AI’s economic impact will be similar to electricity

Generative artificial intelligence will be a transformational form of technology which will have the same global impact that electricity or the invention of the steam engine did in the past, according to Gen Re’s chief technology officer Frank Schmid.

Schmid, an economics PhD, said generative AI would be a “general purpose technology” which would have a dramatic effect on efficiency, including in the insurance industry.

He said the general benefit of AI was its ability handle multimedia, for AI robots to handle the unexpected and to generalise new tasks not seen during training, although this was not flawless and could lead to hallucinations. 

But he said the productivity benefits would take some time to be tangible, although the cycle was speeding up with each new innovation. Where electricity took 25 years to spark a boom, the personal computer took 15, he said.  

Schmid said general purpose technologies typically complemented innovation in production processes, organisational design and products. It was also capable of ongoing technological improvement and enabled innovation in application sector. 

He said they spawned “a feedback cycle” which played out over a long period of time. 

He said the velocity of the feedback cycle with AI was expected to be higher than previous technical advances. 

He said past experiences showed that those companies with the highest cost of transition were the slowest to make the change were a warning because it meant those companies lost out on the early benefits of participation in the feedback cycle of continued improvement and risked being left behind. 

“Technology choices that allow the insurer to benefit from a high velocity of complementary innovation emphasise the importance of learning and reversibility,” he said.  

Schmid said the benefits from AU should be looked at as an “augmentation” rather than "replacement" of labour.  

Simply replacing human skills with automation saved money but did not lead to technical growth, meaning there was no greater economic benefit. 

He said the tech decisions should be made that are “highly reversible”, meaning managers would need to keep changing and adjusting workflows. 

The arrival of a GPT also meant that organisations needed to make complementary investments and in the early stages of the adoption process, they needed to build intangible assets which would serve as inputs into the production process which would be difficult to measure in the early stages. 

Schmid said companies should initially start with small in-house teams to manage the transition and to facilitate transfers from external  sources. 

Did you get value from this story? Sign up to our free daily newsletters and get stories like this sent straight to your inbox.