Professional investors are convinced increased computational power will lead to a breakthrough in artificial intelligence ( AI ) capabilities this year, according to a recent study.
Almost all ( 96% ) of senior executives at insurance asset managers, pension funds, family offices and wealth managers believe that the expected 10 times increase in computational power being applied by the big five US model developers to current AI models will deliver a breakthrough in what AI can achieve, finds fund manager Robocap’s study, which surveyed institutional investors and wealth managers managing assets of US$513 billion.
That will translate into increased global growth this year, the study notes, with on average professional investors surveyed believe productivity gains from AI this year will account for 22% of global growth.
In three years’ time, the percentage of productivity gains and global growth attributable to AI will increase to an average 30%. Around one in five questioned believe the increase could be 40% or more.
The impact of increased AI capabilities will be felt first by marketing and sales, with 31% predicting the sector will be the first in line for disruption, followed by services, such as law, finance, accounting and tax, which was ranked first by 29%.
One in five ( 20% ) of surveyed firms based in the UK, US, UAE, Saudi Arabia, Singapore, Hong Kong, Germany and Switzerland highlight IT as first in line to feel AI’s impact, while 9% pointed to transportation, 7% to healthcare and 6% to industrials and consumer goods.
Real estate and utilities are seen as the sectors likely to be disrupted last by AI and robotics. Real estate, around 28% say, will be among the last to be disrupted, followed by 20% choosing utilities and 11% selecting materials.
“The advance in AI capabilities over the past few years has been phenomenal as major model firms have increased the computational power used to train their models year on year,” adds Jonathan Cohen, Robocap’s founder and CIO. “This year, it is estimated that the computational power being applied by the big five American model developers to train their next model will actually be up to 30 times more, relative to the compute power current model, and investors are convinced that the world is on the verge of a breakthrough in terms of AI capabilities considering weekly progresses.
“That is translating into real-world effects with the productivity gains from AI making major contributions to global growth, which are expected to increase over the coming years.”
