Bottom-up approach towards AI isn't working since it's inception. We need a top-down approach and sincere efforts in this regard by grad students.
What makes AI difficult is that the traditional science/engineering approach of "divide and conquer" doesn't work: there are too many parts, and they're too tightly bound together. We need a different approach.
-
-
Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
-
-
are there any other viable approaches? Because I'd argue the divide and conquer method is too engrained with how humans/scientists tackle problems and view the world, or not?
Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
-
-
To provoke or create curiosity.
Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
-
-
D and C is not the approach in engineering. On the contrary, engineers build integrated systems. Its the opposite approach of science, which relies on analysis. The problem of AI is the lack of engineers.
-
Take e g. control engineering vs. reinforcement learning. Computer scientists working in RL often ignore years of achievements and sucessful techniques developed by control engineers (which equip from our air-conditioners to our cars).
Keskustelun loppu
Uusi keskustelu -
-
-
Similar to the problem it the natural sciences where people behave like it was geometry. It's all about the architecture. And that is NOT like normal in IT. It must have hierarchies achievable but they must arise spontaneously. Luckily we've got it: GWT.
Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
-
-
I am guessing we augment the brain first and let the brighter minds figure it out.
Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
-
-
We need a top-down, holistic approach
Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
-
-
I rather think that it's the process, it's more interesting to try the process of biomimetism.
Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
-
-
There are mathematical fields that literally study systems made of many parts that are tightly bound together: complex systems and dynamical systems have a lot to give to ML and AI in general.
-
They've been involved in ML from day one, but so far there isn't much to show. Hopefully this will change in the future.
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.