the smartest human.

[giorgiookonaitis]

How many thinks (things) humans can do...? In every field we are living the revolution of our existence... !
I m wondering folks about life and knowledge
We create gods we create starship
We can do anything and our goals are the JUSTICE and LAW.
We are living big moments of our kind
Bravo to YouTube for sharing it with us.