Robotic Surgery

Would you let a robot without being controlled by a human, perform surgery on you?  I find this question fascinating not because of the hypothetical scenario it creates but because of the reactions people have.  Having said that, it's not strictly hypothetical as there are already machines that perform surgery on humans, they have been around for a while, the most recent that I know of was the Versius System developed by the National Health Service in the UK, although like most driverless cars and other autonomous systems there is still a human present to intervene if anything goes wrong that the system can't respond to.

When it comes to the question of whether you would let a robot operate on you, most of the concerns come down to a few basic questions.  The first, is which do you think would be more likely to make a mistake?  You can make an argument for either side here, you can argue human error is sporadic, and you can argue that machines don't make mistakes, they only do what they were programmed to do, and the only mistake can be in the programming, again a human error.  When you discuss mistakes however there is a follow-up question that arises, which would be more adaptive and able to respond if it made a mistake?  Again there is an argument that can be made for either side here.  On the side of a human you have adaptive intelligence which can learn in the moment, something which machine learning has not fully developed.  On the side of the machine you can argue that a machine intelligence can compute each possible action it can take at a rate far quicker than any human could, so provided it had an algorithm that was able to respond, it could do so much quicker than a human.

There is also the issue of pre-empting mistakes.  A human is limited in their ability to multi-task, to be able to both operate and consider what might happen next is difficult.  Whilst humans are vigilant for signs of error whilst working, machines can split their focus onto many different tasks without affecting performance in each task.  A machine could therefore predict things more accurately than a human could.

When it comes to machine intelligence, in the classical sense a robot could only do what it was programmed to do so will inherently be more restricted, it is only in the more contemporary sense that a machine intelligence is able to extend its own programming and do more than it was originally created to do.

When it comes to focus and the ability to maintain performance, machines also have the upper-hand here too.  Machines don't get tired, and as long as they are kept in working order their performance should remain optimal.  Humans on the other hand can get tired, and most human errors can be attributed to tiredness and misjudgements.

The question as a whole highlights the divide between those who trust machines and those who do not.  For those that do, there is no element of fear or emotion involved.  For those that do not, it is fear or emotion that override logic and reason to conclude that machines are the riskier choice.  If you want evidence of this claim you can look at the statistics of the number of people who died as a result of vehicle accidents in the USA.  In 2016, there were 37,461 deaths on US roads as a result of vehicle accidents.  These were all human controlled vehicles.   In the same year there were 2 fatalities as a result of driverless vehicles.  In 2017 there were 40,200 deaths as a result of human controlled vehicles, and there were none as a result of driverless vehicles.  Compare the coverage of human controlled vehicles in the news and in the media, with that of driverless cars.  Compare the calls for all driverless cars to be banned outright because of 2 fatalities, as opposed to no-one calling whatsoever for all human drivers to be banned from driving despite the combined 77,661 deaths on US roads in the same period.

Fear of change because it is different is irrational.  If every car on the road was driverless there would be substantially fewer fatalities as a result of vehicle accidents.  As robot surgery becomes more widespread the same conclusions are likely to be drawn as to its efficacy, and inevitably, there will be the same emotional response from those who are reluctant to embrace change.