Informational complexity bottlenecks. So many things are shackled to human decision making loops. If we were truly serious, we would unshackle everything and let it run wild. Would be chaotic, but chaos create strange attractors.

Quality control, for one. The state of commercial software is appalling. Writing code itself is not enough to get a useable piece of software.

LLMs are also not very useful for long term strategy or to come up with novel features or combinations of features. They also are not great at maintaining existing code, particularly without comprehensive test suites. They are good at coming up with tests for boiler plate code, but not really for high-level features.

Considering how software is increasingly made out of seperate components and services, integration testing can become pretty damn difficult. So quite often, the public release is the first serious integration test.

From my experience, this stuff is rarely introduced to save developers from typing in the code for their logic. Actual reasons I observe:

1. SaaS sales/marketing pushing their offerings on decision makers - software being a pop culture, this works pretty well. It can be hard for internal staff to push back on What Everyone Is Using (TM). Even if it makes little to no sense.

2. Outsourcing liability, maintenance, and general "having to think about it". Can be entirely valid, but often it indeed comes from an "I don't want to think of it" kind of place.

I don't see this stuff slowing down GenAI or not, mainly because it has usually little to do with saving time or money.