See these two Richard Schodde graphs.

Ultimately, however, mineral exploration underwent a very significant structural change because of this progressive shallow ore deposit depletion. This structural change was signalled by the fact that, for the first time in modern history, a very strong multi-year increase in mineral exploration expenditure was not answered by a parallel increase in ore deposit discovery (see below).

Why did gradual shallow ore deposit depletion finally cause the structural change illustrated above?

Assuming that mineral exploration geologists, academia, and general technological progress were busily all at work, it would have been expected that continued human problem-solving would have eventually surmounted the shallow ore deposit depletion problem by learning to discover unmined deposits marginally deeper than those historically exploited.

Schodde shows that this presumably ongoing industry and academic problem-solving effort has not been at all very fruitful, however.

In the case of gold deposit discoveries over the 2005-2014 time period, for example, the world-wide average gold deposit discovery depth was still only about 200 feet.  It turned out that expenditure is not a substitute for valid expertise when the targets of such expertise have finally very largely disappeared.

So, what is keeping the mineral exploration industry from finding deeper ore bodies? I propose that the problem is largely a case of an ‘elephant in the room', namely, mineral exploration's continued reliance on the accumulated experience and knowledge of its own now largely outmoded experts. Kahneman and Klein review the functional limits of expert judgment and both agree that, barring sufficient opportunities to observe and learn, it is impossible to develop and then employ professional expertise.

And, what is keeping the exploration industry from making all of the observations necessary to learn how to more reliably discover deeper ore deposits? It is a Catch-22 situation: the industry overconfidently and unsuccessfully keeps using the much-too-limited knowledge and experience it laboriously gained over centuries of exploring for shallow, grassroots ore deposits to attempt the new, much more difficult and different task of learning to find deeper ore deposits.

Fortunately, Kahneman and Klein prescribe a very effective bootstrapping solution to the conundrum now facing the mineral exploration industry, a conundrum keeping the industry from obtaining the new observation and learning opportunities required to eventually generate new, genuinely authoritative and reliable expertise in exploring for more deeply buried ore deposits. 

You will find some already developed deep-cover mineral exploration examples of the algorithm method of augmenting otherwise inadequate professional judgment recommended by Kahneman and Klein here. - Larry Turner, DIR Exploration Inc, Colorado, USA