kwooding

The incoming administration wholesale replaced the contents of whitehouse.gov

Web content from previous administrations, is archived by the National Archive. (Eg https://bidenwhitehouse.archives.gov/)

You can grab the archived copy at https://bidenwhitehouse.archives.gov/oncd/briefing-room/2024...

Of course it’s been solved many times before. This is not a great write up and a terrible headline.

In fact, the students came up with a trigonometric proof of Pythagoras which was not circular: I.e. did not rely on anything derived from the Pythagorean theorem itself, and presented it at a regional AMS meeting.

Better write ups are: https://peopleofcolorintech.com/articles/have-two-black-teen...

Or

https://www.theguardian.com/us-news/2023/mar/24/new-orleans-...

Absolutely. Lots of humans live in a post-truth world. They learn not to trust anything they read or hear. Think totalitarian regimes wi try tight controls over media.

A more painful and pertinent question might be: Can democracy adapt to a post-truth world, and the answer to that I fear is probably no. How can a democracy function if it’s citizenry can’t remain informed?

Hahah. We did, but we were a few years too early for many parts of it. The Global Learning XPRIZE was a $15M challenge to build basically this: tablet software that could let kids without access to schools or teachers teach themselves.

We (Learn Leap Fly) were a semi-finalist in that competition. We bet heavily on AI, but again, were a few years early (Note the constraints of the problem said that the devices were NOT connected to the internet, which makes things still challenging today)

We were also a semi finalist (round 3) in the AI XPRIZE (with the same tech).

We’ve thought about whether to resurrect the core ideas and give it another go. For that we need partners. Education isn’t the most lucrative of fields, it turns out.

That said, anyone who wants to go down this path, send me a note. I’m happy to talk nerdity on this topic all day, be it AI, literacy education, pedagogy or grand challenges!

That’s a weird comment. Were you trying to be sarcastic?

There are 3 major federal parties (Liberals, Conservatives, NDP), who received 33%, 33%, and 18% of the popular vote respectively.

There are two minor federal parties (Greens and the People’s party) who together took around 7% of the popular vote, and

Finally there is the Bloc Québécois. It is a federal party but it exists only in one province. It grabbed around 8% last time.

Of course those percentages aren’t entirely representative, because seats are won in a first-past-the-post style in each riding.

Here’s a book I don’t expect to see on too many lists (and I’m surprised it’s on mine, as I don’t typically like big-property universes like Star Trek): The Final Reflection, John M Ford. It’s an ancient book (by Star Trek standards-1984) set in the original ST:TOS universe, but it did Klingon world-building before that was a thing (it’s not canon, as it did this long before Klingons were fleshed out in any real way). Loved the worldbuilding, loved that it made a then 2-dimensional alien into a rich 3-d society.

Ever since my PhD work (where I worked with some pretty obscure computing architectures), the phrase “Von Neumann Architecture” makes think immediately of D H Lehmer’s quote [1] on working with ENIAC:

> Can we use the high speed computer to do the sieve process? This was a highly parallel machine, before von Neumann spoiled it.

[1]: https://history.computer.org/pioneers/lehmer.html

In fact, in many cases, deposits in provincial Credit Unions in Canada (e.g. Alberta and BC) are 100% guaranteed. (As opposed to the big banks, which are only guaranteed to $100,000). Some provinces hve other rules, but in general, deposits to Canadian Credit Unions have better deposit insurance protection than the big Canadian Banks.

Source: https://www.canada.ca/en/financial-consumer-agency/services/... (and the marketing campaigns of the credit unions)

Exactly what I was looking for!

A quick suggestion, I’d love to see you add the “cyclic sighing” pattern from [1] as it’s recently gotten some amount of attention for it’s effectiveness, and is potentially less common than the others.

[1]: https://www.cell.com/cell-reports-medicine/fulltext/S2666-37...

Not sure what you mean by >20 years to be right. I built and trained a 3-layer back-propagating neural net to do OCR on an Apple 2 in 1989 based on that paper. Admittedly, just the 26 upper case characters. But it clearly worked better than the alternatives.

RocketLab (https://www.rocketlabusa.com/) services a slightly different segment (i.e. not as heavy) but does take things regularly to space (and in the case of Capstone, to the moon!)

Kim Moir has a great blog post on this: https://kimmoir.blog/2021/11/30/my-favourite-books-for-new-m... which includes many of the choices talked about here (and many others)

Kim Moir has a great blog post on this: https://kimmoir.blog/2021/11/30/my-favourite-books-for-new-m... which includes many of the choices talked about below, and a great discussion of why

What is the actual point of a generative text model that emits plausible-sounding scientific technobabble? How this is not just a DDOS against the entire system of peer review?

Honest question: how can this be used for any kind of good? What’s the actual point?

Dr Seuss:

One fish Two fish Red fish Blue fish.

I highly recommend Dr. Kelly Starrett for all things mobility. Thereadystate.com (his website) is my daily source of mobility instruction, both for daily mobility work, and for getting me through many an injury. Reading the Supple Leopard (his book) is useful, but it’s hard to beat the follow-along videos for usefulness.

Plus, he’s a huge Dune nerd, which comes through in occasionally hilarious ways.

The way it’s often been quoted to me: “practice makes permanent.”

Or, put another way, “perfect practice makes perfect.”

It absolutely matters what you practice, or you’re just training your fingers to stumble at the same spot every time, which is exactly what I did for decades before being introduced to deliberate practice.

Note: 2017 article

In any live event, things go wrong. Tearing in to what went well, what didn’t go well, and suggesting process changes for the next event is an important process not only for making future events easier, but also for improving the mental Health of the organizers who tend to focus on the “what went wrong” parts.

So technically, we only hold one if things went wrong.

If we ever held an event where everything went right, then yes, we’d totally call that one an afterparty.

Recently, I was a co-organizer for a conference. The event was almost over, and we were in the process of scheduling our event postmortem when one of our fellow organizers suffered the loss of a close family member.

As you can imagine, we all became instant members of the “we need a better name for this process” club. Thank you, OP for giving us a completely viable alternative. I’m calling it a “disquisition” from here on in.

This is, hands down, the greatest HN comment I have ever read.

I was really, really hoping this was about enabling emacs to be scripted via siri shortcuts. (Though this is super useful, too).

Honestly, just being able to set emacs full/split screen via a siri shortcut (and hence, any Shortcut script) would be unbelievably useful for automating pieces of my workflow.

I use mine as the living room computer. Something I can use with the kids to let them learn about computers and xode. And as a MAME box for playing through the history of video games with them.

I really love having a computer in a shared space. It helps turn “screen time” into “social time”

Nah. This guy was awesome. Of course he can supersede rules beyond his position. That’s “free speech” and tenure, and exactly what Universities have traditionally been about. He’s making a stand based on what he thinks is right. The students are free to do the same, of course, with all the repercussions that entail. In this case, the repercussion was obvious. Don’t play chicken with someone who has nothing to lose.

Astrazeneca was offered in pharmacies, and that part of the rollout was a total mess, I would certainly agree. It was also only a small fraction of the total doses offered, however, and was primarily aimed at individuals who wanted to get a shot earlier than was possible through the provincially administered system (which offered Pfizer, and once Moderna’s logistical issues were sorted out, Moderna)

My experiences with the provincial registration system in Ontario, were the complete opposite of yours. They were very good. Once my demographic became eligible, it was a single form to book an appointment. The actual vaccine clinics were quite efficient, and the followup (if you opted-in) was quick and easy.

The hardest part was waiting. Vaccines supply was the issue, but given Canada’s bargaining position (essentially none?) I’m shocked we got them as quickly as we did.

Did your province not give you an option to report followup symptoms? In Ontario, at the time of shot, we could opt-in to followup symptom reporting. Doing so got you a mail like this, which was followed by a pair of surveys, several months apart:

> The Canadian Immunization Research Network is performing active surveillance for adverse events after immunization against COVID-19. You are being invited to take part because you have been vaccinated against COVID-19 and agreed to be contacted about vaccine studies. The purpose of the study is to see how often health events occur after COVID-19 vaccination. If you might be interested in helping with this surveillance, please visit https://CANVAS-COVID.ca for additional information on how to participate. Your participation is entirely voluntary.

This may have been true at the beginning, but nowadays I use “pip install” extensively in conda-created environments. Are there particular packages that you have trouble with?

This is the main reason I use conda to manage python virtual envs. It easily supports multiple python versions.

I'm guessing "instructional design" is referring to how you model the problem domain being learned so that it's amenable to use with adaptive learning systems?

Most research in this area is drawn from the domain of “Intelligent Tutoring Systems,” (ITSs), thoroughly described in Woolf’s book “Building Intelligent Interactive Tutors.” [WOOLFE2009].

I'd say the classic Intelligent Tutor System construction was pioneered by ADVISOR [BWB2000]. It's a two-agent architecture that first trains a student model on real-world data, then uses the student model to train a pedagogical agent using a more resource intensive algorithm. Most modern ITSs make use of Representation Learning (RL) to train the pedagogical agent.

One of the main techniques for structuring the problem domain is Bayesian Knowledge Tracing (BKT) [CA1995], where you model the domain being learned as a Bayesian skills network; i.e. a 4-tuple of probabilities (init, learning, guess, slip), updated in a Bayesian fashion. An excellent survey work in this area is given in [BGTPF2010].

If you're interested in the RL part of the problem, and how it can work with this setup you might want to look into partially observable Markov decision process — POMDPs. Emma Brunskill has done some nice work in this area.

[WOOLF2009]: Woolf, Beverly Park. 2009. Building Intelligent Interactive Tutors: Student-Centered Strategies for Revolutionizing e-Learning. Amsterdam ; Boston: Morgan Kaufmann Publishers/Elsevier.

[BWB2000]: Beck, Joseph, Beverly Park Woolf, and Carole R. Beal. "ADVISOR: A machine learning architecture for intelligent tutor construction." AAAI/IAAI 2000 (2000): 552-557.

[CA1995]: Corbett, Albert T., and John R. Anderson. "Knowledge tracing: Modeling the acquisition of procedural knowledge." User modeling and user-adapted interaction 4, no. 4 (1994): 253-278.

[BGTPF2010]: Brunskill, Emma, Sunil Garg, Clint Tseng, Joyojeet Pal, and Leah Findlater. "Evaluating an adaptive multi-user educational tool for low-resource environments." In Proceedings of the IEEE/ACM International Conference on Information and Communication Technologies and Development, pp. 13-16. 2010.

Here’s a framework we’ve been developing for this purpose, delivered as a python cookiecutter:

https://github.com/hackalog/cookiecutter-easydata

The framework makes use of conda environments, python 3.6+, makefiles, and jupyter notebooks, so if that tooling fits into your data/ML workflow.

We gave a tutorial on the framework at Pydata NYC, and it’s still very much in active development - we refine it with every new project we work on. The tutorial is available from:

https://github.com/hackalog/bus_number

While it works well for our projects, the more real-world projects we throw at it, the better the tooling gets, so we’d love to hear from anyone who want to give it a shot.