I saw a CNBC news video that hundreds of robots dancing. That reminded me of the mass Tai chi demonstration in 2008 Beijing Olympics. (In fact, the robots are made by a Chinese company, aiming at the Guinness world records for the number of dancing robots.)
Dancing robots https://www.facebook.com/cnbc/videos/1848834661844690/
They say Artificial Intelligence (AI) will take people's jobs in near future, and to some extent, it will be true.
Even my scientist job may not be safe. A couple of months ago or so, in a prestigious magazine "Science", there was a cover article titled "The AI revolution in science". In an associated article, they were analyzing scientific process, breaking down the science research process into five major steps, and discussing if/how each of the process may be performed by AI.
Link to http://www.sciencemag.org/news/2017/07/ai-revolution-science
Adding some hope to our side, there is another prediction.
"If you find yourself on the side of programming (or "teaching") AI, your job would be safe" (for a time being).
Such teaching of AI involves step-by-step breakdown of what humans do, and enabling robots to imitate each step.
Humans are not made perfect for everything. I'd predict fishes will be better swimmers than humans, so will be fish-shaped robots. But if we limit the ideas to humanoid robots, standard instruction to humans should have usefulness.
The news footage made me think if you can teach robots to dance Argentine Tango (or other partner dances). It's a challenge.
It is a challenge involved in engineering. It may be difficult, but it may not be impossible.
This is a thread appeared in my Facebook with a friend JL when I posted the robot video.
****************************
HY: Next challenge: a partner dance. (dare to dance Tango)
JL: NEVER!!!
HY: In 2036 first Tango robots are created. In 2037, by mass-production, a large milonga is demonstrated. In 2039 four variations are added to the production line, indicating "individuality" of the couples,......engineering wins
JL: No connection however!
HY: They can program basic patterns like walking and ochos. Axis-based auto-balancing may be done with gyro and weight shift sensors. Musical interpretation could be a challenge. Another challenge, "connection", may be translated to a responsive feedback circuit attached to pressure/balance sensors. For emotional connection, program them to kiss after a dance (easy fix). 😉
JL: Ain't buying it
*******************************
The chat was actually thought provoking.
Joking on emotional connection aside, how do you teach musical interpretation to robots? How do you define connection? (which is poorly-defined and arbitrarily-used word in Tango world, by the way).
With motion capture technology and shape molding, the robots may even look like champion dancers in appearance and in motion. Making 2-dimensional "fantasy sports" to 3-dimension robots.
Such technologies may be used for refining interpersonal (or robot-human) interactions. They would be immediately useful in medical or nursing industry, for example.
On the downside, of course someone may want to use the technology to create killer Karate master robot or some other robots with "refined inter-robot-human technology" for military and police. For civilian use, UFC guys may rejoice for having great training partner.
It's easy to imagine. And if it is easy to imagine, it can come true easier.
I don't work at MIT nor at the Chinese robot company, so I don't fancy myself building Tango-dancing robots any time soon. But diagnostic and breaking-down-the-steps skills for dancing should certainly help to teach robot's AI. Any robot maker interested in hiring me for helping to build Tango-dancing robots ?(with a good pay, of course). Tango dancing robots may impress higher-ups in Pentagon with the underlying technologies. Haha.