Ai has always been seen as a double-edged blade, with its many benefits yet equally dangerous ability to displace humans in the workforce. The latter is a worrying fact that is further articulated by Asia School Of Business CEO Sanjay Sarma during the AI@Work Leadership Summit 2023. He quoted Goldman Sach, saying that over 300 million jobs will be either lost or degraded by AI.
Sanjay added that generative AI will likely produce 7% of global GDP, moving forward. As AI becomes more and more advanced, it is likely to increasingly take over routine jobs.
However, AI will empower knowledge workers, instead of replacing them, as demonstrated by several case study by Massachusetts Institute Of Technology Director (Computer Science and Artificial Intelligence Laboratory) Daniela Rus.
According to her, AI had been used to review lymph node cells to diagnose cancer. On its own, the machine has an error rate of 7.5% which is worse compared to the 3.5% rate by a human pathologist.
“However, when both AI and humans are combined, the error rate went down to 0.5%,” she said.
An overworked doctor might not be able to review every new clinical trial, but working in tandem with AI, he or she can gain access to the world’s knowledge in an area that is most relevant to the patient.
With AI synthesising information, the doctor will not be required to read through an entire library of journals. Inversely, a recent article in the Europe Times talked about a prominent lawyer in New York who used ChatGPT to identify cases related to what he was arguing for.
The system generated the right structure and docked number for the cases. But upon closer inspection, the docked numbers correspond to either different cases or they were all made up.
“We have a powerful tool, but the powerful has to be checked. You should use it as an assistant to generate information for you, but this information is not always correct. You the people need to ultimately make the decision,” she said.
Departing from this ill-usage, AI holds immense opportunities for businesses. Daniela said AI can be used to forecast demand, optimise price and inventory. Analyst reports and CAD models can be established, and intricate, highly sophisticated coding for websites can be written in moments.
However, this benefit does not come without its problems. As more data is absorbed by the AI system, privacy will be increasingly compromised, take for example, the deep fakes. Authoritarian governments will use AI tools to stifle dissent and the inequality of wealth will grow now that jobs are being displaced.
The message of hope here is that AI is a technology whose advancement is something that humanity has already foreseen, unlike the Covid-19 pandemic. Hence, the human race should have the necessary time to react to this substantial change that is certain to occur in the future.
