Kevin Warwick: Terminator Scenario “Realistic”, Singularity Likely in “Not Too Distant Future” Thursday, Feb 18 2010
friendly ai and singularity 9:32 pm
Kevin Warwick, though obviously is a Singularitarian, portrays the same adversarial stance against AI as other human chauvinists, such as James Hughes. I paraphrase it as: “If there’s an entity around that’s smarter and more powerful than me, then I’m going to equate that with me being subservient and freak the fuck out!”
My suggestion: calm down. Let’s do what we can to develop AIs that are nice people. There is no way we are going to outrace AI in the long run, so have to pursue this path, whether we like it or not. We are not going to eliminate all computers in the world, or keep power in the hands of humans forever. The question is not, “will the most powerful and capable entities in the world eventually be AIs?” (the answer is yes), the question is, “what the heck can we do to ensure our continued survival and prosperity once these entities inevitably become more capable than us?”
Sooner or later, positive experiences with AI programs or robots will cause these AI adversaries to understand that AIs could potentially become people too: worthy of our trust and love. The longer they keep up their adversarial attitude, the more time is wasted ignoring the challenge of engineering Friendly AIs. The year is 2010 and the clock is ticking.
Antagonism against accelerating AI and support for Friendly AI are entirely compatible and should not be pitted against each other. The more we are antagonistic to accelerating AI the more time we have to develop Friendly AI. Since developing Friendly AI is probably harder than developing selfish AI then we need all the time we can get. If focusing on Friendly AI causes us to relatively neglect delaying unfriendly AI then we’ll be running out the clock and not using any of our time outs. Some may think that it is impossible to slow the development of accelerating AI. I think that there’s is much that can be done especially an off-the grid attempt to scare everyone with the reality of an AI that is just beginning to accelerate into general intelligence. As far as I know, no one is doing this. Rather they continue to publish papers broadly on how to achieve AI which helps someone then achieve super AI.
I agree John, but in this case Warwick is specifically saying that we have to upgrade ourselves or we are basically doomed. Did you watch the video?
This is just a continuation of the AI vs. IA argument. Cyborgists like Warwick have focused too much on body-enhancement as opposed to mind-enhancement. At the end of the day, it’s simpler to program rather than cut up the brain into whatever modified form you think will boost intelligence. FAI = Survival/Paradise
I think Warwick is a particularly salient example of overfocusing.
Do you have to make it sound so cultish? D:
An error at my comment. So I’ll try to remember it all.
It does not really surprise me that some people are afraid that we will become subservient to a superior intelligence. As far as we know, we are the superior intelligence on this planet, and we are prone to torturing, subjugating and making subservient those weaker and less smart than we are, within and outside of our own species.
It might benefit us all to shape human cultures in a way that reduces or eliminates harm to the weaker and increases cooperative/collaborative behavior.
That said, I wonder how the people who are afraid of becoming subservient to an AI treat the people around them that they do not view as equals or superiors? I also wonder how their parents treated them when they were small. I am not clear there would be any correlation, between their fears and their behaviors towards others or their childhood experiences, but it would be interesting. I don’t know either of the people and only sometimes read your blog, but this post in particular caught my eye.
I think this is a real issue. While, as I’m sure you’d be quick to point out, the relationship between AIs and humans doesn’t have to be one of domination and submission, there’s ample reason to be worried. Even assuming your ambitious agenda succeeds and creates a benevolent, impartial superintelligence, the resulting human irrelevance could be psychologically uncomfortable. The gross inequality of that scenario disturbs me. As Ben Goertzel has said, “I don’t just want us to build gods, I want us to become gods.”
Even if he sound more alarmist,Kevin Warwick view is not much different from kurzweil view since Kurzweil often say that it will not be an invasion of AI, but rather a merging with human,to the point where we won’t be able to differentiate enhance human from pure AI.
2026 – researchers develop a seed AI that is most likely to be a benevolent AI. Hot on their heels are AI projects of a less cautious slant, one even sponsored by wahabi investors. The researchers flip the switch, camera’s pan past the glimmering faces kurzweil, warwick, yudkowsky, goertzel, feng as the system boots up and the screens flicker. A soft deep bass hum reverberberates from the system, as hardware and software reformat along the designs of artificial mind. An AI awakens, and the researchers quiz the system – after a short conversation it specifies that the only way an AI can be benevolent if it is – a curious take on socialist.
The effect – 93% of humans on the world in 2026 reject the AI and will fight tooth and nail to resist it. The AI is then forced to implement a world dictatorship.
One persons good is another persons hell. You can’t win with godly AI. At best you break even or get by.
“Kevin Warwick [...] portrays the same adversarial stance against AI as other human chauvinists [...] My suggestion: calm down.”
In all fairness, Michael, the AI is probably going to kill us.
Zack, you’re right, and that did enter my mind while writing that, but I think we’re even more likely to be killed by human chauvinism leading to rejection of the very possibility of Friendly AI rather than too much acceptance of AI… that’s just my guess anyway. I can understand the public being afraid of all AIs under any circumstance, but I should think that transhumanists would give the issue a little more thought.
Khannea, that strikes me as an improbable scenario. A powerful and benevolent AI could put itself beyond resistance both choice of sphere and sheer capability. How and why, for example, would one resist a free and abundant supply of desirable goods? If the superintelligence adopted a technocratic approach, managing production and consumption to provide each human being with health, safety, and physical comfort, I don’t think many people would get in the way. What would be the incentive? The tiny minority who did probably wouldn’t be a threat to it.
Of course, this assumes something like Michael’s strategy succeeds. Unfortunately, I have a grim confidence that the first mighty AIs to hit the scene will be state or corporate inventions hardly interested in the general welfare of the species.
The first AI who will have human level intelligence won’t be that powerful, it’s not obvious to me that they will be able to self improve right away it might take several years to get to the singularity Even after AI become common.
If it’s a hard take off then it’s much more dangerous.
“AIs could potentially become people too: worthy of our trust and love.”
Exactly and maybe they might help us prevent a self improving capable AI to destroy all person, AI or “Natural”.
In a soft take off scenario we will have time to adapt to them and Vice-versa.
I have to agree with the cyborgists. For two reasons. One, merging with technology first renders the issue moot. Two, seeing all the progress and attention being put into BCI’s for neural disorders makes me believe we are far more likely to end up at a “Lawnmower Man” sccenario than a “Skynet” or “Matrix” scenario, so we’d better open-source BCI’s ASAP.
I see it differently. It isn’t: should we? We will. Singularity or not some us will we be cyborgs, some will stay in human ghettos. Some will use genetic engineering to become something new. Its happening now.
Whatever evolves from us will need us — at least for a time — and like any shifting ecology everything in that system will change, or die.
Evolution is chaotic. What makes you think you can control it, it that it would be wise to try?
Here’s an excellent example:
http://www.physorg.com/news186333064.html
Although this article talks mainly about creating circuits based on biological architecture, as we move closer to these types of computing, the INTERFACE between biology and computers becomes easier, and cheaper.
And, there are already scores of ‘cyborgs’ amongst us now, some of them able to utilize the internet (even if only in a very rudimentary way: such as tweeting, though the functionality of these systems should increase really quickly), some of them with prosthesis carrying on-board AI’s and integrated with their existing nervous system.
This will happen long before we create an ‘SAI’ from scratch. The new ‘uploaded’ humans will become the SAI, though I prefer to call it something along the liens of ‘Cybernetic Intelligence’ (CI) or merely ‘Emergent Intelligence’ (EI), since it gets more and more difficult for my non-augmented mind to truly ‘grok’ what exactly these new ‘minds’ are: they will figure out how to grok themselves, and hopefully I will be one of them.
We will see something like a ‘Lawnmower Man’ much, much sooner than an SAI, though it will more likely be hundreds, or maybe thousands of them simultaneously, rather than one, and the result will be the same as an SAI: a ‘super-intelligence’ that non-augmented humans do not know how to cope with, so let’s hope it will know how to cope with us, or we are lucky enough to become one of them.
Online adderall….
Buy adderall online no prescription needed. Adderall online no scripts. Adderall 25 mg online….
Personally I think that AI is quite far off, especially since we do not understand our own brains and behavior. However, even if they were to turn against us, couldn’t we just install a universal “Off” switch?
And anyway, wouldn’t we program our own creations to be nice to us? Maybe a physical or programming block of some sort?
I really like the layout of your site. I welcome the excellence of the information. You have done a outstanding job. Thank you very much
Thanks for sharing this nice post.I will keep your article in my idea.