Show newer

@grimalkina On the hiring front, when I've been involved in making hiring decisions my issue has been that I recognize really great contributions can come from people who are self-taught or have less conventional backgrounds, but what isn't clear is how to recognize those people. Part of me thinks "just spend time with them and talk to them and you'll figure it out," but that also seems like the sort of highly subjective gut judgement where unconscious bias looms large, so I fear it might do more harm than good.

It is ironic that in the hiring where I've been involved, I'm often the only one in the discussion with a PhD, but I often find myself arguing not to weight education too strongly. My feeling is just that if you are trying to find someone to do task X, having actually done task X before is better evidence of capability than having been in a program that nominally teaches how to do X (especially if X isn't domain-specific research).

@bogosity @danilo

@grimalkina Oh yeah, certainly if one is subject to prevailing biases that would question one's competence, then any formal recognition of expertise is more valuable. I guess I hadn't fully considered that beyond the more typically discussed dimensions of privilege there's also the question of whether your degree is in something considered to be "mathy" or "technical" nor that some of my experience with coding specifically may be highly historically contingent based on when I was learning it. @bogosity @danilo

@bogosity That was my reaction as well. I both came of age at a time where a large number of coders were self taught (due to the personal computing revolution followed by the WWW revolution) and worked around physicists, electrical engineers, and astronomers who generally seemed to assume they could teach themselves to code and didn't need instruction (for good or ill...definitely ill in some cases 😄 ). It hadn't really occurred to me that times had likely changed and also norms might be different in other scientific disciplines. @grimalkina @danilo

Nick boosted

Please subscribe to my newsletter! Mondays are free! Tuesdays and Thursdays will cost you, but just a little bit, though, so I can, y'know, eat.

badastronomy.beehiiv.com/subsc

Show thread

@grimalkina Honestly, I'm in awe of that. I can't even conceive of how one can find enough hours in the day. @analog_ashley

Nick boosted

@jared You say that you don't consider Goodfellow's book to be the best one on the topic. If it's also quite expensive, it might be useful to mention what you do think are the best books (as an alternative). @TeaKayB

Nick boosted

Homotopy Type Theory by the Univalent Foundations (2014). The book itself is more an explainer as to what UF participants do. Much of the activity around automated proof assistants revolves around this project.

Deep Learning by Goodfellow, Benigo, and Courville (2016). Some might say it’s a computer science book, and not a mathematics book, but I’d argue the combination of linear algebra, calculus, and statistics just happens to be implementable in silicon. Not the first or even the best book on its topic, yet certainly the standard reference

Poincairé Conjecture website by the Clay Mathematics Institute (2010). They provide links to Perelman’s three papers solving the problem as well as the papers and lectures by experts explaining the proofs and importance. Another great example of how intensely collaborative mathematics can be, even if just one person succeeds in solving a difficult problem.

Four Colors Suffice by Robin Wilson (2002). An expository book, beautifully produced with high quality paper, binding, and print, on the controversial and long solution to the four color problem. While automated provers and a plethora of deep learning techniques have quickly become standard, the world had to get used to having machines “solve” problems. Wilson tells the story of how this problem was solved and then improved upon, and effort among many that allowed mathematicians to accept the usage of machines in their work.

@TeaKayB

Nick boosted

**
What are the most important/influential #Maths #books of the 21st century
**

This is a very loose question, I know. That's intentional: I'd like to see how people interpret things like "important", "influential", and even "maths" and "book".

Given that, an explanation with your chosen book(s) would be most appreciated.

Also: I know that confining it to the last 24 years makes it more difficult as there hasn't been a lot of time for anything published in that period to have much influence at all. Humour me!

If you don't have an answer I'd appreciate a #boost all the same.

Thanks!

Nick boosted

@collectifission @cstross So, I agree with you about the missed opportunity, but I feel the "where's Eugen" dig is very unfair. Mastodon gGmbH is like 4 people and they're all coders, not marketing. The fediverse is an intentionally decentralized collective, not organized like a startup running user-acquisition campaigns.

Nobody executed the highly skilled labor you (and I) wish would have happened because there is literally no one being paid to do that, because Mastodon is not a company,.

Nick boosted

The EXCITE team will share their adventures in scientific ballooning over at @nasaexpeditions at the beginning of October. Keep an eye out for more information!

bird.makeup/@nasawallops/18300

@dgoldsmith Writing "viscous" instead of "vicious", they must be thick! 😉

Nick boosted

You don't have to be racist to participate in systemic racism. US policing produces racist outcomes even from "not-racist" people. I'm going to explain (again):

* why I've probably been pulled over way more times than most people you know, even though most of the times I'm pulled over, I get no ticket (because I did nothing wrong)
* how I stopped getting pulled over so much (because I understand the system)
* And why lots of cops say that the average voter is more racist than the average cop

Nick boosted

The first solar flare observations were made #OTD in 1859 by Carrington and Hodgson, preceding one of the largest geomagnetic storms on record.

Both amateur astronomers made their observations independently, and published them simultaneously in MNRAS.
articles.adsabs.harvard.edu/cg

Nick boosted
Nick boosted

This graph guessing game is pretty interesting! You're shown a graph, and you have to guess (from five options) what is being displayed. A good tool for statistical literacy in math and statistics classrooms!

graphs.world/

#Math #MathEd #Statistics

Nick boosted

Your SSH honeypot fakes a Linux system and logs the threat actor's commands.

My SSH honeypot hijacks the threat actor's terminal to play the music video of Rick Astley's 1987 pop hit "Never Gonna Give You Up" while ignoring Ctrl-C.

We are not the same.

Nick boosted

Check out this link to follow nearly 100 journalists working on climate, energy, and the environment.

go.bsky.app/2TNd1Ya

Nick boosted

Yesterday I listened to the first episode of the "Change, Technically" by @grimalkina and @analog_ashley and (unsurprisingly) really liked it. I think it had a lot of interesting things to say about how we think about who can code or do science as illustrated in part by their own paths into these fields. If you're interested in , , , etc. I think it's well worth a listen.

changetechnically.fyi/

@rdviii I was honestly surprised to discover Tim Horton's in Ohio at all when I ran across it earlier this year.

Show older
Qoto Mastodon

QOTO: Question Others to Teach Ourselves
An inclusive, Academic Freedom, instance
All cultures welcome.
Hate speech and harassment strictly forbidden.