Hacker News: smath

New comment by smath in "Regressive JPEGs"

smath — Sat, 18 Jul 2026 15:38:21 +0000

Is it possible to acquire lower frequency components with a camera too?

This also reminded me of MRI where low frequency is acquired first in a space called k-space

New comment by smath in "The End of Creativity"

smath — Wed, 15 Jul 2026 22:39:37 +0000

My conclusion is that this in not the end of creativity but an increasing premium on creativity that is certified to be human made. It’s like how if I cook something from scratch that’s considered more meaningful than if I simply heated frozen food or ordered in for a guest.

New comment by smath in "Silent speech with ultrasound"

smath — Sat, 11 Jul 2026 14:52:23 +0000

Very cool! While this is a different type of inverse problem, it reminds me of a radiolab story [0] about a device to help blind people “see” using electrical impulses on a metal strip under the tongue.

[0] https://radiolab.org/podcast/seeing-tongues

New comment by smath in "Mir Books – Books from the Soviet Era"

smath — Sat, 04 Jul 2026 10:58:34 +0000

Very fond memories of Mir Publishers' science and math books growing up in India in the 80s and 90s. I think English translations were freely available in India. My grandfather would buy them for me to encourage my interest in science and math. If I could find physical copies of those books I would buy them in a heartbeat today.

Slight digression: Russian cartoons from that era are also very interesting. One of my favorite short cartoon from that era (I still hum its music involuntarily): Ikarus and the Wise Men [0]

[0] https://www.youtube.com/watch?v=0Yk1mz23YFA

New comment by smath in "From brain waves to words: a new path to communication without surgery"

smath — Wed, 01 Jul 2026 02:37:54 +0000

Are they trying to infer characters/words from brain waves? I would have thought the brain is thinking in concepts rather than actual words

New comment by smath in "AI learns the “dark art” of RFIC design"

smath — Sun, 28 Jun 2026 02:21:11 +0000

GA’s optimize only combinatorial problems though — where you have discrete set of choices (~genes) for each variable, and therefore do not have a gradient

New comment by smath in "Egyptian Fractions (2006)"

smath — Sat, 20 Jun 2026 12:27:23 +0000

Mind blown, I had no idea about this, thx for sharing

New comment by smath in "Zen and the Art of Machine Learning Research"

smath — Sat, 20 Jun 2026 12:17:00 +0000

I believe that is correct. Part of the problem is that similar words in different languages are actually not quite identical. “Not having emotions” is not quite the essence of stoicism: I understand it to be “avoiding highs and lows - being equipoised by recognizing what is and isn’t under our control”. A very closely related idea is proposed in the Bhagawad Gita.

New comment by smath in "Zen and the Art of Machine Learning Research"

smath — Sat, 20 Jun 2026 12:09:04 +0000

Tangential tidbit about etymology of the word Zen:

Zen is a Japanese word that comes from the Chinese “Chan”, which in turn comes from the Sanskrit word “Dhyana”, which roughly translates to focus/meditatiin.

That trajectory Sanskrit —> Chinese —> Japanese reflects the geographical trajectory of the spread of Buddhism out of India.

Same word in Vietnamese and Korean is “Thien” and “Seon” respectively.

New comment by smath in "SpaceX to buy Cursor for $60B"

smath — Tue, 16 Jun 2026 22:18:32 +0000

Right, value != price. Now, is this a bug or a feature? Discuss (serious question)

New comment by smath in "Nobody ever gets credit for fixing problems that never happened (2001) [pdf]"

smath — Fri, 12 Jun 2026 02:36:53 +0000

I'm looking for some data -- if anyone has it -- on the fraction of companies that are led (CEO) by a technical person, over the years/decades. I have the (anecdotal) impression that this fraction has been falling (stories like Boeing), but it would be cool to support or refute this with hard data. Anyone know where to find/assemble something like this? Also, if this trend is true, then why?

New comment by smath in "Mathematicians issue warning as AI rapidly gains ground"

smath — Wed, 03 Jun 2026 11:49:28 +0000

This reminded me of my 11 yr old who, when I give her math problems to solve, is too focused on “getting the right answer”. I’ve told her plainly, I don’t care if you get the right answer right now, I want to see your reasoning. She has yet to understand this.

New comment by smath in "Appearing productive in the workplace"

smath — Wed, 06 May 2026 21:17:25 +0000

Here is a solution to this problem I think: make an LLM. Summarize everything. If there is fluff then it should get dropped? Basically we only care about the relevant information content, regardless of the number of characters used - so we need a compressed representation

New comment by smath in "John Ternus to become Apple CEO"

smath — Tue, 21 Apr 2026 11:56:39 +0000

I for one am happy to see an engineer at the helm.

New comment by smath in "Introduction to spherical harmonics for graphics programmers"

smath — Thu, 16 Apr 2026 22:06:47 +0000

Loved the writeup.

I'll just drop a note here to say that these spherical harmonics are also used in creating specialized neural network layers that are useful for modeling 3D objects like point clouds and moleculues, proteins, etc. Basically whenever we want to make sure that rotating / translating the object doesnt make a new object. [0] is a good reference for this.

Even more interesting is that these are the same spherical harmonics that appear as solutions to Schrodinger's equation in quantum mechanics (s, p, d, f orbitals in an atom) [1]

[0] https://arxiv.org/abs/2312.07511. [1] https://en.wikipedia.org/wiki/Atomic_orbital

New comment by smath in "Having Kids (2019)"

smath — Fri, 20 Mar 2026 18:34:13 +0000

My 2c is that it is not 'joy' or 'happiness' that kids bring to parents universally (although they might bring those things for some parents), but 'meaning'. Meaning is harder to define than joy/happiness, perhaps because it is less objective and more subjective.

New comment by smath in "Nearby peer discovery without GPS using environmental fingerprints"

smath — Tue, 25 Nov 2025 15:47:08 +0000

Shameless plug: More than a decade ago, I wrote a paper [1] on how the random perturbations in the wireless channel between an ambient RF transmitter (FM radio, TV) to the two devices, allow nearby devices to authenticate locality because the perturbations are correlated only if they are nearby (where nearby is relative to the wavelength being monitored)

[1] https://dspace.mit.edu/handle/1721.1/121443

New comment by smath in "The history of Casio watches"

smath — Wed, 12 Nov 2025 14:45:15 +0000

As a kid I loved casio digital watches (metal band, digital display). But as a grown up I found I like analog watches better -- my brain is quicker at interpreting the visual image of the hands. For the last few years I wore this [1] very simple and robust casio watch and eventually gave it to my son to help him learn to tell time. Very clean crisp design and 1/3rd the price of a similar looking swatch.

[1] https://www.casio.com/us/watches/casio/product.MQ-24-7BLL/?u...

New comment by smath in "Eye prosthesis is the first to restore sight lost to macular degeneration"

smath — Wed, 29 Oct 2025 16:46:37 +0000

Very very cool. I have this condition - I got it randomly ("idiopathic" as opposed to age-related) when I was 22. At the time it wreaked havoc on my mental health.

New comment by smath in "Synthetic aperture radar autofocus and calibration"

smath — Sat, 11 Oct 2025 19:12:16 +0000

A benevolent monarchy maybe - like some places in the east (maybe).or maybe UBI? Some way to not have to worry about basic health and needs

What places have this today? I see an answer suggesting AUS below. ChatGPT says Switzerland