Earlier quoted context omitted.
What would the cost of the “next machine” be? Is it going to be tens of billions or can we make progress with lesser money. If it is going to be tens of billions, then maybe we need to invest in engineering to reduce this cost, because it’s not sustainable to suspend thirty years, tens of billions for every incremental improvement.
This kind of slow, incremental improvement that costs tens of billions of dollars and takes decades gave us the microchips that ultimately enabled you to type this comment on your phone/computer. The return on that investment is obvious. But it is not just about making money: The entire field of radiation therapy for cancer exists and continues to improve because people figured out ways to control particle beams with…
There is a lot of handwaving going on here to justify the incredibly cheap, mostly privately funded investments that launched the computer generation with the massively expensive, extremely gradual gains we are making now with particle accelerators. Part of it is that people just can't imagine how little was invested in R&D to get these stunning results, given how much we have to invest today to get much less impressive results, so they just assume that semiconductors could not have been invented without tens of billion dollars of research.
There is diminishing returns, just as a 90nm process is really all you need to get 90% of the benefits of computerization -- you can drive industrial automation just fine, all the military applications are fine, etc. But to go from a 90nm process to a 3nm process is an exponential increase in costs. In a lot of fields we are at that tail end where costs are incredibly high and gains are very low, and new fields will need to be discovered where there is low hanging fruit, and those fields will not require "tens of billions" of dollars to get that low hanging fruit.
Even with particle accelerators, SLAC cost $100 million to build and generated a massive bounty of discoveries, dwarfing the discoveries made at CERN.
To pretend that there is no such thing as a curve of diminishing returns, and to say that things have always been this way is to not paint an accurate picture of how science works. New fields are discovered, discoveries come quickly and cheaply, the field matures and discoveries become incremental and exponentially more expensive. That's how it works. For someone who is in a field on the tail end of that process, it's not good history to say "things have always been this way and have always cost this much".