Technical debt is the single largest obstacle to innovate according to research done in Norway. Today we know how to avoid technical debt, at least in greenfield projects. This doesn't imply we don't accumulate it on new projects though, due to humanity's nature being to gravitate towards repeating the same mistakes over and over again. But at least in theory, we know how to avoid it, and the recipe is quite easy actually.
- Never reinvent the wheel, instead use pre-existing building blocks where you can
- Use micro services where you can to loosely couple together your end result
- Automate as much as possible
- Outsource as much as possible
If you follow the four above simple guidelines, technical debt cannot occur, not even in theory. This is because what you're doing is to basically "outsource" your technical debt - Either to "the machine" (automation), open source libraries (components), or 3rd party vendors (components and outsourcing). I happen to run one such company, who's sole purpose is arguably to "take unto ourselves your company's technical debt" for a fee. The idea is that this holds value for others, freeing up internal resources, such that they can innovate more freely, without being hindered by their internalised technical debt.