I think that we'll just end up gluing neurons onto transistors for the near future. With the latest research they have rat brains flying flight sims.
The primary stumbling block is going to be the ethics committee or finding an animal that has a large brain that ethics committees aren't particularly fond of. Or networking the brains of rats together to form a much larger neural net that interfaces to ethernet. Think EC2 with a rat brain on each server. Or conversely a network interface for the human brain.
I do think it's possible to think with a UTM but I don't think that discovering that ability is going to be economical compared to fusing transistors to neurons.
I tend to hold with Anne Foerst, who was the theologian working on the ethical side of the MIT robotics project back in the 90's. http://www.cs.sbu.edu/afoerst/
If it is relatable and seems to have human-like consciousness, we have to treat it like any other human, because consciousness is an empirically meaningless term.
So, how at what ratio of biology-to-technology does something seize being a thinking person, and start being a thinking computer?
Clearly, a few rat cells hooked up to big processors are a thinking machine. But someone with an electronic implant in their brain that prevents seizures is a thinking person. But at which point do we say, "you're a robot, beep boop"?
Oops! Well, you're right - my first language was Russian - but at this point English is my primary language. I'd attribute that mistake less to my heritage, but rather to all the wine I drank last night :P
I do think it's possible to think with a UTM but I don't think that discovering that ability is going to be economical compared to fusing transistors to neurons.