One of the world’s leading experts on artificial intelligence has advised companies to use the technology very carefully and always in conjunction with human control.
Erik Brynjolfsson, professor and senior fellow at the Stanford Institute for Human-Centered AI, told an audience at the Davos World Economic Forum on 20 January, that artificial intelligence represented a burst forward of technical capability “that we’re not keeping up with in the way we work”.
We must address to what extent do we keep humans in the loop rather than focus on driving down wages” – Erik Brynjolfsson, Stanford Institute
He reminded the audience at the AI and White Collar Jobs session that when electricity infrastructure became available, 100 years ago or so, it took at least 30 years to fully realise its productivity benefits because to use it properly required the redesign of the office and the factory. “We’re in a similar period now with AI,” he said.
“What AI is doing is affecting job quality and how we do the work,” said Brynjolfsson. “Keeping humans in the loop is the main focus. We must address to what extent do we keep humans in the loop rather than focus on driving down wages.”
This was a primary task for CEOs and HR leaders, he added.
Brynjolfsson was adamant that AI should not be left to its own devices when used in creative and decision making. “I would not advise turning it on and walking away. These systems have too many flaws. They don’t understand truth very well and kind of hallucinate facts. Right now it’d be downright dangerous to use AI tools without a human in the loop.”
AI in action
Employee monitoring software spots ‘time theft’
Minister signals light touch approach to AI in the UK
Why algorithms at work aren’t all bad
Hybrid working legal and practical issues: Your common questions answered
He cited the example of Cresta, which supplies AI technologies to call centres. He said: “Cresta keeps the human at the forefront with AI behind giving prompts. When the human is at the front firms did better in terms of productivity and closed the wage and skill gap.”
Crucially, he said, research revealed that “the workers who benefited most were the less skilled and lower paid – their productivity increased.”
He added that the main focus of study should be around where the relative strengths of AI and humans are. Where it was used as a complement to humans, AI drove up wages and productivity, he said, but where AI tools replaced humans, wages went down.
He joined members of the panel, including: Martin Sorrell, chairman of S4S Capital; Mihir Shukla, founder of Automation Anywhere, and Lauren Woodman, CEO of non profit DataKind, in advising businesses to not be afraid of high functioning tools such as ChatGPT. “Work with it, embrace it.” he said. “One of our natural strengths is flexibility. Humans and machines can solve problems in business and environment together.”
He said that at Stanford academics created a “skills adjacencies map” by using AI to evaluate people’s skills on CVs. He said that this map had shown how machine learning could unveil how skills interrelate and connect, revealing how many more skills people possessed than the ones their employers were using.
When the human is at the front firms did better in terms of productivity and closed the wage and skill gap” – Erik Brynjolfsson
Woodman said she was concerned there was too much commercial pressure to jump to AI quickly, and with Shukla, reflected that more than a third of the world’s population had never even used the internet. Thus the advent of AI could exacerbate inequalities.
However, Shukla said research in poorer areas of the world including parts of India, China, and the US, had shown that people could go from “flipping burgers to qualifying for $150k jobs within three months’ training” working with AI and new tech. “Reskilling is a huge part of this,” he said.
Sorrell said people were becoming more fearful of AI as the technology spread. He added: “HR will be very taxed by this.”
Journalist Camilla Cavendish asked from the floor how businesses could prevent workers feeling demoralised by the effectiveness of AI. Brynjolfsson admitted this was a “pernicious problem” and was illustrated by Google’s driverless cars. “First they had to have a safety driver to watch the system… but they got bored so then they added a second safety driver to watch the first driver.” Similarly, if we weren’t careful, workers would be demoralised if they were forced into secondary roles by AI.
Nonetheless, Brynjolfsson concluded, there was a “Huge opportunity for HR to unleash assets… imagine how many people are not in the right job; the real value of AI is to give people the chance to live up to their potential.”
Latest HR job opportunities on Personnel Today
Browse more human resources jobs
