AI + ML
Labor groups say new technology could widen inequality if not careful
AI-pocalypse More than half of British people are concerned about the impact of AI on their jobs, according to employers’ unions, which are calling on the UK government to adopt a “worker-first” strategy, rather than simply allowing companies to abandon workers for algorithms.
The Trades Union Congress (TUC), the trade union federation in England and Wales, said it found that people are concerned about how companies are implementing AI and want a say in how the technology is used in the workplace and across the economy.
It warns that without such a “worker-first plan”, the use of “intelligent” algorithms could further widen social inequality in the country and cause the kind of social unrest that comes with it.
The TUC has said it wants to put conditions on tens of billions of dollars of public funding spent on AI research and development to ensure workers are supported and reskilled, rather than de-skilled or replaced.
It also calls for governments to put in place guardrails to protect workers from “AI harm” in the workplace, rules to ensure workers are involved in decisions about how machine learning is used, and support for people who euphemistically “experience job changes” as a result of AI disruption.
These are set out in the report “Building a pro-worker AI innovation strategy” published by the TUC. [PDF].
Perhaps unions are alarmed by reports, such as those in the Financial Times, that British companies see AI simply as a way to reduce investment in their employees.
Or perhaps recent research shows that machine learning models appear to be automating some entry-level roles and taking jobs away from younger workers.
Perhaps it’s a move like the one the UK’s Department of Science, Innovation and Technology (DSIT) signed earlier this summer with a memorandum of understanding (MoU) with Silicon Valley’s OpenAI to leverage its technology to revitalize the UK public sector.
This is despite the fact that a prominent AI professor at University College London said: register It says, “OpenAI is a very unstable company that could collapse at any time. They’re just working hard to increase their immediate value.”
Sam Altman, head of OpenAI, also acknowledged last week that the world is currently in the middle of an AI bubble, although he did not predict how long it would take for the bubble to burst.
The aforementioned “AI harms” include “improper use of algorithmic management tools to monitor, assign tasks, reward, and sanction employees,” which would “adversely impact employee health, productivity, and labor relations.”
Of course, such tools don’t necessarily have to be “intelligent” algorithms to make the workplace miserable. In fact, if it’s AI-driven, employees may need to monitor automated tools to ensure they’re working properly, rather than the other way around.
However, as the report notes, predictions about how AI will impact workers vary widely. Some predict mass layoffs, while others predict only gradual shifts across the industry. Some hope that redefining the economy will increase productivity and create new jobs, while others see only modest gains.
The TUC argues that rapid technological advances can only bring about widespread social progress if working people are empowered, and believes this requires a whole-of-government approach in partnership with workers and trade unions.
“AI has transformative potential and, if developed properly, workers can benefit from the productivity gains this technology brings,” said Kate Bell, TUC Assistant General Secretary.
That means public money comes with strings attached and cannot be siphoned off into the pockets of billionaire technology executives, but the alternatives are bleak, he added. If left unmanaged, the machine learning revolution could destroy jobs and entrench existing inequalities as AI owners become increasingly wealthy.
Meanwhile, a study by the UK Government Digital Service (GDS) found that giving civil servants access to Microsoft 365 Copilot AI can actually save an average of 26 minutes a day on administrative tasks. No doubt the software giants are happy about this.
Again, a recent report from MIT in the US found that despite spending $35 billion to $40 billion on generative AI initiatives, 95% of organizations see little or no return from their AI projects. This is something boards must think through as they consider their own developments. ®
