# As an ML/CV engineer of 7+ years, the current generative AI hype train makes me pessimistic. Where are the non-LLM, real-world hard problems still waiting to be solved? — retained record

> Retained reddit record from reddit:artificial.

Canonical: https://highsignal.app/data/records/c25456c619f0aace

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:artificial)  published 2026-11-09 12:43 ingested 2026-11-10 02:37

[original source ↗](https://www.reddit.com/r/artificial/comments/1xrqpq1/as_an_mlcv_engineer_of_7_years_the_current/)

## retained text



```
I love machine learning. I spent years training custom vision models, working with time-series data, and architecting multimodal pipelines from scratch. But the current shift toward treating AI exclusively as "prompt engineering" and massive LLM wrappers feels reductive and disappointing.

I want to build software that creates genuine, positive real-world impact. Specifically in fields like environmental science, specialized B2B workflows, or complex physical monitoring, where off-the-shelf generative APIs simply don't cut it.

For engineers and researchers still working on core ML/CV:

1. What are the high-impact, non-LLM problems that still desperately need custom models and deep domain expertise?
2. How are you maintaining enthusiasm for foundational data science and ML craftsmanship in an industry currently obsessed with generative text and image tools?
```



document fetched 2026-11-10 00:19 · canonical https://reddit.com/r/artificial/comments/1xrqpq1/as_an_mlcv_engineer_of_7_years_the_current
