I don’t have any data on me to be honest, most of what I hear is based on people I know purchasing second hand EVs, and from what I know of EVs. They suffer from battery degradation which right now is an expensive swap, and charger compatibility is heavily dependent on infrastructure that takes forever to roll out where I am. Internal combustion engines will run great even after 150,000 kms, whereas conventional electric cars and even hybrids show degradation at 80,000 on a few models.
Newer EVs likely have less of these problems, but when I talk of purchasing second hand, I’m talking cars that are 10+ years old.
I’m conflicted because more people joining the open source community and contributing is great. That said, open source is an ecosystem that covers everything from the application, to the build system and the code editor used to write the code. A lot of the people contributing to FOSS care a lot about the quality of their code, and being part of it means respecting the community.
But most of the AI models being used to write vibe coded or ai assisted code is not at all sustainable, it’s a huge part of the toolset that is locked behind corporate infrastructure. A lot of code spit out by LLMs is rarely workable or designed to be extended or transformed, and very few are critical of the quality of the code. I know even at work I’ve been frustrated at having to fix, work with or re factor code that has been written with the help of ai.