Software development really has gone off the deep end. In any other field, people actually document what they do, and verify that it works, before releasing anything. Restaurants have recipes and food handling requirements, manufacturers have tolerances and verification, architects have building code, warehouses have inventory management, and so on and so forth. Because non-software development relies on products that actually work, and are not built around the meta game of abusing arbitrary metrics, workers can rely on other departments to make sensible choices.
Fun fact: Waterfall development never existed; it's a straw man argument against the common sense idea of finishing what your working on, before starting something new.
AI is going to pop the bubble of software development, not because it's good at it, but because because the entire field is too broken to compete against it.
This is how software development compares to other fields: https://xkcd.com/2030/
Software development really has gone off the deep end. In any other field, people actually document what they do, and verify that it works, before releasing anything. Restaurants have recipes and food handling requirements, manufacturers have tolerances and verification, architects have building code, warehouses have inventory management, and so on and so forth. Because non-software development relies on products that actually work, and are not built around the meta game of abusing arbitrary metrics, workers can rely on other departments to make sensible choices.
Fun fact: Waterfall development never existed; it's a straw man argument against the common sense idea of finishing what your working on, before starting something new.
AI is going to pop the bubble of software development, not because it's good at it, but because because the entire field is too broken to compete against it.