Transform and roll out
The arrival of generative pre-trained trans-formers (GPT) has unleashed a new generation of large language models (LLM). Amid this rollout, there is a lot of scope for significant economic disruptions. For the humans in this equation, this discussion of an uncertain future also creates an odd mixture of hope and anxiety.
A subsidised byte
For now, these frontier LLM models are being offered as loss leaders to encourage their adoption (and to train the next model). This is not a sustainable business model, with the next logical steps involving higher token costs, tighter token budgets and greater market segmentation across geopolitical, quality and pricing lines.
Rewiring the economy
As the AI sector matures over time, we may see profound changes to the economy. As with all disruptions, certain jobs, industry sectors and economic regions will likely be more affected than others. It is in these market segments that we should be looking for early signs of economic and labour market dislocations.
Our new AI overlords?
Four years into this virtual relationship, there are limited signs of labour market impact from the rise of the chatbots. In Australia at least, job numbers are holding up, hiring appetite is strong and wages growth – the price of human labour – remains relatively steady. Thus far, AI technology has not displaced labour demand.
Source: ABS, Job & Skills Australia, Karpathy, Stats NZ, MaxCap (June 2026)
Implications for real estate
All told, AI tech is just one part of a long series of structural changes – e-commerce and telecommuting –affecting on real estate demand. To date, the demand impacts are relatively small. Over time, we would likely see earlier signs of disruptions in specific cities (Canberra) and market segments (office).
