Shropshire hills landscape ith yellow gorse foreground

Is my job safe?

History suggests new technologies usually create more work than they remove — but the mix of jobs changes.

On 29 April 1770 Captain Cook’s Endeavour sailed into Botany Bay, Australia. On the shoreline people were at work gathering shellfish for food. Never had these members of the longest continous culture on earth seen anything remotely like the Endeavour. An observer aboard recorded that their reaction was to go back to the task in hand and collect more shellfish.

The “too unworldly to comprehend” response can be seen around the world of business with regard to the power and potential of Generative AI. Human history gives the long view on the effects of other great inventions.

Historic inventions and their overall effect on work
Invention (date) Effect on work
Writing (cuneiform)
c. 3200 BCE
Positive — created new occupations (scribes, clerks, administrators) though reduced the need for purely oral record-keepers.
Paper and books
1st–2nd centuries CE
Positive — supported bureaucracy, education and record-keeping; job creation outweighed copying losses.
Double-entry bookkeeping
14th–15th centuries
Positive — enabled complex business management, generating clerical and financial professions.
Movable type printing (Gutenberg era)
15th century
Positive — revolutionised publishing, teaching and literacy industries, multiplying knowledge-related work.
Scientific method / statistical inference
17th–19th centuries
Positive — created systematic research professions, laboratories and data analysis work.
Steam power
18th century
Negative → Positive — massive early disruption of manual labour; later expansion of industrial and transport employment.
Semaphore → Telegraph
Late 18th – mid 19th centuries
Neutral — messenger and courier roles declined; telegraph operators and communications industries grew.
Jacquard weaving loom (punched cards)
Early 1800s
Negative → Positive — some weavers were displaced; however, skilled operators, pattern designers and textile volume expanded overall.
Railway time standardisation
1840s
Positive — enabled large-scale rail, logistics and tourism employment; few direct job losses.
Public postal systems & stamps
19th century
Positive — generated mass employment in postal, logistics and related administrative work.
Electric motor
Late 19th century
Positive — increased productivity and safety; enabled entirely new industries without major net job loss.
Haber–Bosch process (synthetic fertiliser)
1910s
Positive — greatly increased agricultural yields; reduced rural labour intensity but supported industrial growth and food industries.
General-purpose computing
1940s onwards
Neutral → Positive — eliminated repetitive clerical work but produced vast numbers of technical and service jobs.
Standards (ISO, screw threads, gauges, etc.)
19th–20th centuries
Positive — made manufacturing interoperable, sustaining skilled and semi-skilled labour globally.
Intermodal shipping container
1950s
Negative → Positive — eliminated many dockside jobs; long-term logistics, manufacturing and retail gains outweighed losses.
Spreadsheet (VisiCalc → Excel)
1979 onwards
Negative → Positive — reduced demand for manual bookkeeping; but analysts, planners and IT support roles expanded.

The common thread is that, heretofore, inventions have been much more likely to create human work opportunities than to take them away.

Whether AI will follow this pattern depends on how the next few years pan out. Essentially: if AI can now perform x, y and z tasks, what tasks can humans do better than AI and what can humans do, now they have AI, that they couldn’t do before?