Skip to content

Yesterday was a beautiful sunny autumn day in Mornington.

My wife and I walked to our local café and ordered our coffee. Even though it was a working day the venue was packed with groups of friends, young mothers and couples enjoying their brew, both inside and outside the bustling café.

While we waited for our order I fed Rebel the free dog treats available to the café’s patrons.

The advent of relatively inexpensive and high quality homemaking coffee options has not halted the growth of cafes such as Dreamer whose considered human touch in all aspects of their operations keeps their customers returning.

Although approaching 500 years old, the humble café seems one of the businesses least vulnerable to disruption by generative AI.

The evidence is quickly piling up that it’s lower-skilled jobs, like café workers, that are less disruptable than higher-skilled jobs.

This is the opposite of the three largest previous mass changes in labour markets brought about by machines (the Industrial Revolution in the 18th and 19th centuries; the replacement of agricultural labour by machines in the early-to-mid 20th century and the computerisation of the office in the late 20th century) where lower-skilled labour bore almost the full impact of the resulting job losses.

The Institute for Public Policy Research (IPPR), last month released its latest research on how generative artificial intelligence could affect work in the UK.

The authors lead with; “The core of our findings is that the world of knowledge work will be transformed by generative AI and that we need to start preparing for this now.”

In a large-scale assessment of 22,000 tasks in the UK economy, the authors found about 11% of tasks are exposed to generative AI right now, and this could increase fivefold if AI systems became more deeply integrated into organisational processes.

I doubt these figures would be surprising to many however the speed with which change occurs is predicted to leave business leaders and public policy makers struggling to keep up.

The phase that we have been in since the advent of GPT4 in 2023, is described by the authors as the experimentation phase.


Source: Transformed by AI by Carsten Jung and Bhargav Srinivasa Desikan (IPPR, March 2024, page 6)

The authors consider the most likely impact of the next two phases of AI adoption:

Phase 1 (ie the implementation in organisations that will likely target ‘low hanging fruit’ use cases. These are the cases where generative AI programs are relatively easily plugged into existing IT processes, without many changes to workflows.

  • Back office jobs (such as personal assistants), entry-level jobs and part-time jobs will be most exposed in this first phase.
  • Women will be significantly more affected (as they are more likely to work in the most exposed occupations, such as secretarial and administrative occupations).
  • While the overall labour market impact in this phase could be limited, it could nonetheless be disruptive in these occupational groups. For example, in administrative occupations, about a third of jobs could be displaced.

Phase 2 (i.e. where generative AI becomes more deeply integrated with existing organisational processes

  • If organisations decide to integrate existing AI technology more deeply into their processes (which is not a given), almost five times more tasks – about 59% of tasks – are exposed.
  • This means a large number of jobs could ‘feel’ its impact, and it will also increasingly affect high paying jobs such as accountants, analysts, IT professionals and creatives.
  • Such integration into existing processes would, for instance, mean giving AI the ability to access proprietary data, providing inputs via apps or giving AI systems the ability to execute tasks
  • Whether this will materialise and who gains and who loses will depend on various policy and organisational factors. Crucially, it is likely that not all organisations will adopt the technology at similar rates, leading to inequalities.

The authors outline three scenarios that represent a continuum of how generative AI could impact labour markets. ‘Augmentation’ means the technology is used to boost worker productivity to produce more or better output, while ‘displacement’ means it is used to lay off workers with less boost to output.

  • Full augmentation: If generative AI was widely integrated across the economy, we estimate it could provide an economic boost of 13% of GDP
  • Full displacement: 8 million jobs could be lost with no GDP gains.
  • The middle: 4.4 million jobs disappear, but still with significant economic gains of about 6.4% of GDP

A peek into what’s ahead was the outcome of last year’s writers’ strike in the United States.

This was not your normal industrial action, motivated by the desire for better pay and conditions.

The swift advance of generative artificial intelligence was the primary driver for both writers and actors reacting to the possibility of unchecked AI dramatically undermining their respective roles.

The newly-agreed writers’ contract does not outlaw the use of AI tools in the writing process, but it sets up guardrails to make sure the new technology stays in the control of workers, rather than being used by their employers to replace them.

Under the new terms, studios “cannot use AI to write scripts or to edit scripts that have already been written by a writer”. The contract also prevents studios from treating AI-generated content as “source material”, like a novel or a stage play, that screenwriters could be assigned to adapt for a lower fee and less credit than a fully original script.

Simon Johnson, an economist at MIT who studies technological transformation, said he hoped the outcome “..will be a model for the rest of the economy,”

“AI is under control of the writers, not under control of the studios,” Johnson said. “It’s not to be used as an automation technology. It’s complementary to humans.”

Although this was a short term win for human labour the big difference, compared to most other highly paid professionals, is the bargaining strength of the writers’ union (Writers Guild of America).

Highly paid professionals in law, accounting, marketing, sales, IT and other similar fields don’t have labour unions to represent their respective interests against the potentially job-undermining actions of their employers when it comes to AI implementation.

As the IPPR authors conclude, “Given the speed and breadth of generative AI adoption that we show is possible across sectors, and given the long lags with which policies take effect, we think it is urgent that policymakers start preparing a policy response now.”

The most secure future might just be in owning a great local café.

Related blogs

Care robots in Japan can’t reduce human labour (so far)

The Daleks are coming and many high skill jobs are not safe

The AI end game for workers: greater focus on critical work


0 0 votes
Article Rating
Notify of

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Inline Feedbacks
View all comments
Would love your thoughts, please comment.x
Scroll To Top