Can enterprises spend their way to AI leadership? Not on investment alone. Compute and specialist hiring can enable AI work, but they do not guarantee broad adoption, better productivity, or financial returns. Durable advantage depends on whether a company can fit AI to real workflows, prepare its systems and processes, and sustain the engineering work needed to put it into practice.
What AI spending can—and cannot—buy
Capital can secure computing capacity, tools, and expertise. Those are important inputs for some AI ambitions, but they are not the same as outcomes. A company may invest heavily without turning that investment into routine use or measurable business value.
| # | Preview | Product | Price | |
|---|---|---|---|---|
| 1 |
|
AI-Powered Leadership: Mastering the Synergy of Technology and Human Expertise | $29.79 | Buy on Amazon |
| 2 |
|
AI-Ready Leadership: Leadership and Management in the AI Era | $4.99 | Buy on Amazon |
| 3 |
|
AI-First Leader | $50.24 | Buy on Amazon |
As an Amazon Associate I earn from qualifying purchases.
Joe Bertolami, co-founder and CTO of Clifton AI, makes this case in a September 10, 2026, CIO opinion piece: “Hoarding all the compute in the world only gets you so far if your talent is fleeing.” That is an argument about the limits of spending, not proof that talent retention alone determines which companies lead. The broader evidence supports a more qualified conclusion: results vary, and companies need to connect investment to adoption and business needs. Read Bertolami’s opinion piece.
Adoption is not the same as value
Stanford HAI’s 2026 AI Index reports that 88% of organizations have adopted AI. That broad measure says little by itself about how extensively organizations deploy AI, whether deployments work well, or whether they improve financial results. Adoption is a starting point, not a leadership scorecard. Stanford HAI’s 2026 AI Index also urges care in interpreting model benchmarks and independent tests.
#1 Best Overall
A 2026 National Bureau of Economic Research working paper, based on a survey of nearly 750 corporate executives, reports positive but varying labor-productivity gains across firms. The authors associate observed gains more with revenue-based productivity and innovation- and demand-oriented channels than with capital deepening alone. This is survey evidence, not a causal guarantee that any particular investment will raise productivity. See NBER Working Paper 34984.
For an enterprise, the practical question is not simply how much it spends or which model it uses. It is whether a particular system improves a specific workflow in a way the business can observe and sustain.
Why organizational readiness matters
AI adoption depends in part on the conditions into which a system must fit. A Stanford Graduate School of Business study published in 2026 measured AI use at 22.8% of US manufacturing plants in 2021, based on a purpose-designed survey of about 28,500 establishments. That figure is a historical, sector-specific measurement—not a current rate for all businesses.
PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchThe study links adoption with more recent digital infrastructure and structured production processes. It also identifies cost, the absence of an applicable use case, and lack of expertise as barriers. The findings suggest why buying access to a model or computing capacity may not be enough: organizations need suitable processes, a real problem to solve, and people able to integrate the technology. Read the Stanford GSB study summary.
Cheaper models widen the choices, not the guarantees
Lower query costs can change what is practical, but a price comparison is meaningful only when its scope is clear. Stanford HAI’s 2025 AI Index reports that the cost to query a model scoring at GPT-3.5-equivalent accuracy on MMLU fell from $20 per million tokens to $0.07 per million tokens between November 2022 and October 2024; Gemini-1.5-Flash-8B is the example for October 2024. This is a benchmark- and model-specific comparison, not a universal estimate of enterprise AI costs. See the AI Index economy chapter.
Bertolami’s opinion piece also argues that open-weight and distilled models, along with smaller or localized deployments, can offer alternatives as enterprises weigh privacy and operating costs. Those options are not automatically suitable for every job. A smaller model may cost less to run yet fail a task’s quality threshold; a local deployment may offer a different data-handling profile while requiring integration and maintenance work.
Evaluate options against the actual workload rather than treating model category or benchmark standing as a decision by itself:
Free tools Windows power users keep installed
One-click scans. No signup required.
- Task quality: Does the system perform reliably on the company’s own work, not just a general benchmark?
- Total operating cost: Include the costs of running and maintaining the deployment, not only a quoted query rate.
- Data handling: Determine whether the deployment meets the organization’s privacy and data requirements.
- Integration and upkeep: Confirm the organization can connect, monitor, and maintain the system over time.
How to judge whether investment is becoming an advantage
Because investment, adoption, productivity, and financial return are distinct, assess them separately when reviewing an AI initiative:
Rank #3
- Specify the business need. Name the workflow or customer problem and the outcome that would count as improvement.
- Check readiness. Look for suitable digital foundations, structured processes, and people with the expertise to implement and support the system.
- Test fit before scaling. Evaluate model quality on the intended task and weigh cost and data-handling requirements against the organization’s ability to operate the deployment.
- Track use and results separately. Record whether people actually use the system, whether the workflow improves, and whether the business sees a return. A high adoption figure alone does not establish the latter two.
Investment remains an enabler: some AI ambitions require substantial computing capacity and specialist work. But spending is not a substitute for a viable use case, prepared processes, or effective engineering. The evidence does not establish a universal ratio between AI spending and business value, nor does it prove that one factor—such as talent retention—determines leadership. The strongest case for spending is therefore a specific, measurable business need backed by the ability to deliver and sustain the solution.
What large capital forecasts do—and don’t—show
An NBER working paper published in 2026 reports $380 billion in 2025 capital expenditure by the five largest US technology firms, with roughly double forecast for 2026. Those figures concern the named firms’ capital expenditure; they are not interchangeable with the CIO opinion piece’s projection of more than $750 billion in AI infrastructure spending. The estimates differ in scope, so they should not be merged into a single measure of enterprise AI investment. See NBER Working Paper 35290 and the CIO opinion piece.
Large forecasts show that major technology companies are committing substantial capital. They do not, by themselves, demonstrate that a particular enterprise is adopting AI effectively or earning a return. For other companies, the relevant lesson is to judge investment by what it enables in a real business context—not by the size of the market’s spending totals.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
