Being an ML engineer sometimes feels more like being a manager than an individual contributor. When writing software, you think through every edge case & behavior yourself. In ML engineering, you shape the behavior of systems through indirect means. Trickier, but more leverage.

Apr 28, 2022 · 5:09 PM UTC

16
20
2
365
Replying to @gdb
And if you get it right the results will be mind blowing
2
Replying to @gdb
Great analogy Debugging & optimizing ML models really feels more like trying to convince people than writing a deterministic computer program
Replying to @gdb
I was a software engineer for 45 years. Unfortunately I retired 10 years ago before the real fun started!
Replying to @gdb
That’s the mark of a good MLE
Replying to @gdb
This is one of the most profound and sincere statements about ML. Thank you very much for the insight. Einstein said: If you can't explain it simply, you don't understand it well enough.
Replying to @gdb
Sometimes like a product manager
Replying to @gdb
Will there be a time when a language nlp like interafce will be used to do this shaping or ML tools to do ML. Aspire?
Replying to @gdb
"In ML engineering, you get to save the time worrying about edge cases by disclaiming responsibility and hoping the model figures them out." cc @filippie509
1
3
Replying to @gdb
Yikes. That has some serious ethical implications for AI.
Load more