On Learning…

I have learnt mostly everything I know about Quantum Computing in the past 7 years. In fact, I’m going to make a little confession: I have never taken a formal course on Quantum Computing, I am almost completely self-taught 🙂 And if there is one thing I’ve learnt, it is that to learn hard subjects on your own you need to learn how to learn. I’m pretty obsessed with learning. It is one of the things that defines me.

So, as I embark upon a new learning task, I start to wonder if there is anything I can ‘learn’ (hehe) from my previous experience. To be clear about my goal: I would like to formally learn more about the field of Artificial Intelligence.

So I thought I’d write down a few things that might help and hinder me in this task, and to try and extract just how comparable the two learning experiences might be.

.

1.) I know a lot more about learning now. I am fairly good at knowing where to get information, how to mine resources and when to give up and go onto something simpler first if you’re just not getting anywhere. So this will all help.

2.) I probably have about the same amount of time to do the learning for AI as I had for QC, in both cases the subjects are tangential to what I am/was doing at the time. Not central, but related.

3.) I am generally more focused than I was 7 years ago! I also pretty much don’t drink alcohol anymore, and I’m probably in better health overall, which will help towards the goal. So this is all positive.

4.) However I *am* older – perhaps I am losing some of my brain plasticity!

5.) I think the core conceptual framework of QC is more tightly knit, whereas with AI it is a bit more sprawling, however to counter this I’d say that the mechanics/mathematics of QC is probably more involved. I may be wrong here of course!

6.) It is easier to ‘start’ with AI, I think there are simple programs, simulations and experiments that you can work on DIY style to learn a lot (I know that I learn quite well that way)

7.) There is a crossover between QC and AI – which makes the learning more interesting, as you can apply one area of expertise to another and see if you can come up with anything interesting along the way.

8.) I am now much better at networking, talking to people and conferencing – and you learn a LOT through these techniques. It is so important to learn what has already been done so you don’t end up spending several years re-inventing the proverbial wheel…

9.) Technologies are accelerating. I can utilise the latest developments to help me learn a new topic. 7 years ago I wasn’t using Wikipedia, ArXiv, Google Books etc, etc…. and I can now start to organise my thoughts more intuitively using wikis and mindmaps. I have an almost constant connection to the internet at all times now – so not only do I have a much greater access to learning resources, but I also feel in a much more information-aware mindset.

10.) Black Swans – of course something crazy might happen, like, say (lets be optimistic here) we suddenly find we’ve gone through the Singularity… then I won’t have to worry quite so much about the learning anymore…

What will happen? Will it work? 7 years?

Let’s see…..

Advertisements

6 thoughts on “On Learning…

  1. Albert says:

    How it is that anything so remarkable as a state of consciousness comes about as the result of irritating nervous tissue, is just as unaccountable as the appearance of Djin when Aladdin rubbed his lamp in the story.

    Thomas Henry Huxley

    … so, AGI (mainly for the artificial consciousness) is extremely more complex than QC.

  2. physicsandcake says:

    It would be interesting to quantify what you mean by ‘remarkable’. Why is consciousness remarkable? It is a vague definition most likely to be explained as an emergent property of a complex dynamic system. But one that as humans we seem to irrationally revere. It’s like worshipping phase locking of the moon’s orbit (which by the way is quite fun and involves blood sacrifices).

    However, we understand the physics of complex dynamic systems much better than we understand quantum mechanics…. so I very much doubt that it is more ‘complex’ than QC, where I will arbitrarily define complex as our ability to realize something experimentally given large computational resources.

  3. Albert says:

    “emergent property of a complex dynamic system”
    “we understand the physics of complex dynamic systems much better than we understand quantum mechanics…. so I very much doubt that it is more ‘complex’ than QC”

    But this is related to the “easy problem of the consciousness”

    The real problem is related to the “hard problem of consciousness”
    http://en.wikipedia.org/wiki/Hard_problem_of_consciousness
    http://en.wikipedia.org/wiki/Philosophical_zombie
    This is the reason because I think that AGI is extremely more complex than QC.
    For example, what guarantee do we have that a future AGI can feel “real” pain or pleasure? It will be a fundamental ethical question.
    Or, what guarantee do we have that a future “mind uploading” will be a “real” (with qualia) uploading and not only a “functional” (behaviour) uploading?

  4. null says:

    Could you clarify number 10. You would not learn AI because the AI problem has been solved and can only be improved by the AI itself (for that matter any scientific endeavor). Or the Singularity will render learning absolute and learning any subject will be delegated to the AI. We will be able to share the learning being intimately connected to the AI. Or I do not enjoy learning it is rather means to an end and after the singularity any leaning will not have an ‘end’. I for one do not find any of those reasons quite satisfying.

    • physicsandcake says:

      Ah, OK. By this I meant that humans may be able to augment themselves to such an extent that low bandwidth ways in which we learn now would be obsolete (for example, reading and talking would probably not be so useful if we had direct brain-computer-interfaces). So the idea behind number 10 is that you wouldn’t be able to compare learning two different subjects because the methods available to you for learning them have changed so much in the intervening time 🙂

  5. David Lemus says:

    I’ve thought about this subject also: how one who loves learning learns and how to continue on learning, but also realizing that you cannot know everything, especially on some domains. I was reading your top ten ways you approach learning, and thought that everyone has their own specific way of learning, however, the internet has provided us with so many ways to actually do so, besides lectures. Which makes me glad to live on the internet age. And,then, when I came to your last one, Black Swan, I absolutely, positively, concur, that the unknown is more intriguing and vexing than everything one can learn on any subject. Of course if you are talking about the Black Swan I had in mind. I just read your clarification to null, and perhaps I’m on different subject, nonetheless I’ll summit my comment. Great website, will definitely continue reading it.: D

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s