Ancestors

Toot

Written by Joost van Baal-Ilić on 2025-01-06 at 06:46

Some things to expect in 2025 By Jonathan Corbet, January 2, 2025: "A major project will discover that it has merged a lot of AI-generated code, a fact that may become evident when it becomes clear that the alleged author does not actually understand what the code does. We depend on our developers to contribute their own work and to stand behind it; large language models cannot do that."

https://lwn.net/SubscriberLink/1003780/64f85a80923c174e/

=> More informations about this toot | More toots from joostvb@mastodon.green

Descendants

Written by 🌈☔🌦️🍄 on 2025-01-06 at 07:57

@joostvb "We depend on our developers to contribute their own work and to stand behind it; large language models cannot do that.". Just thinking of all those devs that just copy-paste of internet examples, using libraries/apis/etc they don't understand. The entire idea of clicking together code... I think we lost understanding of the code we provide a good while ago now. Llm's are a next step in that.

=> More informations about this toot | More toots from wmd@chaos.social

Written by 🌈☔🌦️🍄 on 2025-01-06 at 08:05

@joostvb the risk of llms and how we handle code these days I think is less some emergent AI, more on how we depend/rely more and more on systems we understand less and less. Moving into the realms of wizardy, witchcraft and magic. Combined with marketing then also religion.

In the world of IT, we've mostly worked on complicating systems, not understanding them.

=> More informations about this toot | More toots from wmd@chaos.social