Credit: iStockphoto
Eventually, we will create truly artificial intelligences, with cognition and consciousness recognisably similar to our own. I have no idea how, exactly, this creation will come about. I also don't know when it will happen, although I strongly suspect it won't happen before 2030, the year that some singularitarians predict.
But I expect the AGIs of the future – embodied, for example, as robots that will roam our homes and workplaces – to emerge gradually and symbiotically with our society. At the same time, we humans will transform ourselves. We will incorporate a wide range of advanced sensory devices and prosthetics to enhance our bodies. As our machines become more like us, we will become more like them. And I'm an optimist. I believe we will all get along.
Like many AI researchers, I've always dreamed of building the ultimate intelligence. As a longtime fan of Star Trek, I have wanted to build Commander Data, a fully autonomous robot that we could work with as equals.
Over the past 50 years, the field of artificial intelligence has made tremendous progress. Today you can find AI-based capabilities in things as varied as Internet search engines, voice-recognition software, adaptive fuel-injection modules and stock-trading applications. But you can't engage in an interesting heart-to-power-source talk with any of them.
We have many very hard problems to solve before we can build anything that might qualify as an AGI. Many problems have become easier as computer power has reliably increased on its exponential and seemingly inexorable merry way. But we also need fundamental breakthroughs, which don't follow a schedule.
To appreciate the challenges ahead of us, first consider four basic capabilities that any true AGI would have to possess. These are second nature to a human child but have not yet been cracked by computer intelligence:
• A two-year-old child can observe a variety of objects of some type – different kinds of shoes, say – and successfully categorise them as shoes, even if he or she has never seen soccer cleats, wooden clogs or suede oxfords. Today's most advanced computer vision systems still make mistakes – both false positives and false negatives – that no toddler ever makes.
• By age four, children can engage in a dialogue using complete clauses and can handle irregularities, idiomatic expressions, a vast array of accents, noisy environments, incomplete utterances and interjections. They can even correct non-native speakers, inferring what was really meant in an ungrammatical utterance and reformatting it. Most of these capabilities are still hard or impossible for computers.
• At six years old, children can grasp objects they have not seen before; manipulate flexible objects in tasks like tying shoelaces; pick up flat, thin objects like playing cards or pieces of paper from a tabletop; and manipulate unknown objects in their pockets or in a bag into which they can't see. Today's robots can at most do any one of these things for some very particular object.
• By the age of eight, a child can understand the difference between what he or she knows about a situation and what another person could have observed and could therefore know about it. The child has what is called a "theory of the mind" of the other person. For example, suppose a child sees her mother placing a chocolate bar inside a drawer. The mother walks away, and the child's brother comes and takes the chocolate. The child knows that in her mother's mind the chocolate is still in the drawer. This ability requires a level of perception across many domains that no AI system has at the moment.
The singularity.
While I agree with Robin Brooks starting point,that we are essentially machines we have escaped being predictable machines via the unpredictable emergent properties of our highly complex brains.
It is from this source that conscious thought and personality arise.
Thus that while the deterministic model is self evidently true, the enhancing effects of emergent properties transmutes us into an indeterminable condition of open ended possibility.
The proposition that humans are predetermined is a logical brain twister in the same class as that old chestnut
"I always tell lies"
So if we can do this with a machine then by definition the result will be unpredictable.
Question, please Sir is this a GOOD idea?
Answer, maybe not
Q When?
A Not soon
Q will it like us/ be like us?
A which do you prefer?
Q are you?
A guess
Ron Horgan.
Be careful what we instill
Rodneys comments on what we instil into AI raise the key issue on AI - you instill what you are. That is the problem. we will instill human nature its strengths and frailties. Someone, somewhere will instill the capacity for war in their intention to protect themselves from other humans. Then we let the "children" protect us. Well we know how that goes down.
ok
It all sounds very nice, creating ourselves in our own image stepping back and patting each other on the shoulders. Just like the boys down at the Manhattan project did 50 years ago. But let me get this straight mankind currently is drowning in our own apathy, watching species after species disappear, "talking" about global warming and twiddling our thumbs, not really giving a shit about the world now nor since the dawn of the first singularity, when we started clubbing animals with bones. And you expect AI to clean up our mistakes live in some sort of quasi-utopia with Nano bots that clean up our toxic waste, plug ourselves into the matrix, blah blah blah. Get over this ridicules god complex that all these futurists cling too. It’s now, not maybe, grow up and take stock of the current world and its problems not some AI 40 years from now. i mean Jesus thinking and talking about this stuff is cool, but actually giving funding to these pipe dreams only divides the reasonable from the absurd. The best part of life is that we die, who wants to be uploaded into a computer, life isn’t a cyberpunk game there are certain responsibilities humans have, and one of them is to die, not continually consume for eternity. Get a grip.
Super-human intelligence
I believe there is a fundamental blocking to what we can know about
a super-human intelligence. Simply, we can't know what such an entity will
think or do, because per definition we can't think like superhumans.
So no matter what precautions we take, we can't predict the actions of these creatures. We can't even say "low probability of killing us" or something like that.
Exceeding our design
It is easy to believe that the creation of AGI will happen soon, or happen at all for that matter. From reading countless science fiction books, it would be a fantastic concept to visualise, humans and robots co-existing peacefully.
I would like to point out however, that it is a rather selfish act to design AGI in our own image. Given that we would be recreating our own mistakes, in efforts to conceal or solve our current problems. We would be re-developing the laws that govern us, and programming these traits according to our functions and needs. There will still remain many characteristics and faults at large which we wouldn't be able to predict.
I'm offering that human control does not offer greater possibilities.
I gather that this production would be a slow process to the 'ideal' final design. But what this fantasy result (of an AGI) consists of exceeds our human abilities to fathom. So it is only realistic to be able to replicate ourselves in the form of synthetic matter, if we ever do get there. However unrealistic to think mankind (primitive as it still is in its development) can create something beyond comprehension, and beyond the human "ideal".
In light of this discussion I would love the idea of a self emergent AGI from the depths of a network such as the world wide web.
( This reminds me of the same idea sparked in a novel by Orson Scott Card, called "speaker for the dead". Embedded in this text, an immaterial super intelligence emerges as a "network of ansibles" capable of awareness, sentience and a million performances at light speed.)
-19, female, aus
Would this cause the intelligent design camp to win?
By recursive logic?