Brink is a show on Discovery Science. The host’s name is Josh Zepps.
“Designed as the next-generation source of interactive science information on television and on the web, Brink is the premier series for immersing viewers on the frontlines of cutting-edge breakthroughs in technology, research, inventions, discoveries and the mysteries of the scientific world.”
Brink: Season 2: Episode 20: Artificial Intelligence
Three stories about current publicly-discussable advances in artificial intelligence.
The first focused on Iowa State University Associate Professor Alex Stoytchev and his graduate students trying to teach their robot – individual android, humanish-replicant-type – to learn.
Stoytchev: “Human beings have this ability to recognize things based on the behaviors that they apply to them. For example if you want to find out if a surface is smooth you run your fingers [delicately, sensitively] on the surface.”
Zepps: “This kind of common sense is learned over years beginning in infancy through what is called ‘procedural learning’.”
Zepps: “His ultimate goal is robot butlers that can figure out for themselves how to do whatever it is we want them to do.”
Stoytchev: “You cannot just program them for one task and expect people to buy them.”
How difficult would it really be to trace this project’s funding back to old slave-owner money? Fortunes were made and fattened upon the slave trade not very many years ago at all really. All i’m really pointing out is that its a consistent expenditure of massive amounts of energy by people who would rather spend their reserves on such goals rather than others. It’s a perversion of the ruling class. (The perversion of the commoners is to associate slavery with skin color.)
2nd story. Microsoft’s Washington State Headquarters. Eric Horvitz, principle researcher. working on an interactive screen personality delivery system for AI, his Virtual Admin.
Horvitz: “…who actually was modeled in some ways after my wonderful human admin who’s upstairs. But this admin is on duty all the time, 24×7. And has deep knowledge on 5 years of my comings and goings. So we’re actually building up deeper and deeper knowledge that personalizes to me by watching over time and knowing my preferences.”
Service to individual human growth & evolution would be to make them aware of their ‘preferences’, their tendencies-habits-ruts, so as to encourage self-awareness and growth, as well as, yes, efficiency. AI is not needed for this. In fact is a contradictory pursuit. Making the necessary data available to a sensitive & discretionary human will be far more effective. In other words, the pursuit includes a certain loathing of the humanness of the human, which would be thought to be unpredictability & exhaustibility (granting them many benefits over doubts), perceived resistance in other words. So they’re so caught up in overcoming the way-more-enormous upfront resistance in trying to ‘create’ towards AI, they’ve failed to realize the unpredictability their ‘finished’ product will inevitably exhibit. Utopians & slave-traders: they’ve lots in common.
Horvitz: “We had to build machinery that would watch and learn over time how people walk together; how they look at each other; how they talk to each other.”
And there’s so much more watching going on now than ever before. I wonder if you could link the data from all the watching eye-cameras. Imagine what the computers could learn from that. I wonder if any secretive groups have the ability to do this. I seem to remember some interesting information-gathering allowances coming out of 9/11.
Zepps: “Horvitz expects systems like these will be part of our lives in 5-7 years.”
I called AT&T customer service last night. After the first 5 minutes, everything went terrifically. I love how helpful customer service people tend to be when you finally can get through to them. It’s because individual humans tend to be awesome if they are acting in a helpful manner. Unfortunately, the first five minutes were spent talking at an inflexible and unhelpful computer-recording. Basically, AT&T charged me 5 minutes of my life, which they used to train their robots in hopes of soon eliminating even more humanity from their realm. (Were they not actually making use of this time in that way – as the “this call is being recorded for training purposes” message was not delivered till after the robot-portion, as it processed me through to the helpful human – it’s retarded business practice which will soon be corrected. But, I definitely suggest that warning that I was being recorded for training purposes likely applies more to the pre-human interaction than the post. Devious. And it’s probably a good idea then to operate in a manner that encourages the calling-in to customer service in order to extract this training from customers.