Hello New Blog!

Moving to a new place can be hectic and tiresome. I am moving my blog, from here, and it’s none of those.¹ /s

I tend towards writing technical posts when I tend towards writing at all these days, and blogger doesn’t give me the presentation options I need. So, for now, its GitHub pages, but with my own domain. That way, if I decide to move again, my (almost non-existent) readers won’t be sent scrambling to find my (almost non-existent) content.

The old blog was titled “Random Thoughts”. I wanted something different and bit more original this time, so I Googled “Not So Random Thoughts”. Obviously.

So many hits it isn’t even funny. So many, that you couldn’t squint and ignore. And that is exactly why you are stuck with “A Not So Primordial Soup”; which, by the way, does a good job of telling you that this is going to be a mixed bag of the deep and the frivolous.

Just for the record, “The Psionic Poodle” was on the list. Since that isn’t the title, joy to us, things could have been worse.

I have been asked about the domain name “quipu strands” (by the 3 and a 1/2 readers I have) . These were a device used by the Incas to record information, and well, the blog is like my recording device. Hence the name. You can read more about them on Wikipedia. And this (Locke, 1912) is probably one of the earliest articles detailing their use.

Now for some quick tests to see if this platform holds up. Grouping them here, so that if I switch to a new platform/theme etc, I can use this post as a quick check to see if the various non-text entities I use are displayed correctly. My regression tests, if you will.

• Math. Trust me this is going to come up.

  \[\begin{align} &\text{Cauchy-Schwarz: }||x+y||^2 \leq (||x||+||y||)^2 \\ &\text{Holder: }\sum_{n=1}^{\infty}\frac{(a_1 + a_2 + ... + a_n)^p}{n} < \Bigg(\frac{p}{p-1}\Bigg)^p \sum_{n=1}^{\infty} a_n^p \end{align}\]

• Images. Of course.

From a trip to Yosemite. Wow, captioning in Kramdown isn't straightforward!

• Code. You bet.

  def dot_product(v1, v2):
    # v1, v2 are lists
    result = None
    if len(v1) == len(v2):
        result = sum(map(lambda x: x[0] * x[1], zip(v1, v2)))
    return result
  

• Citations. Here are some papers I co-authored (Ghose & Ravindran, 2020; Ghose & Ravindran, 2019) and a software I maintain (Ghose, 2021).

• Tables. Basic. Just-in-case-I-need-them-but-I-really-don’t-think-I-need-them.

  Default aligned	Left aligned	Center aligned	Right aligned
  First cell	Second cell	Third cell	fourth cell

During my research to move, I discovered, quite to my surprise, that having a reliable comment system is something of a challenge. In 2017. I mean in a few years I could be sitting in my self-driving car (Update in 2022: yeah, this dated well.) and worrying about comment systems.

More updates from 2022: When I had initially moved to GitHub pages, commenting systems left much to be desired. I did have Talkyard on here for a while, which is a nifty system. But now we have options like utterances and giscus, which feel like a more natural fit for something that’s hosted on GitHub. I am going to integrate one of them soon.

See you later then?

References

1. Locke, L. L. (1912). The Ancient Quipu, a Peruvian Knot Record. American Anthropologist, 14(2), 325–332. http://www.jstor.org/stable/659935
2. Ghose, A., & Ravindran, B. (2020). Interpretability With Accurate Small Models. Frontiers in Artificial Intelligence, 3, 3. https://doi.org/10.3389/frai.2020.00003 PDF
3. Ghose, A., & Ravindran, B. (2019). Learning Interpretable Models Using an Oracle. CoRR, abs/1906.06852. http://arxiv.org/abs/1906.06852 PDF
4. Ghose, A. (2021). compactem: build accurate small models (Version 0.9.6) [Computer Software]. https://compactem.readthedocs.io