AI will only be truly gamechanging when it starts making companies bigger (not leaner)
Heat trend
Collecting trend data
The percentage is based on available heat signal, not comment count or independent people.
The true game-changing potential of AI will be realized when it drives company growth and revenue expansion, rather than just cost compression. While AI is currently reducing enterprise-level costs, this does not equate to productivity growth. The significant spending on AI does not guarantee a return on investment until it demonstrably leads to real revenue expansion and impacts GDP figures, shifting from a cost-saving tool to a genuine growth engine.
AI is clearly compressing costs at the enterprise level, but cost compression is not the same as productivity growth, and so far the two are being conflated quite a bit.
Margin expansion through headcount reduction is a one-time efficiency gain with a ceiling, while productivity growth means each remaining employee produces meaningfully more output. This second outcome is what compounds and ends up being reflected in the broader GDP.
Right now, aside from the infrastructure buildout itself (which is a capital expenditure cycle that could just as easily be based on overestimated demand) there is no broad macro signal yet that AI is expanding real output. What we have instead is enterprises using AI primarily as a workforce reduction tool, and notably, some of the companies most vocal about AI's transformative potential are among the most aggressive at cutting headcount.
If AI were genuinely driving productivity, you would expect output per employee to rise alongside investment, not layoffs to precede any output story. A shrinking denominator can only take earnings so far.
What would actually validate the AI investment case in my view is visible revenue growth at the client companies deploying AI (not just the AI suppliers). This would be evidence that AI is helping them sell more, build more, or serve more customers, not just do the same with fewer people. A handful of developers shipping apps nobody uses does not clear that bar.
Only when we see that shift, can we safely claim that AI-driven output growth showing up as real revenue expansion is the moment the macro case gets genuinely strong (and will start showing up in GDP figures). Until then, this is not a productivity story, and the massive spending does not guarantee ROI)