Comment on Learnings from building AI agentsparentComments−AndrewKemendo1yOtherwise known as science1:Observation 2:Hypothesis 3:test 4:GOTO:1This is every thing ever built everWhat is the problem exactly?−wrs1yFor one thing, what you learned can stop working when you switch to a new model, or just a newer version of the “same” model.−AndrewKemendo1yAll that means is that you verified the null hypothesis which should be that it doesn’t workIf you create hypothesis tests that are not written in or specific enough then you’re right you’re not gonna be able to do scienceIncidentally 99.9% of people I know have no instinct for how to actually do science or have rigor or focus to actually do it in a way that is usable
Comments
Otherwise known as science
1:Observation 2:Hypothesis 3:test 4:GOTO:1
This is every thing ever built ever
What is the problem exactly?
For one thing, what you learned can stop working when you switch to a new model, or just a newer version of the “same” model.
All that means is that you verified the null hypothesis which should be that it doesn’t work
If you create hypothesis tests that are not written in or specific enough then you’re right you’re not gonna be able to do science
Incidentally 99.9% of people I know have no instinct for how to actually do science or have rigor or focus to actually do it in a way that is usable