The New York Times | Interview with the robot Bina48
June 20, 2011
The New York Times | Bina48 was designed to be a “friend robot,” as she later told me in one of her rare (but invariably thrilling) moments of coherence. Per the request of Martine Rothblatt, the self-made millionaire who paid $125,000 for her last March, her personality and appearance are based on those of Bina Rothblatt, Martine’s living, breathing spouse. [...]
Related:
Wikipedia | Martine Rothblatt
Video Source: The New York Times
Related:
The New York Times | "Science: Making Friends with a Robot Named Bina48"
Comments (4)
by John N
What ever happened to Ramona? She should interview Bina48.
by melajara
I tend to agree. Chat bots are very delusional to the general public. However, I was really enthralled by Watson. So, actually there is some progress.
The biggest trouble with anthropomorphic robots is their total lack of genuine drive. I think unless we make strides to endow them of true proprioceptive abilities and a knack at (self) composing sensori-motor schemes to build up perceptual invariants and genuine cognitive maps, they only will be more elaborate fakes.
But then, we cannot truly emulate feelings, desires, pain, love and every emotion and those are fundamental components of human condition.
So all those “companion” robots will stay lame cheaters for quite a long time unless we force them to rely on some true wetware, chilling isn’t it?
Much more fruitful, interesting and actually possible is to strive for robots as workers and ultimately, scientific research assistants, IMHO.
by jesliphius
I agree with you melajara!
We can emulate needs, emotions and feelings via software. That isn’t the problem. The problem is that emulations don’t do what we need them to do. At some point our brightest minds confused emulation with actual consciousness. The Chinese room experiment illuminates this fact quite effectively. We need the real qualia of a need… the gripping sensation of fear…. not the emulation of ‘oh my goodness, I am really scared’ because some firing software neuron caused a flag to go off. In some settings it looks real but when we need actual understanding to be present our carefully crafted demos leave us wanting.
Even language acquisition itself is founded upon needs. We cannot create a native AGI that acquires language in a dynamic setting because we have not figured out a way to truly motivate the AI with non-emulated needs. It must fight for its survival for it to be real.
I would say we should give up on trying to create a human-level intelligence because we cannot create an artificial general intelligence that has real human needs. If you want a human-level intelligence, have a baby. Human intelligence depends on real qualia. The real “CORE” needs of an AGI would be battery life, core-temperature, RAM and long-term memory (etc). More advanced needs could emerge from intelligence layers as they do with humans, but we must get the core correct first.
If we admit to creating a machine-intelligence we could actually start off on the right foot. Machine-intelligence is different than human intelligence. A machine-intelligence would still be able to ‘emulate’ a human. It could eventually understand humans and even teach us about ourselves through observation. It would even ‘know’ it was not a human intelligence but it could ‘act’ like one.
To get here I believe you must create machine intelligence first; NOT the emulation of human intelligence.
by jesliphius
This is a very poor demonstration of a very poor attempt at AI. I’m not being negative just to be negative. I have worked in the industry. I walked away for personal reasons; and ever since I have been shocked at how far behind these so-called AIs are.
They have no context pools, incorrect memory structures, no ability to learn new concepts or resolve ambiguities on their own and they have been engineered (on the software side) by people who do not understand the human condition. The result is….. what you see in the video.
Hardly the AI of our future and a step in the WRONG direction, if you ask me.