> This seems to challenge belief. There are almost always rapidly diminishing returns on increasing allocation of resources once you reach a certain baseline. A

> This seems to challenge belief. There are almost always rapidly diminishing returns on increasing allocation of resources once you reach a certain baseline. And almost is a weasel word there -- I can't think of anything where this is not true. Just like 50 programmers won't work 50 times faster than 1 programmer on a given task, 50 times the amount of money dedicated to a task will often provide far less than 50x the return. And the article's premise is suggesting not only a linear return on investment, but a largely exponential one. I see research, especially fundamental research, like a lottery in which we don't know what we will win. State invests X billions of dollars, and a breakthrough will likely happen, but maybe not in the expected problem (principle of serendipity). More X billions means a higher chance of breakthrough, and hopefully on the desired problem. Surely experts and thinkers looked at efficiency in research, but by definition researching is about looking at the unknown, and when there is uncertainty, a lottery model is probably a better way of thinking about it.