A clinical trial is not a single experiment but a sequence of studies, and each one has its own clock. Phase I asks whether the drug is safe in a small group, typically a few dozen people, and usually runs for several months to about a year. Phase II asks whether it shows signs of working, in a few hundred people, and typically takes one to two years. Phase III confirms benefit against the current standard of care in hundreds to thousands of people and commonly runs two to four years. Add the time to design the protocol, obtain ethics and regulatory clearance, recruit sites, and analyze results, and a single drug program routinely spans eight to twelve years from first human dose to approval.
Cost accumulates unevenly across that timeline. Early phases are comparatively cheap because they involve few participants and short follow-up. Phase III dominates the budget because it enrolls the most people, runs the longest, requires the most sites and staff, and carries the cost of manufacturing and monitoring at scale. This is why the financial risk of drug development is concentrated at the end: by the time the largest bills come due, the sponsor has already spent years and cannot easily walk away.
The reason the clock cannot simply be sped up is that the measurements are biological and clinical, not computational. A trial must wait for enough participants to be enrolled, for the drug to be taken long enough to show an effect, and for enough clinical events — recoveries, relapses, side effects — to accumulate before the result is statistically meaningful. No amount of computing power changes how long a human body takes to respond.