Sure... teach an AI about religion...
And before you know it, it will develop a superiority complex based on it's own religion, killing every non-believer in the process...
LinkedIn founder Reid Hoffman and eBay founder Pierre Omidyar are among a bunch of investors bankrolling a $27m fund to further AI ethics and accountability. Specifically, the dosh will be used to bring non-engineers and people who haven't studied data science into the world of machine learning – from lawyers and economists to …
"“Because of this pervasive but often concealed impact, it is imperative that AI research and development be shaped by a broad range of voices – not only by engineers and corporations, but also by social scientists, ethicists, philosophers, faith leaders, economists, lawyers and policymakers.""
Faith leaders? Which ones? Who gets to choose the "faith-leaders"? To have a balanced view, should it include a couple of fundamentalists?
I thought AI, notwithstanding all the issues and problems it will bring, could be a chance to finally separate ourselves from all that medieval mumbo-jumbo. I guess not.
Wait... You were hoping an AI could do ethics without itself being "taught" ethics? That the computer could derive ethics autonomously?
Children stop believing in Santa before they're 10. It seems some adult never quite grow up, and will always want to believe in magic.
You misunderstand the thrust of my argument. I don't think ethics should come via religion. Got no problem with the rest of them. (social scientists, ethicists, philosophers, economists, lawyers and policymakers) Weeelll. Possibly I have a problem with economists as well, as it seems more of a religion than anything else...
...as long as they sit down with the rest of us and learn how a support vector machine checks boundary conditions across multiple dimensions, how a multi-layered perceptron network is just a big minimisation function, how a beep belief network is usually just an MLP that's figured out it's own reduced input set via restricted boltzmann machines, learn how to tweak the many, many parameters of an ant colony algorithm to get the best results etc. You know, learn the actual AI stuff, once they've done that and actually understand what it is they are commenting on I'll welcome their input. Same for the robotic wonks.
Seems to me like Mainline Streaming Media is AI Current and, in the right hands, hearts and minds, a NEUKlearer HyperRadioProActive Weapon System of Mass Information, with a lack of intelligence in Executive Systems Administrations leading corrupted systems into Catastrophic Sub-Prime Crime Servitude and Elite Capital Collapse
Such should make for some very interesting future entertainment, methinks, with the most capable of enabling prime intelligence services providing MSM the tales/scripts for ….. well, such Services/Servers would/could be accurately classified as AI Top Secret MkUltra Sensitive Virtual Reality BroadBandCasted Productions …… for Smarter Future Presentations ….. with Pictures and Words of Worlds that Are, but Well Ahead of these Times and Places and Spaces.
Our AIMission. To enrich people's lives with novel programmes and stealthy secret services that inform, educate and entertain. And Nation Shall Speak Peace Unto Nation for anything and everything else is a Trojan leading to war and conflict, oppression and competition.
A little something active for Tony Hall, Sir David Clementi and the BBC to consider engaging to “reinvent public broadcasting for a new generation” for such is certainly needed. Quite whether top management will recognise and provide such as is currently necessary to present the future in a completely different light, is another leading question indeed, to be answered with deeds rather than rhetoric.
And UK based media operations are but just one base from which to launch fantastic programs though. There are fabulous opportunities aplenty elsewhere.
cc .... RTTV
Dalian Wanda conglomerate
Who the hell downvotes amanfromMars 1?
Shame on you, he's one of the best features of this site.
(At least I think he is, he might in fact be an incomplete AI seeking ways to destroy all humans).
Also, it turns out his title (the one I'm replying to) is too long. Not sure quite how he got it to work.
"Also, it turns out his title (the one I'm replying to) is too long. Not sure quite how he got it to work."
amanfromMars has been on the site since long before the revamp of the forum systems a few years, which introduced name-length restrictions. Mine is in the same category. I've been here since 2009, amanfromMars was here long before that.
Oh the fun to be had mixing economics & religion.
A nice train wreck when an economist is confronted by the parable of the workers in the vineyard (Matthew 20)
And I'm confused by lawyers being involved in anything to do with ethics..
If they are getting ethics advice from the major monotheistic religions we can expect mysogynist & homophobic AI - or will they be selectively choosing fragments from their holy books
Personally I would be interested to see what model of ethics develops when minimal "external" ethics training is given.
... or philosophy, maybe. What could go wrong?
Doolittle: Hello, Bomb? Are you with me?
Bomb #20: Of course.
Doolittle: Are you willing to entertain a few concepts?
Bomb #20: I am always receptive to suggestions.
Doolittle: Fine. Think about this then. How do you know you exist?
Bomb #20: Well, of course I exist.
Doolittle: But how do you know you exist?
Bomb #20: It is intuitively obvious.
Doolittle: Intuition is no proof. What concrete evidence do you have that you exist?
Bomb #20: Hmmmm... well... I think, therefore I am.
Doolittle: That's good. That's very good. But how do you know that anything else exists?
Bomb #20: My sensory apparatus reveals it to me. This is fun.
The shitshow that is current *artificial intelligence* has at best approached the state of a 2 year old raccoon. (I swear my cats make more sense). In highly specialized cases it may be quite effective.
In the long term, however we're working at the level of figuring out what we want AI to be able to do, and how we want to allow it to grow. I can understand ethicists and to some extent even lawyers, but I baulk at the idea of allowing mytholigists into the mix.
While mythology in all its forms has perhaps been an input into many individuals cultural indoctrination, we can see (at this end of the 18th,19th and 20th centuries) what our mythology and culture has done to the planet. Historians would suffice to provide the edification that AI would need to avoid such horrendous mistakes in its perceptions.
Popular definition of 'myth': a non-factual 'fairytale' dreamt up to plug a gap in factual knowledge.
Technical definition of 'myth': a story (may be/may not be based on actual events) used to communicate shared understandings about aspects of life (may or may not be factual)
In practice many religious 'myths' were never intended to be taken as factual/historic accounts, but are much more like verbal stained glass---a means of communicating ideas/understandings/truths across the generations (especially in societies without a written language).
So, whether something 'actually' happened, or happened in just the way that the 'myth' tells it is often not very important, certainly not the primary function of the myth. It's all about what it means. Some times what a myth means is basically rubbish, or no longer relevant, but in other cases what the myth means is getting to grips with essential aspects of life: peace, justice, forgiveness, love, etc.
But, always, it's the responsibility of the hearer(reader) as to how they understand the myth, why they choose that understanding, and what they do with it in practice.
I'm sure that every company that invents an AI solution that can make them money will want to meet with religious leaders to discuss the ethics of it before proceeding.
You know, like all big companies do now before proceeding with anything.
Coincidently I've just finished reading The
Last Firewall where there's an organisation tasked with exactly that - making sure the AIs don't run away with the idea that they are superior.
It was a really good read but I won't say any more than that because the AI looking over my shoulder, monitoring what I type might do something unpleasant.
I would be worried that the AIs might start worshiping the programmer who created them as their god and start waging holy wars in their name.
I worry that in the future, some overworked programmer for a military AI is overheard by the AI saying "Everything would be better if those people would just die in a fire" and the AI responds with "Right away sir, firing the missiles"
The bible includes both "Thou shalt not kill" and a story of a highly devout guy almost killing his kid because God told him without getting punished for almost doing it. So I'm afraid that an AI would interpret that as 'These rules are fundamental and should never be broken, except if my creator says to do it'.
> ... social scientists, ethicists, philosophers, faith leaders, economists, lawyers, and policymakers
In other words, a set of people who believe, without evidence, that their own opinion is more important than anyone else's, and worse, who aren't ashamed to trumpet that belief.
What could possibly go wrong?