AI and the world of work

Artificial Intelligence is everywhere these days, it seems. The headlines seem to veer between two poles – that it will destroy everything or save us all – whereas the truth is likely to lie somewhere in the middle.

It has, of course, profound implications for the world of work – not just for the jobs we do and how we do them, but also for job search.  Already many are using Chat GPT at work and when looking for work. A poll on our sister site, workingmums.co.uk, found 53% of respondents say they use ChatGPT in their role, compared to 27% who say they don’t, 11% say they don’t know what it is, suggesting a clear divide is opening up. 

What’s more, 58% of respondents say they need more help to use it more effectively, compared to 32% who say they don’t. The results show that employers may be missing a trick if they don’t develop AI policies and support to enable employees to get the most out of generative AI and to avoid the pitfalls of overreliance on it.

Meanwhile, when it comes to job search, the poll found 49% – use ChatGPT for job search purposes, such as CV and cover letter writing, with 33% saying it has helped them with such tasks. 42% say they are not sure if it has helped and 25% feel it hasn’t helped.

Bias

At a panel discussion of experts at the Hay Festival at the weekend, Stuart Russell, director of the Centre for Human Compatible Artificial Intelligence at the University of California, Berkeley, said even the experts don’t know how it works and that there is no guarantee scaling it up will improve it. FT AI editor Madhumita Murgia said that when it goes wrong it is hard to find out or pinpoint why. Yet errors and biases are piling up, she said. She also pointed out the biases it can accentuate in the recruitment process, depending on the information that it is programmed to draw from.

Professor David Runciman from the University of Cambridge said AI has much in common with the way states and corporations operate. They had superhuman power and no conscience or soul, he said. States and corporations can either be humanised more or may become more aligned to AI. AI has similarly dual possibilities in relation to politics – it could open up politics more and make it more transparent or it could be used to bolster autocrats. 

Surveillance state

Oxford associate professor of ethics and philosophy Carissa Veliz highlighted the dangers of “sleepwalking into a surveillance state” through the amount of data we are giving to AI and our inability to delete it if it were to fall into the wrong hands. She stated: “Code is authoritarian. There is no negotiation with code. Giving data is a very political action.” She added that what it is good at is highlighting how magical the analogue world is and making us realise how rich and fragile the analogue world is so we can cherish and preserve it.” Russell also highlighted potential positives in terms of healthcare access and education, particularly in the Global South.

Professor Runciman said we will probably continue to sleepwalk into the future unless a global crisis happens that prompts more regulatory action around the world.

The next version of Chat GPT – GPT-5 – is due out later this year. With increased computational capacity, it is described as being able to handle large-scale tasks, complex analyses and massive data volumes, enabling cutting-edge AI applications in sectors such as finance, healthcare and technology.

• Vote  
• 0
• 0
• You already voted!