As a certified AI Hater, I do have to say: We seem to have found one (1) use-case for LLMs where they're useful and (can be) prosocial: Finding software vulnerabilities.
This wasn't true a few months ago, but it seems the scales have finally tipped.
It ticks the boxes for me:
- Verifiable
- "Generative" aspect is limited
- Utility that isn't just replacing human labor
(I don't *like* it, but... this does seem to be legit. Just be wary of the hype.)
@varx "prosocial"...hard to be prosocial when you are contributing to increased climate change and scarcity of water.
On the software front, I guess one could say LLMs have brought the reckoning that the software community has had coming for a long time, but it is not a salvation; instead of leading to recognition of the actual problem (lack of valid analysis models and metrics that could lead to improving abstraction), it doubles down on the current poor practices.
#DumbAllOver
@lwriemen Yeah, that's what I mean about the cost/benefit calculation. Even if the use-case is prosocial, it's not clear if it's worth it.