Matt's 2018 Goals


#1

:confetti_ball: Happy New Year!!! :confetti_ball:

I like to plan out my goals every January, and since I serve this community, you should know what I’m up to. If anyone has any ideas on how I could improve, I’m open to suggestions and discussion.

Here is a brainstorm of my professional goals for 2018:

Each branch/leaf in this tree is either an action that needs to be taken or a thing that needs to be maintained or fostered in some way.

Most of my time is going into “Finish HTM School Layers and Columns” this year. Including the one I’m working on now, I think there may be 3 episodes here. I’m adding a lot of custom animation to explain, so it is taking longer than expected. But I think it will be worth the wait.

I also want to find some talk venues to submit to. If anyone has ideas, please share.

Note: Highbrow is an HTM visualization library that @lscheinkman is working on. It converts HTM cellular states within columns / layers into generic 3D representations for visualization. I need to use it for an object recognition demo I want to put in HTM School.


Also, thank you, everyone who took time out of their lives to read strangers’ questions on an internet forum and help them get something working. It makes a difference.


#2

I believe it would be great to add one more project to your plan: weekly podcast. You could make it in interview format online, so it would be much easier to arrange such conversations comparing to your YouTube interview project (which you can change a bit to have a podcast version of them in addition to YouTube).
There are some great podcasts about AI and related topics, but unfortunately, there is nothing at the junction of AI and neuroscience. I’m sure there are a lot of people who will include such one to their list of subscriptions.


#3

A post was split to a new topic: Object Recognition Touch Demo


#4

I’ve thought about a podcast, but I just don’t think I have enough to talk about each week. I’d just be reviewing the conversations I’ve had online with you all.


#5

It could have several sub-topics, like:

  • interview with neuroscience
  • interview with people in mainstream AI who have some connections to cognitive science (like studying human attention with AI)
  • other approaches to AI which have biologically plausible background (like models build on grid sells, swarming algorithm, genetic algorithm, etc.)
  • relevant paper reading/discussion
  • examples of HTM implementation discussion
  • discussion of the most popular topics from the Numenta’s forum
  • hardware and computational frameworks/solutions which are promising from the HTM perspective

It’s definitely not the complete list, but an illustration of the potential for the podcast format.


#6

I’m spending my time on video for now. I might podcast in the years ahead. I still see a return from making HTM School videos. You’ve drafted out a nice recipe for a podcast, though. Maybe you want to do it? Or someone else in the community? I’ve done podcasts in the past, they are fun.


#7

I’m not as good an interviewer as you are 0:-)
I hope somebody in the community will raise the flag. I believe it can be a very interesting and useful podcast with a lot of fans.


#8

I’m too late to submit for the AAAI 2017 event, which happens in February. But I might submit for next year: http://aaai.org/. Is this an appropriate event? Has anyone attended?