I use the things to help me make more bulletproof writing, when I care to pre-empt the attacks.
It could just as easily be used to identify bugs before committing code, but I’d want multiple different models trying to find bugs, & then using … perhaps a quorum of the things, to decide if the bugs were real, before bothering the actual-human-programmers…
The false-negatives/false-positives problem that all systems have, when it comes to identifying problems… ( airport-screening, infected-with-pathogens, etc: the problem exists in lots of domains ).
The thing is, though, that the 1 thing that the LLM isn’t providing, is the 1 thing that no for-profit-operation seems to be valuing: good judgement.
& when we select against good-judgement, … that’s a DarwinAward in the making, at civilizational-level, isn’t it?
But that’s fine, so long as it’s what people want…
Natural Selection will take care of its work: we only have to force its hand the way we unconsciously-want, & it’ll do the rest.
to me, that article looks exactly right.
I use the things to help me make more bulletproof writing, when I care to pre-empt the attacks.
It could just as easily be used to identify bugs before committing code, but I’d want multiple different models trying to find bugs, & then using … perhaps a quorum of the things, to decide if the bugs were real, before bothering the actual-human-programmers…
The false-negatives/false-positives problem that all systems have, when it comes to identifying problems… ( airport-screening, infected-with-pathogens, etc: the problem exists in lots of domains ).
The thing is, though, that the 1 thing that the LLM isn’t providing, is the 1 thing that no for-profit-operation seems to be valuing: good judgement.
& when we select against good-judgement, … that’s a DarwinAward in the making, at civilizational-level, isn’t it?
But that’s fine, so long as it’s what people want…
Natural Selection will take care of its work: we only have to force its hand the way we unconsciously-want, & it’ll do the rest.
_ /\ _