Hacker News Reader

Comment by throwaway1183 | Hacker News Reader

I have been gathering experience in computer vision, graphics, and natural language processing to break into the tech industry as a junior research engineer rather than a research scientist. However, with recent advancements in technology, I feel like I have no place in the field anymore. I have lost all of my curiosity and inner drive. It feels like I was just about to start my career, but now I am not sure what to do. I want to switch fields, but I am worried about all the time and effort I invested in learning these models and their implementation. If I had invested that same energy in a different field, I would have more reliable experience and skills. All I have now is pre-historic knowledge. And if I were to catch up now just because I am interested, by the time I do, the goal line will have moved further away. It's a difficult and frustrating situation to be in.

View on HN

Don't be too discouraged. I'm trying not to be, and one thing I have definitely noticed in my career in corporate jobs is this: Just because a technology exists that could do a job does NOT mean that anyone in the company will implement it in a way that actually replaces the humans doing it. A startup may avoid a lot of work (and therefore labor expense) by being smart and nimble about technology, but established companies rarely do. As an example, most big companies still run multiple Exchange servers in 2022 (the multiple part is usually related to acquisitions), while a startup started in 2022 would of course never dream of doing so, since it would be madness. There are a lot of jobs out there that are the equivalent of running an on-prem Exchange server. They aren't glamorous but they will keep you fed and housed.

Also, there will likely be plenty of work in the coming decade unwinding half-baked attempts to integrate large language models into business processes where it doesn't ultimately make sense and is a net negative. Many of these negative-value projects are probably being conceived at this very moment.

All is not hopeless.

This gives some hope! I want to learn that skill that is equivalent to "managing these exchange servers" that might actually last throughout the career. This "latest and greatest" hype chasing has seriously burnt me out.

Something to think about btw: a lot of big companies have a _ton_ of proprietary, internal complexity that is not exposed to the public and therefore cannot be assisted by these general purpose LLM’s.

Seems a bit strange but being a “Company Man” might actually bring a new degree of job security…

Another thought: perhaps LLMs will "commoditize the complement" of lots of IT work. I imagine lots of CoPilot-generated code will be deployed by naive people and organizations without proper "IT 101" stuff like backups, RBAC, PCI/SOX/GDPR compliance, etc. It might INCREASE the demand for some types of IT talent.

Reply on news.ycombinator.com