←back to thread

334 points mooreds | 1 comments | | HN request time: 0.315s | source
Show context
mgraczyk ◴[] No.44485742[source]
Important for HN users in particular to keep in mind: It is possible (and IMO likely) that the article is mostly true and ALSO that software engineering will be almost completed automated within the next few years.

Even the most pessimistic timelines have to account for 20-30x more compute, models trained on 10-100x more coding data, and tools very significantly more optimized for the task within 3 years

replies(1): >>44485784 #
yladiz ◴[] No.44485784[source]
It's naive to think that software engineering will be automated any time soon, especially in a few years. Even if we accept that the coding part can be, the hard part of software engineering isn't the coding, it's the design, getting good requirements, and general "stakeholder management". LLMs are not able to adequately design code with potentially ambiguous requirements considered, and thinking about future ones too, so until it gets anywhere close to being able to do that we're not going to automate it.
replies(1): >>44485792 #
1. mgraczyk ◴[] No.44485792[source]
I currently use LLMs to design code with ambiguous requirements. It works quite well and scales better than hiring a team. Not perfect but getting better all the time.

The key is to learn how to use them for your use case and to figure out what specific things they are good for. Staying up to date as they improve is probably the most valuable skill for software engineers right now