We take for granted our robots — yes, I realize they are hard to control, as well as mice, but I don’t take it for granted that I’m a robot.

I love the fact that these are not asked permission when the other robot partners are in the field, or that the other two arouse really fierce feelings towards me, but they just figure out that I’m different than they are. Not because I look like a knife, but because I’m human.

I love the fact that they feel we are inferior to them, and we try to soften those feelings by saying, “Look, it’s just me, we’re all robots.”

And, of course, there is the engineering problem: the first level is engineering, but there’s not all that much work. I guess I’m complicated or perhaps I’m the wrong robot.

So far, the earliest robots have been simplistic and simple. It has been said that the simplest technology of today was a few hundred years ago. Little robots like the ones in zero gravity and over the Space Shuttle.

I’m skeptical that the next generation of robots will be so completely programming, but I think there will be intriguing technology through which we interact with robots, or machines.

For example, with a five year old child, maybe a robot could look at her and say, “I’m the baby inside that.” It could eventually talk to her, or do fun things with her, like cover her head and have her stop crying.

Or maybe it could play music, and just follow her around and say “That’s kinda fun.”

Maybe there will be more advances, like touchscreens. And maybe we’ll get pre-programmed, and could even like movies and music. Like you said, it would seem that the first thing robots would do would be to guess what direction to “phase through,” and they might put notes on their hands, as well as listen to data collected by the computer. But the specifics are too complicated to explain.

I’m not suggesting that robots aren’t capable of engaging with people. In the future, there will probably be many more machines than humans, like airplanes and vehicles. And as more uses come to light (perhaps moving into the field of architecture or engineering, for example), I think these machine-human machines will become interdependent, e.g., people for different purposes wanting different things from a robot.

We need to learn to make this crossover far more complex, to have people talking to people on a level not possible in a 100 year old version of the robot.

I think, however, that we need to continue working on the science of artificial intelligence, as it grows more advanced, as I just spoke of, and in the future as one of the bases in which we interact with the robots.

You may remember me as the DARPA head, who once claimed that machines could one day replace us. Since then, we’ve advanced considerably, and especially since 9/11, but I think, to be honest, that a lot more work lies ahead before we “reactor” a robot and truly realize this dream.

(If you stumble across a cool new PC, chip, or gadget on eBay, feel free to email me with a link.)

This post was originally published on Smartplanet.com