Very Interested in HTM. What should I learn?

question
community

#1

I’m 18 and am very interested in HTM.Should I pursue a degree in Computer Science or Neuroscience?
Any other advice is also appreciated! :slight_smile:


#2

We sometimes say that we are at the intersection of computer science and neuroscience. So either direction would be valid. If you plan on implementing HTM systems in software, especially creating production applications, you need background in software development. If you are more interested in theorizing how intelligence works in the brain, the neuroscience background is very important.


#3

Hi Niszoig, you are at a very important milestone in your life, the great launching point for the development of your dreams. Very exciting and turbulant times, indeed. After graduating college 4 years ago I’ve mentored high school robotics students and they’ve asked me similar career questions. Indeed I asked the same questions to my mentors way back when.

I invite you to consider an idea that may be of use to you. The best, most crucial piece of wisdom I can give to you is this: always go to virtues first. Ask yourself “who am I becoming?” Are you courageous to take risks to move yourself in the direction of your dreams? Are you dedicated to your craft and willing to accept failure to learn from it and grow stronger because of it? Are you willing to spend time learning how to learn, to “sharpen your axe before chopping down the tree”? There are many other virtues(love, sacrifice, discipline, etc), but my point is character building is foundational and will apply to every aspect of your life. Ultimately a person of strong character will thrive in any environment and in any task. Understanding this will help you be a stronger person.

Now, I will get back on topic. A fundamental question you need to ask yourself when establishing your vision of the future is “what do I want to do?”. Are you interested in developing algorithms or understanding the brain? Do you like sitting in front of a computer and solving problems? Or do you prefer researching ideas, forming a hypothesis about how something works, and thinking of potential experiments on how to verify your ideas? You need to figure out what you enjoy and could see yourself doing in a career. Spend some time, an hour or two, writing down and clearly defining what you enjoy.

In the end, pursuing an understanding of HTM will require knowledge of both compsci and neurosci. It doesn’t matter which you choose as your degree, as long as you pursue both in some way. Get comfortable supplementing your learning by doing your own research outside of class. The internet has LOTS of knowledge ripe for the taking if your dedicated enough to learn it. Do some hobby projects and start building a portfolio. Also, don’t be afraid to go to advisers and professors for help. You will find valuable mentors who are willing to give you more specific advice.

Good luck!


#4

@ddigiorg Thank you for your valuable advice! (I’m sure your advice is going to help develop my frontal lobe faster :stuck_out_tongue: )
I think I’ll stick with Neuroscience.I can learn the software part from the Internet as you mentioned. I would prefer forming hypothesis over coding all day.


#5

Yeah its definitely be useful to code too, so that you can test your hypothesis. That way you can engage in the research/development cycle.


#6

I have an excellent resource for you. It’s a free, at your own pace course from Udacity where you’ll learn Python, probably the best language for coding quick scripts to test ideas and model your ideas. You don’t have to install anything, all the coding practice sessions are done in the browser and the lessons are taught via video. I took some helpful Machine Learning courses here before switching my focus to Machine Intelligence. Check it out:


#7

Thanks! Already got started with Python and a Git tutorial by thenewboston.As I am already quite familiar with OOP concepts and Java, I am finding it quite easy :). I’m confident that I’ll be able to understand NuPic for python in a couple of weeks!