This question is as relevant today as it is controversial. We have pondered about this as well. After all, it seems as if our CEO, Graham Gaston, is working to make his former role as a DOS (Distributed Optical Sensing) specialist redundant.
D/SRUPTION have published an interesting piece on exactly that issue. They discuss how radiology specialists may become obsolete because AI (artificial intelligence) will be able to perform medical analyses. Why then would anybody go through 12 years of education and training to become a radiologist in the future? Will this, in turn, mean that specialists will die out, and we will make ourselves entirely dependent on machines?
The article suggests that we use AI to enhance rather than replace human experts.
The idea is called Artificial Swarm Intelligence (or Swarm AI®).
If that term makes you think of bees and fish, you are not mistaken. Unlike those species, though, human intelligence has not evolved to work in a ‘hive mind’. This is where AI comes in. The article suggests that the power of AI-driven analysis combines the opinions of a group of experts in real time, effectively creating an ‘artificial expert’. This approach has been tested in different applications and shown to be more accurate than humans or machines on their own.
QDOS™, our artificial Graham if you like, is software we are developing to integrate downhole well data from several sources, draw AI-driven conclusions and present those in ready-made reports. For our AI to learn, it will require training by human experts. Ideally, it will receive input from a group of human experts to form a consensus.
Yes, eventually QDOS™ will enable non-experts to produce decision-ready reports on downhole events.
However, we see great value in the human specialist. What if there are events that the AI has never encountered and cannot relate to anything in its database? What if novel sensors are developed that provide new types of data about the well environment? Human experience and intuition will be needed to inform the constantly learning AI. Analysis accuracy and reporting efficiency will increase dramatically in the future.
We see AI and humans upgrade each other.