Robotic surgery. And the future of medicine


This story begins with Laparoscopic surgery that has become so popular in our world. The surgeon goes in through keyholes instead of making huge incisions and is able to perform several surgeries without the trauma of regular surgery.
It's all fly by wire. He views the scene through a monitor attached to cameras on the ends of the equipment and uses robotic arms instead of his own.
Surgeons were not trained to work this way,  but that is another discussion for another time.
The point I am making is that the surgeon sees through a computer and his instructions relayed through computers.
The next logical extension would be that the patient and the doctor need not be in the same room, or city, or even the same country.
This, by the way, is already happening all over the world. A gifted surgeon in one country is able to perform incredibly complex procedures in any country that has a reliable communications infrastructure. All that is needed is a competent crew to set up the puncture incisions and the placement of the equipment in the patient.
The next step in this crazy road would be,  that the system records every move every surgeon makes and it ranks the quality and outcome of every stitch and incision, and logs them all in a vast database.
In the near future, a keyhole surgeon will be able to click on "stitch" button in his heads up display and choose stitches that were done by the masters.
The system would simply reproduce those, exactly as the best surgeons in the world would.
Given enough data the system would improve the efficiency of these actions to reach levels of perfection that exceed the ability of the human surgeons.
Initially, all the human would need to do would be to choose the correct incision or stitch, and the robotic system would simply carry out the instructions. But when it becomes apparent, as it will,  then the machines reaction time is significantly faster than the operators, and that the behaviour and performance of the machine, more reliable and dependable than the human, more decisions will be allowed to the machine.
The trajectory is that given enough time, the system would not require a controller, but somebody to oversee the operation. Theoretically, that remote surgeon would be able to oversee many operations, being carried out, all over the world. He would step in if or when something went wrong or advise the robot when a decision needed his approval.
All of this speculation and projection ignores the probability of deep Ai in our lifetime.
Everything described in this does not rely at all on deep Ai.
We are simply describing the evolution of existing surgical systems and the refining of existing software.
When deep general Ai does come to exist. 
(the skeptics would have you believe that this is not ever going to exist).
If this Ai reaches maturity, everything will change, again, all at once, and the overseer would not be required. 
No human in the loop.

While it all sounds far fetched.
The above video will maybe give you a preview of the levels of technology that exist, today.

For the purposes of discussion, we need to examine two other areas of rapid development. Remote controlled military drones, and self-drive vehicles. Both of these have been hurtling ahead at breakneck speed.
Some of the new drones resemble regular fighter jets and have very similar offensive characteristics.
The remote pilot in the earlier versions was expected to land and take off, very much as if he was in the cockpit. The only difference is that he was safely back home. The later versions were able to handle all the intricacies of takeoff, landing, and flight. All he had to do was send the instruction to land, or take off.
The drone was smart enough to carry out the instructions perfectly.
Recently this was illustrated in having a drone land and takeoff from the limited runway of an aircraft carrier ship over rough oceans.

This is a daunting task for a human pilot, and today there are drones that can do this better and more reliably than their human counterparts.
The talk about self-drive vehicles was the stuff of SciFi but with millions of kilometers of testing in populated cities all over the world, we have all come to realise that it's happening.
It simply is a question of when and in what form.
So the take away is that we have been able to do all of this without the advent of deep general Ai.
All of this is happening. We need to either get with the program or get out of the way.
Imagine what this landscape would look like, with Deep Ai.
M Parak 2017

Comments

Popular Posts