@trurl@mastodon.sdf.org avatar

trurl

@trurl@mastodon.sdf.org

Also @naturaltonic on Twitter.

This profile is from a federated server and may be incomplete. Browse more on the original instance.

weilawei, to random
@weilawei@mastodon.online avatar

@ned ๐Ÿ˜ณ

Wow. Question: in Canada, are bicycles required to broadly adhere to the same laws as motor vehicles, or do they get totally separate treatment?

(I ask, because that's illegal here in the US to be in the road at a light and cross when it goes red.)

trurl,
@trurl@mastodon.sdf.org avatar

@weilawei @ned in the USA, the legality of riding a bike through a tree light depends on what state you're in: https://en.wikipedia.org/wiki/Idaho_stop

mekkaokereke, to random
@mekkaokereke@hachyderm.io avatar

Rest in Power Shafiqah Hudson.
https://www.netrootsnation.org/profile/shafiqah-hudson/

Shafiqah, known online as SassyCrass, was a brilliant teacher, writer, and Black feminist that we owe so much to.

We talk a lot about how online disinfo campaigns from Gamergate to the Alt-right to Moscow's Internet Research Agency, target the Black community. But the Black community doesn't fall for it. "Famous security researchers" often pretend that they first found these disinfo networks, but they didn't.

https://slate.com/technology/2019/04/black-feminists-alt-right-twitter-gamergate.html

trurl,
@trurl@mastodon.sdf.org avatar

@D_J_Nathanson @mekkaokereke The "Testifying While Black" paper (PDF here: https://www.linguisticsociety.org/sites/default/files/LSA952102.pdf) is straightforward to read and downright infuriating.

It should also be required reading for anyone who builds any tech that touches language. A software engineer who is capable of wading through mathematical notation they don't quite understand in ML papers should be able to do the same with IPA.

alcinnz, to random
@alcinnz@floss.social avatar

There's numerous ways to achieve the task of a "computer mouse", with the design I described yesterday being quite outdated. So what are the common designs today?

Most mice (or wireless "hamsters", lol!) shine a laser light askew at a surface to reveal any roughness it might have. An image sensor rapidly compares the last 2 frames to determine how far the previous has to be shifted to match the next.

Involves a lot of very clever arithmetic, & a "digital signal processor" to run it on!

1/5?

trurl,
@trurl@mastodon.sdf.org avatar

@alcinnz typical touch interface inputs are still extremely noisy, even after signal processing in the hardware and OS driver. See https://github.com/google/ink-stroke-modeler for an example of a software-based solution to smoothed, low-latency drawing with a stylus.

johncarlosbaez, to random
@johncarlosbaez@mathstodon.xyz avatar

No. Succeeding in love is not easy, and there's no formula for it.

Here's the essence of this bad take:

"Since life is itself simply a game in disguise, having a few mathematical tricks up your sleeve can also give you an edge in the game of life."

Life is not simply a game in disguise. There are no fixed rules, apart from possibly the laws of physics. More importantly, there's no fixed definition of what counts as "winning". In fact the whole concept of "winning" doesn't apply, except in very limited realms.

I'm reminded of an anecdote I heard from the statistician Persi Diaconis. I'll probably get the details wrong, but it goes something like this:

Persi Diaconis was friends with an economist who had just gotten two job offers, one on the east coast of the US and one on the west coast. The economist was having a lot of trouble deciding which offer to take: each had its pros and cons. So Diaconis said "Hey, why not use the mathematics you're always talking about? Compute the expected utility in each case, and pick the offer that maximizes it!"

And the economist said "Come on, Persi! This is SERIOUS!"

.....

The article is here:

https://www.theguardian.com/science/2023/nov/05/how-maths-can-help-you-win-at-everything

and I thank @pigworker for pointing it out.

trurl,
@trurl@mastodon.sdf.org avatar

@johncarlosbaez @dduque on the bright side, a deeper understanding of mathematics has helped me with "understanding life," in at least some small ways, in that it has equipped me to look for structures and patterns which I might not otherwise recognize, or in situations where I might otherwise use the wrong models/metaphors.

But the map is not the territory, and we can't blindly apply metaphors rooted in mathematics without acknowledging their assumptions and blind spots.

foone, to random
@foone@digipres.club avatar

It'd be unlikely to have much use outside the one specific situation I find myself in, but it'd be interesting to have a variant on a one-click hoster that works by setting up a temporary one-off SFTP account. Like, you sign up to provide a file named X, and you can securely upload a thing and it gets stored in a server, which won't let you download X back.
Someone else creates an account to download the same-named file, and it gives them a similar temporary SFTP account

trurl,
@trurl@mastodon.sdf.org avatar

@foone can the downloader retry if they disconnect before it finishes?

Wish I had an afternoon to hack this together.

Sender: ssh -i upload.key service@host file <somefile
(server: stash file)
Receiver: ssh -i claim.key service@host file >download.key
(server: move file, generate a keypair, update authorized_keys, print private key to stdout)
Receiver: scp -i download.key service@host:file .
(??? need to reap file when done)

Horribly insecure idea but fun to imagine trying.

mekkaokereke, to random
@mekkaokereke@hachyderm.io avatar

Yesterday I spoke to an MIT sophomore hoping to get an interview for a Google internship. The student has already done one summer internship at another company, enters programming competitions, and has coursework in machine learning, distributed systems, and neuroscience. So far they have submitted over 40 internship applications.๐Ÿคฏ

From all their applications, they've only gotten two OAs (online assessments), and no interviews.

It is so much harder for kids today than when we were young.

trurl,
@trurl@mastodon.sdf.org avatar

@mekkaokereke @taatm it may depend on which country you're talking about, though. I heard a bit of a fuss from Mexican interns in recent years about pay disparities.

kierkegaank, to random
@kierkegaank@mastodon.sdf.org avatar

The wheel of time (pilot)โ€ฆ oh god, iโ€™m holding tolkien responsibleโ€ฆ and whatever soap opera and dungeons and dragons clone that taught the creators of this how to write dialogue and visualize a scene

trurl,
@trurl@mastodon.sdf.org avatar

@kierkegaank you're not accidentally watching the 22-minute TV pilot from 2015, right?

foone, to random
@foone@digipres.club avatar

hmm. I may solve a minor calculus problem by instead using multithreaded python very slowly

trurl,
@trurl@mastodon.sdf.org avatar

@foone ... dump it into Octave? (Because if there's one thing that's better than hacking together a numerical solution on a platform you know well, it's paging through HTML-ized info pages for a language you might never use again.)

emilymbender, to random
@emilymbender@dair-community.social avatar

When multiple scientists independently come up with the same or similar ideas, we call it convergent discovery. What do we call it when multiple scientists come up with the same or similar bullshit?

Screencap of arXiv abstract: Exploring the Use of Large Language Models for Reference-Free Text Quality Evaluation: A Preliminary Empirical Study Yi Chen, Rui Wang, Haiyun Jiang, Shuming Shi, Ruifeng Xu Evaluating the quality of generated text is a challenging task in natural language processing. This difficulty arises from the inherent complexity and diversity of text. Recently, OpenAI's ChatGPT, a powerful large language model (LLM), has garnered significant attention due to its impressive performance in various tasks. Therefore, we present this report to investigate the effectiveness of LLMs, especially ChatGPT, and explore ways to optimize their use in assessing text quality. We compared three kinds of reference-free evaluation methods based on ChatGPT or similar LLMs. The experimental results prove that ChatGPT is capable to evaluate text quality effectively from various perspectives without reference and demonstrates superior performance than most existing automatic metrics. In particular, the Explicit Score, which utilizes ChatGPT to generate a numeric score measuring text quality, is the most effective and reliable method among the three exploited approaches. However, directly comparing the quality of two texts using ChatGPT may lead to suboptimal results. We hope this report will provide valuable insights into selecting appropriate methods for evaluating text quality with LLMs such as ChatGPT.
Screencap of arXiv abstract: ChatGPT Outperforms Crowd-Workers for Text-Annotation Tasks Fabrizio Gilardi, Meysam Alizadeh, Maรซl Kubli Many NLP applications require manual data annotations for a variety of tasks, notably to train classifiers or evaluate the performance of unsupervised models. Depending on the size and degree of complexity, the tasks may be conducted by crowd-workers on platforms such as MTurk as well as trained annotators, such as research assistants. Using a sample of 2,382 tweets, we demonstrate that ChatGPT outperforms crowd-workers for several annotation tasks, including relevance, stance, topics, and frames detection. Specifically, the zero-shot accuracy of ChatGPT exceeds that of crowd-workers for four out of five tasks, while ChatGPT's intercoder agreement exceeds that of both crowd-workers and trained annotators for all tasks. Moreover, the per-annotation cost of ChatGPT is less than $0.003 -- about twenty times cheaper than MTurk. These results show the potential of large language models to drastically increase the efficiency of text classification.

trurl,
@trurl@mastodon.sdf.org avatar

@emilymbender It's probably Baader-Meinhof, but I've heard "zero-shot" thrown around very casually lately (as though it's a given that out-of-class data will be handled appropriately). Has it come to mean "we are going to throw an LLM at it and claim it does a good job"?

tess, to random
@tess@mastodon.social avatar

I posted a rant in someone else's mentions which I regret and then deleted but I need to vent.

I use git every day. I'm very comfortable with git.

Also: git is fucking terrible, and I suspect it's the reason a lot of people get scared away from writing software.

trurl,
@trurl@mastodon.sdf.org avatar

@tess 100% agreement about git not being the right thing to hand to people who are new to source control.

When I was a TA, the professor wanted students to turn in assignments with a "real" tool. I opted for SVN over git.

Students still did silly things when committing to the course repository, but I wasn't about to tell them to pick up a tool whose handles consisted almost entirely of sharp edges. The course was "make Roombas play soccer," not "how to lose your source code like the pros."

jessica, to random
@jessica@mastodon.heavymusic.rocks avatar

Tag yourself Iโ€™m the lone sixteenth note ending a measure

trurl,
@trurl@mastodon.sdf.org avatar

@jessica "go, then. there are other clefs than these."

foone, to random
@foone@digipres.club avatar

wow the autodesk fusion 360 API docs are interesting. it seems like they're trying to see how much documentation they can write without possibly being useful

trurl,
@trurl@mastodon.sdf.org avatar

@foone sounds downright septic to me

trurl,
@trurl@mastodon.sdf.org avatar

@foone could it have been generated from UML or some XML schema?

trurl, to emacs
@trurl@mastodon.sdf.org avatar

Looking around a little bit to see if I can find a replacement for shared shopping lists in Google Keep.

Anyone in the community ever try to share org files over WebDAV between and ? I'm already serving SVN-revisioned org files via WebDAV with Orgzly, which works pretty well for me, but I have family members on iOS.

trurl, to random
@trurl@mastodon.sdf.org avatar

Malapropism on a work mailing list today brought me a new phrase: programming language echo system.

I shall use this to mean hype that boosts a programming language. For example:

While the tooling around Java may have been weak when it was introduced in 1995, its patrons were wise to invest in a strong language echo system.

mekkaokereke, to random
@mekkaokereke@hachyderm.io avatar

๐Ÿคฏ This year in social media is wild. The latest chapter is no exception. Fascinating data and anecdotes.

Data:

  • Over 430K Fediverse accounts created in the past week. And it's accelerating.

https://mastodon.social/@mastodonusercount/110576430285028932

Anecdotes:
Major communities may be moving soon.

https://kbin.social/m/RedditMigration/t/59559/Removed-as-moderator-of-r-Celebrities-after-14-years

Here's the Fediverse version of "celebrities" with exactly 1 subscriber so far... the person who claimed to be the former moderator of the celebrities sub-reddit.
https://kbin.social/m/celebrities

trurl,
@trurl@mastodon.sdf.org avatar

@mekkaokereke @justinmwhitaker as a point of reference, @SDF has provided services on the Internet longer than the Web has been around and uses this membership model extensively.

Green_Footballs, (edited ) to random
@Green_Footballs@mastodon.social avatar

Web pages are full of rectangles, often with borders.

Which border style do you prefer?

trurl,
@trurl@mastodon.sdf.org avatar

@Green_Footballs I've idly wondered if you could quantify the time and energy lost over the years from drawing rounded corners and drop shadows everywhere just because Apple did it or something.

fribbledom, to random
@fribbledom@mastodon.social avatar

Everything is shit until you try to improve it yourself.

trurl,
@trurl@mastodon.sdf.org avatar

@fribbledom "I think I like the speckled ax best."

gregeganSF, to random
@gregeganSF@mathstodon.xyz avatar

Tesla announces that โ€œphantom stopsโ€ are due to the carโ€™s sensors detecting dark matter vehicles and pedestrians. โ€œInstead of complaining, just give us the Nobel Prize!โ€

https://www.theverge.com/2023/5/25/23737972/tesla-whistleblower-leak-fsd-complaints-self-driving

trurl,
@trurl@mastodon.sdf.org avatar

@gregeganSF "our advanced quantum chips occasionally interact with traffic from other world-lines. Patent pending."

(But Pratchett clearly has prior art in the Dis-organiser.)

  • All
  • Subscribed
  • Moderated
  • Favorites
  • โ€ข
  • provamag3
  • thenastyranch
  • magazineikmin
  • InstantRegret
  • GTA5RPClips
  • ethstaker
  • Youngstown
  • everett
  • slotface
  • osvaldo12
  • rosin
  • mdbf
  • kavyap
  • DreamBathrooms
  • megavids
  • ngwrru68w68
  • Durango
  • modclub
  • cubers
  • khanakhh
  • Leos
  • tacticalgear
  • cisconetworking
  • vwfavf
  • tester
  • anitta
  • normalnudes
  • JUstTest
  • All magazines