Random AI thought of the day: the ML boom has suddenly put continuous maths centre stage in comp sci departments. When I was a student the curriculum was dominated by discrete math - sets, logic, etc. Fascinating to see the pendulum swing the other way…
The early decades of computer science were dominated by numerical analysis. Then in the early 2000s I had to persuade my colleagues that it wasn't passe. Now it's gone too far. AI will not be solved by continuous math alone. We need things like this: https://homes.cs.washington.edu/~pedrod/papers/ijcai15.pdf …
-
-
I tried to say this earlier as well… ai isnt going to be solved by contiguous math alone… cannot lose track of the goal & focus too much on prevalent methods
Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
-
-
AGI won't be solved by math at all, imo. Math is irrelevant to it. Neurons are way too slow and low-powered to be solving any kind of math other than simple additions and subtractions. AGI will be solved by applying simple general principles to large sets of neurons.
Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
-
-
Nonconvex nonlinear integer mathematical programming has been investigated for decades by operational researchers, mostly in industrial and chemical engineering departments. ML people often ignore the developments in OR or just reivent the wheel.
-
Training deep learning models in academia should have triggered research on initial value problems, in addition to exploration of nonlinear dynamics. However, I rarely see these paths for research being taken.
Keskustelun loppu
Uusi keskustelu -
Lataaminen näyttää kestävän hetken.
Twitter saattaa olla ruuhkautunut tai ongelma on muuten hetkellinen. Yritä uudelleen tai käy Twitterin tilasivulla saadaksesi lisätietoja.