With AGI supposedly around the corner and (more realistically) current LLMs performing at least incrementally better -- why are we even thinking that Microsoft's or Google's solutions will provide enough value vs competitors in 3, 5 or 10 years? Cheaper or free alternatives might soon reach feature parity, and even previously complicated deployment is now aided by AI.
Multiple years of AGI around the corner and still no demonstrated improvement in submissions to FOSS projects. If anything it's slowed them down by wasting maintainer time reviewing slop.
Thats a skill issue. Maintainers can get through a whole lot more than ever before because they can review PR much faster. Thats great for velocity. Features still need that expert touch since no software system is simple but it is getting easier.
