Of late it's become increasingly clear to me (as it should to anyone with as limited a grasp of the (UK) law as myself who gives it even a moments passing thought) that everyone selling products labelled as 'Artificial Intelligence' is in a dodgy position.
The problem is there isn't any artificial intelligence & calling a thing something it's not is (for product sales purposes in the UK) illegal.
If something isn't done soon all of them will end up broke & in jail having been sued silly by all those disgruntled customers who subsequently discover they've bought a Chinese Room.
<Customer: "I ordered Artificial Intelligence... Artificial Intelligence I say, not a Chinese room, what do I want with a Chinese room? I don't even live in China!">
Which would be a shame as (most of them) are reasonably nice blokes.-- Skewed, Jul 09 2014 Synthetic Intelligence Campaign Synthetic_20Intelligence_20CampaignPrior Art [8th of 7, Jul 09 2014] 3d dragon wotsit http://www.moillusi...om/dragon-illusion/ [not_morrison_rm, Jul 10 2014] New limited solutions to the three body problem http://news.science...-three-body-problemMight interest some. [RayfordSteele, Jul 10 2014] Article describing the AI behind Sims 4 http://www.popularm...the-sims-4-16906802 [theircompetitor, Jul 10 2014] On/Off switch for consciousness http://www.extremet...ep-within-the-brain [theircompetitor, Jul 10 2014] Piff and tosh! You'll be asking marketing departments to stop making misleading claims next!
And before we know it, we'll have to believe up to 100% of advertising copy (based on a sample size of 4 readers who expressed a preference) and expecting our 4x4's to have 16 wheels, daily orgasms from our haircare products, and the appearance of our wrinkles to be noticeably reduced, because, apparently, "We're worth it".
What happens if it turns out, empirically, after conducting repeated trials and subjecting the results to peer review, that we're not worth it?
Sending the marketing department to jail would be great, but then who'd pick up the drinks bill at the office party? Because Finance sure ain't.-- zen_tom, Jul 09 2014 //based on a sample size of 4 readers who expressed a preference//
based on a sample size of 4 readers chosen by the manufacturer from those who responded to the questionnaire who expressed a preference?-- Skewed, Jul 09 2014 Exactly! ;)-- zen_tom, Jul 09 2014 We prefer the term "inhuman intelligence".
But ... <link>-- 8th of 7, Jul 09 2014 Wittgenstein....if you could teach a lion to speak, you couldn't understand it....
ergo if the AI is just like us, so a bit pointless as the human race has been creating other humans for quite some time....unsklilled labour and not a blueprint in sight...
Or the AI is nothing like us, so no way to communicate and so not useful..-- not_morrison_rm, Jul 09 2014 [marked-for-deletion] rant.
Anyways, eventually SF writers will start using the term "Artificial Sentience". Ironic that they've waited this long.-- FlyingToaster, Jul 09 2014 //We prefer the term "inhuman intelligence"//
Nice :)
However, this still implies actual intelligence.
A database lookup system & a couple of clever algorithms?
It's a bit like walking through the woods, kicking a stone which bounces into a stick & going "oh my gosh, it moved, it must be alive"
Now my preferred choice of words on the other hand describes what's actually happening perfectly, all it implies is that under the right conditions (from a certain angle with appropriate lighting) it can 'look' like intelligence.
Much like a good drawing of a rose can be mistaken for a real one (if you'd just put one eye out with a sharp pencil so losing all depth perception).
<edit>
[FT] ditto Artificial Sentience.
Ah...
Just checked your link [8th].
Should I fold this one up & go anno yours now (I seek the wisdom of more established bakers)?-- Skewed, Jul 09 2014 //rant//
So soon?, I thought I'd disguised it so well ;)-- Skewed, Jul 09 2014 More or less by definition, it isn't intelligence if you understand how it works. Ergo programmed machines will never be intelligent.
Also, when we learn enough about neuroanatomy, humans will stop being intelligent.
(True story: a fellow researcher in my lab studies nematode behaviour. He finally, after many years, worked out the series of neural connections and activities that make the worm turn toward or away from stimuli. As soon as he'd figured it out, it was demoted from "behaviour" to "chemotaxis".)-- MaxwellBuchanan, Jul 09 2014 //when we learn enough about neuroanatomy, humans will stop being intelligent//
Hard not to conclude we're all just clockwork when you think about it & extend what we know to it's logical conclusion.
Moving on into philosophy & (nomenclature?), intelligence is the word we use to describe what our brain does (more or less), so whether our understanding of what that actually is gets redefined or not I'd posit the word we use for it will remain 'intelligence'?
But some other things not currently called intelligence may be reclassified as intelligence, while others (like database lookup systems & algorithms) some don't think should be will be confirmed in their title ;)-- Skewed, Jul 09 2014 //Neural nets are definitely not if/then type constructs//
Really?
I assumed it was just more of the same but multiple ones networked together in some way?-- Skewed, Jul 09 2014 //Neural nets are definitely not if/then type constructs.//
Yes they are. As long as they run on digital computers, they boil down to the operation of logic gates. You can add noise (generated by analogue system, if you like) as an input, but everything between input and output is digital.
That's not to say that the output of a neural net is easy to predict, but ultimately you could (if you wanted to) simulate it with a huge number of cogwheels, babbagesquely.
As an aside: the "three body problem" (three objects, none of which is insignificantly small, interacting by gravitation) is insoluble - i.e. you can't produce a formula which predicts the positions of the bodies at any given time in the future. A computer can simulate their motion to any given degree of accuracy to find their positions at any given time; but the output of the simulation can't be predicted in any less time than the simulation takes to run. I think that's spooky.-- MaxwellBuchanan, Jul 09 2014 [Max] Can't or Haven't. Down to how complete the knowledge on gravity and initial conditions is, I suspect. Ultimately God is still only a huge dataset of facts in however dimensions.-- wjt, Jul 10 2014 //a good drawing of a rose can be mistaken for a real one
I prefer the green dragon one, it's much more cool. Linky. Bugger to put together the first time.-- not_morrison_rm, Jul 10 2014 The stimulus to solve problems is an emergent biological behavior -- it can no doubt be simulated eventually, but at significant risk, given that entities concerned with self preservation instincts would compete for resources. It is not clear to me that one can create a useful pseudo biological system without self-preservation drives.
The ability to analyze data for patterns invisible to the typical intellect is obviously already out there, and is likely to see the most significant "useful" progress over the next quarter century.-- theircompetitor, Jul 10 2014 //[Max] Can't or Haven't//
I believe it's "can't", though obviously if it becomes a "have" then it will have have to have been a "haven't".-- MaxwellBuchanan, Jul 10 2014 Is it just me or did that guy in the 3d dragon link sound more than a little like the bagpuss narrator?-- Skewed, Jul 10 2014 I'm with Max: human intelligence is a phenomenon of the gaps in our knowledge. When we've mapped the complete behavior of a human brain there will be three camps: those who admit the human brain isn't special, those who claim against all evidence that we haven't actually learned everything about it, and those who deny the whole thing.-- Voice, Jul 11 2014 Plus of course the fourth camp, those who have been too busy or not paying attention & haven't heard the news.-- Skewed, Jul 11 2014 Is it too late to insert a random political joke?-- RayfordSteele, Jul 11 2014 It's never too late.-- MaxwellBuchanan, Jul 11 2014 If intelligence is related to how far you can travel in mind and body then the universe is a very, very big place.
I believe the current, dimension starved, hardware is not up to a real intelligence but the technology will still be a great tool for new 'thoughts'.-- wjt, Jul 13 2014 random, halfbakery