Five-year capacity model

The cost of waiting to start with AI

Compare starting your operating practice now, starting later, and doing nothing new.

A transparent reasoning aid for executives. No email required; your inputs remain in this browser.

What does a later start leave out of the next five years?

Two kinds of improvement can happen at once. The AI base available to an executive may improve with calendar time. The executive’s instructions, context, tools, and review practice improve only after operating work begins. Neither rate is automatic; both are assumptions to challenge.

The model keeps those effects separate. A delayed start can use a newer AI base, but it does not receive operating practice for the months it sat idle. A third path, do nothing, represents no new agent practice and therefore no modeled incremental capacity. It is not a forecast of the company’s existing performance.

Compensation provides a common time-value proxy. The result is neither lost cash nor a forecast of business results.

Read the reasoning behind the model in the executive brief →

Change the six assumptions

Set assumptions you can defend, then compare the five-year paths for starting now, starting later, and doing nothing new.

Set your assumptions

Two kinds of growth. Three choices.

$1,000,000

Used only as a time-value proxy for your working capacity.

12 months

Operating practice starts later; the available AI base can still improve while you wait.

10% of the week

The ongoing investment subtracted from each active month.

Share of workload
20% of workload

100% equals one current executive workload; higher values represent parallel agent capacity. The slider ends at 1,000%, but compounded capacity is not capped.

3% per month

Compounds only after you start building instructions, context, and review practice.

2% per month

A calendar-time assumption applied to both adopted paths, including the months before a delayed start.

Five-year timing gap

Waiting one year leaves an estimated $1,889,699 of five-year executive capacity unrealized.

Compensation is only a time-value proxy. This is neither lost cash, promised upside, nor a forecast.

Net workload-equivalent capacity now
4 workload-equivalent hours/week
Net workload-equivalent capacity at Year 5, starting now
143.2 workload-equivalent hours/week

Net workload-equivalent capacity subtracts the ongoing management commitment. Values above 40 hours reflect parallel agent output, not recovered clock time.

The chart compares smooth monthly curves for starting operating practice now, starting after the selected delay, and doing nothing new. The delayed path still receives calendar-time improvements in AI base capability. Do nothing means no modeled incremental agent capacity and is not a forecast of existing company performance. Year 0 through Year 5 appear as axis ticks; the table gives annual values in text.

Start now Wait 12 months Do nothing

Hover or tap the chart to inspect a month.

Month 60. Start now: $5,537,724; wait 12 months: $3,648,025; do nothing: $0; timing gap: $1,889,699. Hover or tap to inspect another month.

Annual capacity comparison
Compensation-weighted cumulative net executive capacity
PointStart nowWait 12 monthsDo nothingTiming gap
Year 0$0$0$0$0
Year 1$166,209$0$0$166,209
Year 2$547,570$237,617$0$309,953
Year 3$1,317,971$748,099$0$569,872
Year 4$2,791,840$1,751,979$0$1,039,861
Year 5$5,537,724$3,648,025$0$1,889,699

How the two growth rates combine

The model holds the horizon at 60 months and uses a 40-hour workweek only to express workload-equivalent capacity. It applies no ceiling to the calculated path. Capacity may exceed one workload or the starting slider’s 1,000% bound as the two factors compound.

Assumptions and calculation
  • For calendar month c, AI base capability is multiplied by (1 + base AI improvement)c.
  • For active operating month a, operating practice is multiplied by (1 + operating improvement)a.
  • Gross workload-equivalent share multiplies the starting share by both factors. Net share then subtracts the ongoing management share.
  • Monthly compensation-weighted net capacity is annual compensation multiplied by net share, divided by 12. The cumulative series sum active months.
  • Start now is active at calendar month 0. The delayed path is inactive through the selected delay, then begins with the calendar-time AI base already advanced and operating practice at month 0.
  • Do nothing is fixed at zero because it models no new agent practice and no incremental agent capacity. It is not the delayed path or a forecast of current company performance.
  • Negative net values are not clamped. They show an investment phase when management takes more capacity than the agents can yet handle.

base multiplier(c) = (1 + base AI improvement)c
operating multiplier(a) = (1 + operating improvement)a
gross workload-equivalent share(c,a) = starting share × base multiplier(c) × operating multiplier(a)
net share(c,a) = gross share(c,a) − management share
monthly net capacity(c,a) = annual compensation × net share(c,a) ÷ 12
cumulative capacity = sum of active monthly net capacity

With the defaults, net workload-equivalent capacity begins at 4 hours per week and reaches about 143.2 hours per week at Year 5 when starting now. The cumulative start-now path is about $5,537,724; waiting 12 months produces about $3,648,025. The five-year timing gap is about $1,889,699; do nothing remains $0.

Compensation is a directional proxy for the time-value of one executive’s capacity. The model does not estimate cash savings, revenue, agent accuracy, implementation effort, security or legal risk, adoption by a team, or the consequences of better and worse judgment. It does not compare fees, apply a team multiplier, or assign a probability to success.

Challenge both improvement rates, not just the total

The result is useful only when the six assumptions describe work you can recognize. Start with the workload share: name what one or more agents could actually handle, then name the weekly work required to direct and review it.

  1. Which decisions, analysis, drafting, preparation, or delegation belong in the starting workload share?
  2. What external evidence would justify the AI base-improvement rate?
  3. What operating evidence from your own use would justify the practice-improvement rate?
  4. If you wait, what specific condition will change, who owns it, and when will you revisit the decision?

At zero delay, the two adopted paths are identical and the timing gap is zero. With both improvement rates at zero, each adopted path is linear after it starts. A negative net path is also possible when management exceeds modeled workload-equivalent capacity.

Take the AI•DOOM Assessment — identify the operating constraint

A directional self-assessment, not a grade or diagnosis. No implementation, legal, or security advice.

Operating Leverage Session — $995
X in f link