Views : 56,068
Genre: Science & Technology
Date of upload: Apr 20, 2024 ^^
Rating : 4.986 (6/1,747 LTDR)
RYD date created : 2024-05-17T05:15:06.691198Z
See in json
Top Comments of this video!! :3
Hahaha, Abstract evaluation that rewards creating public repos for no use except boosting profile stats, next generation of GitHub farming.
Can't wait to see someone use purely AI code so another AI can evaluate it positively, giving them a high rating while they just copy paste LLM code hahaha
199 |
The problem with AI as a candidate filter is very simple. It can have significant amounts of both false positives and false negatives. If you have a filter that only gives either false positives OR false negatives, that can still be useful because it still removes a large amount of work for people to do.
If your filter gives false positives AND false negatives at a significant rate, it isn't really doing anything useful for you.
2 |
So in the future I won't only have to impress some HR with high quality visuals because they can't understand anything of my actual engineering processes, I will also have to impress some AI that my github with personal projects aimed at learning stuff and showing that I am actively learning stuff isn't representative of my production ready engineering?
Is there an option to speak to an actual human I can professionally relate to in those options somewhere?
13 |
AI image generators train on photographs and 3D renders (as well as human art), so they have a massive advantage over LLMs (as humans have always been impressed by highly realistic artwork).
An LLM can only train on the (imperfect) source code that people created (by hand). There's nothing analogous to a photograph or 3D render to teach an LLM what a flawless program would look like.
1 |
@onegamingmoose
3 weeks ago
According to GitRoll I'm a "Senior-level AI/ML Developer". I've never touched AI/ML but I expect a massive pay rise any second now
445 |