Double pendulum prediction - can HTM do this?

[Bitking]

Double Pendulum (pendulum_bot)

It would be amusing to see this fed to HTM instead of the hot-gym dataset.

[mthiboust]

It doesn’t sound easy

If someone makes it work, it would be a very high achievement for HTM!

[AMZ]

Yeah… I saw HTM being used for the double pendulum prediction when I was at Nice workshop 2018.
The presentation was by a guy from IBM, his name is Campbell Scott. You can find more details about the workshop here: https://niceworkshop.org/nice-2018-agenda/

[Falco]

That’s quite amazing. I didn’t know IBM is using Numenta’s technology.

In his presentation Campbell Scott says they worked together with Numenta until they parted ways. Does anyone know why?

[Bitking]

IBM's Context Aware Learning

Thanks Sergey. We used to have a working relationship with the Dr. Wilke’s team at IBM, but they don’t keep in contact much with us anymore because of corporate blowback from this blog post where we called Watson “classic AI”. They continue to work with HTM, but they don’t call it that anymore.

[Falco]

Yep, I just found that thread too, by googling Campbell Scott.
XD

(thx)

[fcred]

The IBM team have published a paper recently in the arXiv: Simulation of neural function in an artificial Hebbian network

[rhyolight]

@Falco

worked together with Numenta until they parted ways . Does anyone know why?

There has been a group at Almaden working with HTM for years, but after we published a blog post calling IBM Watson “Classic AI”, IBM stopped cooperating with us. It was not our choice.