This paper models an entrepreneur’s choice between investing in a safe activity or experimenting with a new risky one, and how much to invest in the “entrepreneurial capital” that would permit more effective use of the arriving information on the latter- how much to learn how to learn. Optimal investment in entrepreneurial capital depends the expected return on the risky activity. It can lead to three learning regimes, two of which can generate a development trap where firms and countries are unable to assess the potential of newly arriving technologies and hence grow more slowly. The first arises purely because it is too expensive to learn to learn, the second because the returns to the new activity are so high that they obviate the need to distinguish between activities and hence invest in entrepreneurial capital. The paper draws on historical evidence to show how the model offers insights into three understudied features of the industrialization process in the Western Hemisphere at the beginning of the 20th century: the disproportionate influence of immigrant/foreign entrepreneurs in driving industrialization in Latin America; the emergence of selective exceptions to this pattern, as well as episodes of entrepreneurial retrogression; and the differing effects of similar economic structures across countries that suggest the possibility of a learning-displacing resource curse. The model can simulate the respective decline and boom in the Chilean and US copper industries at the turn of the century, arising either from initially high relative returns or low initial endowments of entrepreneurial capital in the latter.