The last few years has seen a proliferation of accelerators for various reasons, including lower barriers to entry and capital needs for tech startups, and increasingly fluid information flows. As a result, lots of testing and learning can happen earlier in the startup process, and accelerators, with their combination of coaching, provision of professional services and introduction to investment, have served to address these earlier phases.

It’s easy to argue that too many emerged for the market: you can look at the ensuing fallout as evidence. Jed Christiansen reveals some dead ones in his list of global programs and just this month a few other casualties were reported.

However, I contend that the changed dynamics of startups can bear this supply; the problem is rather with the mission of the accelerators rather than their number. I’ve been privileged to run a program unique in our focus and priorities on technology innovation over ROI potential. Considering the overall duration it takes to generate returns (let alone any decent returns at all) in early stage investing, accelerators focused on shorter-term (as in, less than say 7 years) ROI are not surprisingly at risk.

Perhaps more importantly, it’s crucial to recharacterize startup support at these early phases, until sufficient time has passed, as an educational rather than financial endeavor (though certainly with financial ramifications…just as a college degree has financial ramifications, but unless you go to trade school, that shouldn’t be your focus while getting the education).

Nearly a decade ago, Wired founder and overall tech philosopher-pundit Kevin Kelly posited that “Wasting time and inefficiencies are the way to discovery.” No wonder schools often feel inefficient and standardized tests are suboptimal ways to measure true learning and progress. For the love of the new and innovative, bring on the inefficiency. And bring on the sponsors of it who have much to benefit from the discoveries that ultimately get made.