> They'll always need a bigger collider. That's where new results will always be. Bigger colliders are a low-risk approach: we'll definitely learn something, e

> They'll always need a bigger collider. That's where new results will always be. Bigger colliders are a low-risk approach: we'll definitely learn something, even if it's just better constraints on our models. This was a no-brainer back when colliders took up rooms or buildings. We've gone so far down this path that colliders are now measured in kilometres (and straddle national borders!), making it harder to justify the expense. Higher-risk approaches can be much cheaper, by looking for new paths (where we haven't hit diminishing returns yet). EmDrive is an example; wakefield accelerators are another. These have a greater chance of going nowhere (again, EmDrive), but cost so little that we can fund a whole bunch for the equivalent of a bigger collider. Of course, in best approach is to strike a balance between these two extremes: don't spend everything on the biggest possible collider; but likewise don't avoid projects just for being expensive. Another good example is ITER versus the various fusion startups.