January 12, 2017
In a move that may just be a harbinger of things to come, a huge Japanese insurance firm is trying out replacing humans with artificial intelligence (AI) systems. Not that this is particularly earth-shattering. AI robots have been replacing workers for years. But this is a bit more next-generation because it utilizes not only AI, but integrates the ability to work with large volumes of data to make “logical” decisions about payments. And industry analysts say the same could start happening in the U.S. this year
According to the sources, Fukoku Mutual Life Insurance of Japan plans to replace a group of workers that do due-diligence in setting up payouts.
Powered by IBM’s Watson, the system will take over handling insurance payouts by culling hospital records, patient medical histories and injury data. Why? It is pretty obvious. Considering the human issues of time off, productivity cycles, motivation, environmental surroundings and the ton of other nuances that affects human workers, they will save a ton of money. The stats go something like this – productivity is expected to increase by 30 percent and produce a savings of $1.65 million on employee salaries. The cost to install the system is $2.36 million. The annual costs to maintain are around $177,000. Doing the math indicates that the AI system should pay for itself in less than two years.
Going forward, companies are widely expected to use this type of technology to sidestep human workers and get work done without needing to pay salaries and provide health care or vacation time. But it isn’t going to be quite that simple.
Aside from all of the same issues that have been part of automation, this is a bit different. Replacing monotonous, repetitive tasks with robots is one thing, replacing thinking humans is quite another. These are somewhat uncharted waters.
AI is becoming more and more capable of accomplishing complex human tasks because such systems are making steady advancements that enable them to increasingly learn on their own. This means that not only can they do the tasks, but make more “human-type” decisions and understand human behavior – the awakening of eventual self-awareness.
But, let’s not get too far ahead of ourselves. It’s not like all of a sudden we are going to see widespread implementation of AI systems on a global scale. The early process will involve acclimating workers to work alongside of such systems. The challenge will be getting workers to see AI as complementary, not adversarial. And that alone may take a decade or more and may be easier said than done, especially with all the sci-fi about robots, droids, and cybernetic organisms we have been exposed to for years.
AI is going to happen and not just with systems. And it will port to human-like form factors eventually. But we are decades away from the “Datas” of Star Trek” and “Bishops” of Alien movie fame. But we all know that eventually, science fiction becomes science fact. After all the first computer was only a 2-bit device.