AI is Starting to Show Up in the Data
Sept 20, 2026
By Jack
Enormous sums of money have been invested in the Artificial Intelligence buildout worldwide. Many have called this a “bubble” referencing back to the dotcom crash in the early 2000s where investors funded massively overbuilt internet infrastructure, only to see their fortunes wiped out as all the new capacity was built substantially ahead of demand. Others have called for an apocalyptic impact on the job market. But, is this time different? Let’s start with jobs…
We Finally Have some Data
Gusto, a large payroll and benefits company, compared headcount trends between small businesses that adopted AI and businesses that were aware of AI but had not adopted it. What they found was that businesses that adopted AI grew headcount roughly 7% more than businesses that knew about AI but didn't adopt it. While this isn’t proof AI drove the job growth, it’s an interesting association.
Can AI eliminate tasks without eliminating jobs?
Here is one possible explanation…
Before AI
10 employees
Significant time spent on administrative work
Limited capacity
Growth requires adding overhead
After AI
Automated scheduling, reporting, research, invoicing, follow-up, data entry, etc.
Existing employees produce more
Business can serve more customers
Sales/output increase
Eventually another constraint appears: capacity to actually deliver the product or service
Business hires another technician, salesperson, therapist, accountant, etc.
AI → Productivity → Capacity → Sales/Output → Hiring
Is the AI Infrastructure being used? Is it having an impact?
Typically, the first phase of a technological revolution is investment in infrastructure with the economic payoff coming later when businesses actually deploy the technology. While we’re still well within the installation phase of AI, we’re starting to see the results of its deployments.
Non-residential equipment investment increased at a 15.8% annualized rate in Q1 2026 and another 13.6% in Q2. — U.S. Bureau of Economic Analysis
Manufacturing output grew at a 5.4% annualized rate in Q2 while hours worked increased just 2.9%, resulting in a 2.4% increase in labor productivity. — U.S. Bureau of Labor Statistics
Manufacturing productivity increased at a 2.4% annualized rate in Q2 while unit labor costs declined 0.3%—the first quarterly decline in manufacturing unit labor costs since 2021. — U.S. Bureau of Labor Statistics
None of these figures can be attributed solely to AI, nonetheless, you would expect the significant technological investments to start flowing through the economy. The data implies that the AI infrastructure is being put to use (to the joy of investors).
What this means for Business Owners
Buy every AI tool you can find? No…
The wrong question is “how many people can AI replace?”. The right question is “How much additional productive capacity can AI create?”. That’s why business owners should baseline KPIs before any AI implementation to ensure they are tracking value appropriately. Measure the impact of what you’re doing. Measure what matters.
Cost to Serve
Revenue per Employee
Gross Margin per Employee
Transactions per employee
Cycle Time
Customer Conversion / Retention
Still early, but watch the Data
We’re starting to see positive economic signals likely tied to the worldwide Artificial Intelligence rollout. One of the most interesting signals is that some early adopters of AI actually need more - not less headcount - to support their business due to productivity gains unlocking significant capacity for growth. The impact of AI may not be so much how many jobs it eliminates, but how much more companies can produce.
Where is your company on this journey? Comment below if these findings are consistent - or inconsistent - with what you’re seeing.
