ad 1: have you tried using Cloze cards for that? I like to group eg 1-2 reasons, then I will see the list with 1 or 2 items missing
ad 3: I had success for similar problems by simply creating a lot of cards that give enough context and just ask for the next step. In chess openings, couldn't you just display the current position and ask something like "In opening X, variant Y, what are the next moves for white here"? In some cases I've written scripts to create cards for all variations of a question I want to ask
ad 4: I think Cloze deletions can help to some extent here (I've basically made Cloze my new default card type), but you are probably running into the limitations of Anki there
ad 5: language learning, specifically vocab lists, has always baffled me as a use case for SRS. there is so much context that you need in order to use words proficiently (in what kind of medium was the word used? what register was used (formal/scientific/informal/...)? was it used ironically, empathetically,...?). the only way to learn language imho is to immerse youself as much as possible, through ways where it gets actually used, not such artificial environments
the one thing that I'd like to see changed about Anki would be to have more options for changing the scheduler, or making it easier to use custom schedulers on multiple device types. I simply don't like the logic of SM2/FSRS of hiding a card from you until a specified date and assuming that you'll be reading it then. if you don't open the app for a while, the review dates get completely messed up (I've had new cards scheduled for review sometime in the 2040's). I love the interface and that you can use HTML to enter cards, but I just want to put knowledge in there and get exposed to it from time to time. I wish there was a scheduler that just randomly shows you cards, with probability roughly proportional to how urgently you need to see them. and do not interpret too much into the fact if I haven't opened the app for half a year but still remember some of the cards well. I don't mind seeing those cards "too often", but I do mind if Anki hides the knowledge that I put in there from me, for years or decades even
It's not the logic of FSRS. In fact, FSRS knows the probability of recalling any cards in your collection, which may means how urgently you need to see them. So it can be implemented via an add-on.
> the funds previously marked for Boeing would be open for new firms even if staffed by former Boeing employees
> delivering a very clear message to shareholders and C-suite officers that this sort of conduct will not make them rich
And the best way to make that point is to create new investment opportunities that will make some former employees of the same company and many of the same investors rich?
1. Boeing investors should take a substantial haircut. If those investors subsequently in on new firms, good for them (though initially these new firms may be private). I see no reason they should expect to "get rich" from the investment. The top institutional shareholder is Vanguard whose involvement will have been minimal and whose future behavior is unlikely to change. But Boeing execs whose compensation is largely in stock packages should take a really large hit, and any shareholders who are more apt to be more activist will be more likely to express concern about this kind of conduct sooner.
2. I have to expect that plenty of employees who did not become dead whistle-blowers were not involved in fraud, or poor engineering practices. Certainly a great many long-time Boeing employees expressed concerns about the deterioration of engineering-first culture.
The investors who got rich through Boeing have cashed in a long time ago. What you're suggesting would be much more interesting to them than Boeing is today. You're basically saying 'this handful of companies is virtually guaranteed to share contracts worth billions per year among them'.
> The investors who got rich through Boeing have cashed in a long time ago.
I don't know which investors you mean, and what you mean by "a long time", but I think people that cashed in a long time ago should be able to invest in new firms. Many people say the culture and safety issues at Boeing stems from the McDonnell Douglas merger which was admittedly a generation ago now. But if the estates of Frank Shrontz or Thorton Wilson or the great great grand kids of William Boeing want to invest in a new firm -- good for them? None of them were involved in the period when this fraud occurred, and if they had actually cashed out long ago then they weren't even shareholders in a position to pressure Boeing officers to do anything different.
But there's also no reason they would have any kind of leg up over anyone else in making such an investment today.
Can you explain this part? The sound of "Leaf" could be represented by any number of possible spellings, eg "leave", "lief", "leeve"... - not all are standard English words, but neither is "Overleaf", and exotic/made-up spelling would be just par for the course for a tech product with an exotic/made-up name.
I actually really liked the domain name writelatex.com, because it pretty much tells you what you can do and it's easy to remember, even if you haven't used it in a long time (which could easily happen for a product that gets used a lot in academia and much less outside, eg someone returning to school after a few years of work).
Ah yes, I missed some important context - one of the reasons for moving to a name without LaTeX in it was because we'd just released the first beta version of the visual editor (the rich text mode at the time), and the goal was to keep lowering the barriers to getting started with LaTeX, and to make collaboration easier for non-LaTeX users. And so Overleaf came from a search for a broader name than writeLaTeX.
That's an interesting point about the pronunciation - overleaf is a standard English word, and certainly seemed less confusing than writeLaTeX when said aloud, but I agree it's not perfect! Was the best we could find at the time (especially given the other needs mentioned above).
> Khanmigo now uses a calculator to solve numerical problems instead of using AI’s predictive capabilities. If you’ve been using Khanmigo recently, you may have seen that it will sometimes say it is “doing math.” This is when the math problem is running through the calculator behind the scenes.
> We’ve upgraded parts of Khanmigo to a more capable large language model, which is the software that generates human language. The more capable large language model is called GPT-4 Turbo. Our internal testing shows an improvement in math after we made the switch.
> We are beginning to test the capabilities of a new large language model called GPT-4o, and we’re evaluating other models too to see if they are stronger at math.
> We’ve improved the way AI “thinks” during a tutoring session before responding to a student. We have instructed the AI to write out all the ways in which the student may have arrived at their answer. This approach mimics how a tutor in real life works with a student. We’ve found it significantly improves the quality of math interactions.
> We’ve built new tools to track our progress on math.
> We’re sharing math examples and learnings with others in our field so that we can learn from each other.
> We’re studying the latest research papers on math performance.
Sounds like most of what they're doing is related to prompting, chain-of-thought reasoning and similar, on top of a 'vanilla' foundation model. Sounds like something an ambitious student could replicate / improve upon, so given their mission, it'd be cool if they published the exact techniques they're using and their benchmark results.
I sometimes wish if money, which Facebook, et.al invested in Indian edtech ecosystem, was given to Khan Academy, it would have done justice for the whole world's education system.
If you look at the details, the branding is that of the Indian startup. Facebook actually associated itself with a badly run startup company with questionable ethics. It is a net loss of money and reputation.
I actually disagree. Archiving is a complex issue that touches on the interests not only of the general public but also those who produced or are covered in the materials to be archived, and the ways in which it might affect their interests might not even have been clear at the time of publication (think about LLM training as just one example), or their interests might not have been heard at the time. I think a public institution would be better placed than a private organization to deal with weighing the - evolving - legitimate interests of all parties concerned by the decisions of whether and how to archive something.
Came to initially disagree but realized I was making the same point. Archiving is in the matter of public interest which defaults to government. Resolving the conflict, the internet archive should be absorbed into the library congress. It would be a win-win, the internet archive would have the backing it needs both financially and legally and the library of congress will be able to modernize and expand it’s already daunting mandate allowing them to gain further public awareness and legitimacy.
The Internet Archive works with the federal government currently, and while they receive benefits and incentives as a bonafide library, it is better (imho) they are distinct from the government to remain segregated from potential political influence and interference. This doesn’t prevent a federal agency tasked with preservation from kicking off archiving operations and operating a replicated copy of the Internet Archive (or a subset of the corpus).
In the current operating environment, it is important to optimize for optionality.
(No affiliation with the Internet Archive, just a concerned citizen)
In today's political climate I can imagine one side or the other demanding that all material that espouses position X must be expunged from the archive as "wrong think"
I recently tried to open an X account, the process was punishing: the captcha was six images of ~5 dice each from an acute angle, and I had to click the one where the top faces added up to 15. That takes considerably more time than "select all the cars", especially if the correct image is among the last. And that challenge was repeated 10 times (with a counter to 10 shown from the start, so it wasn't just that many because of retries after failure). I actually went through with it, but in the end I failed to sign up because the verification email never arrived.
I've since been wondering whether Musk now wants it to fail, so he can draw a line under this chapter and shift focus back to his more succesful ventures.
> the captcha was six images of ~5 dice each from an acute angle, and I had to click the one where the top faces added up to 15. That takes considerably more time than "select all the cars"
This has been a fairly standard captcha for the past few years across the "more attacked" providers.
That one, Rotate the [object] to match this direction, Click the item in this image with the highest value, Click the dice that add up to [X], Which of these items doesn't belong, etc.
They are all much more difficult than previous ones, and I've long wondered if they are compliant with various disability laws.
I get it, but 10 puzzles in a row? It's not like the number of puzzles makes a difference to a bot when they're all of the same type (except for the bots that guess randomly, but there would be smarter combinatorial ways to trip those up that are less user-hostile; eg simply saying "select all" instead of "select the one" increases the possibilities tenfold while increasing effort only ~twofold). You're going to be filtering out a lot of humans this way. I personally only went through with it out of curiosity, because I was half-expecting it to not work in the end - as it did, so at least I have a story to tell now.
> It's not like the number of puzzles makes a difference to a bot
They do actually.
A lot of the captcha process isn't just yes/no to the answer anymore (though this signal seems to be increasing lately). They also monitor timings on clicks, movement of the mouse, etc to identify a bot vs a real user. More data on these signals = better reliability for detecting bots.
To be fair, I tried to create an Instagram account with my own website's domain as e-mail address, because I didn't want to use my phone number. When I tried to confirm my e-mail I got a simple "there was an error, try again" message. I looked it up on Google and it seems they just instantly suspended my account. I had to login in the account that I didn't even know was created to see I was suspended, so I could appeal, only to get told to confirm my e-mail again, and then they asked me to link a phone number, by which point I gave up. I hear they may even require a photo on top of that.
I'm sure that was super annoying, but Twitter/X has always had a terrible spam/bot problem. This at least is showing an attempt to take that seriously, even at the cost of new user acquisition.
I signed up for an account so that I could direct message a vendor to let them know their site was down. I managed to get the account made, sent off a message to the vendor, and my account was immediately banned. The message on my homepage told me it would likely take a week for the ban to be lifted.
That was in August of last year. Absolutely nothing has changed since then. I'm not even sure how to reach out to them to appeal the ban. Maybe nobody's home.
I took that as a sign that maybe the platform wasn't worth my time.
Luckily were reaching a tipping point where it takes a similar amount of effort to rig up a CNN to just solve whatever god-awful puzzle they've dreamt up for you this time.
You seem to equate research as a leisure activity with amateur research. If you read 'leisure' simply as 'outside your main, bread-earning job', then it absolutely can be cutting edge research published in peer-reviewed outlets. It's not even that uncommon, look for people who have a university affiliation (that isn't just teaching) as a side gig. And that's just one way you could be pursing serious research outside your main job, you don't even need a university affiliation - it's certainly helpful with access to resources, feedback, ideas (and, let's be honest, it probably helps with visibility/credibility too), but it's not strictly necessary. As a different example, there are some (rare) examples where legitimate mathematical results were discovered by hobbyists going completely on their own. I also remember a documentary about a crew of retirees, including former physics professors and a precision mechanics engineer, who were performing physics experiments of masses at extremely low accelerations to test some limits of special relativity iirc - which in itself was technically the result of 'leisure research', like several of Einstein's most famous discoveries.
If it's something you're interested in, my suggestion would be to decide on a rough area and then seek out experts to brainstorm ideas for a suitable problem for you to work on, maybe one that leverages your existing skills in a creative way (like the engineer in the above example).
Since you mentioned Einstein, it reminded me of a comment by Max Born on the subject, from the Born-Einstein letters:
“Einstein expressed over and over again the thought that one should not couple the quest for knowledge with a bread-and-butter profession, but
that research should be done as a private spare-time occupation. He himself wrote the first of his great treatises while earning his living as an employee of the Swiss Patent Office in Bern. He believed that only in
this way could one preserve one's independence. What he did not consider, however, was the organizational rigidity
of almost all professions, and the importance which individual members of a profession attach to their work. No professional pride could develop
without it. To be able successfully to practice science as a hobby, one has
to be an Einstein.”
> look for people who have a university affiliation (that isn't just teaching) as a side gig.
Can you expand on this? Like people who have full time jobs in perhaps a different field but do research on the side using the facilities/resources of a university? If so how does one slide into a gig like that?
Having done that, I can answer. I'll be a little vague on the topic to keep some anonymity.
Without a plan, per se, I nurtured a burgeoning personal research interest that was well outside my field. The initial hobby was very rudimentary but it gave me enough feedback and fuel that over a number of years I built more advanced software and hardware to investigate it more scientifically. Eventually, having reached the limits of the data I had access to, I reached out to a local university to start a collaboration.
If you already have an idea, I think, as long as you're sufficiently interested in actually doing research, you'll be able to find some professor that is as well. On the flip side, I know for a fact that most professors have more research ideas than students, so they can certainly provide you with things to investigate.
As a bonus, for the professor, since this is a side gig for you, there isn't an cost for them (or much less of one). Access to facilities and resources depend on the university. But I was certainly surprised by what I was given access to in the service of doing the research.
Wow ive been hoping to do something very similar! If you're open to chatting a little more, i would love to learn a bit about how you do this - my email is in my profile or i will contact you if you drop an email here.
I work as a software developer at a major research university. I got a PhD there long ago and I've worked in a wide range of jobs ever since including another job at that university. PhD connections helped me get the first job but not my current job. The university hires a large number of people to do everything from food service, building maintenance, the trades, to technical and administrative. All of those employees have a number of privileges such as being able to take a free class every semester (a grad level seminar can open opportunities like nothing else) and if you have flexibility in hours you can go to talks, etc.
Let's put it this way. If you can write a (1) good grant proposal with a good chance of getting funded you are required these days to (2) get a professor to be the principal investigator if you want to spend it and I'd say (1) is much harder than (2). Not like it is easy, but it is possible.
ad 3: I had success for similar problems by simply creating a lot of cards that give enough context and just ask for the next step. In chess openings, couldn't you just display the current position and ask something like "In opening X, variant Y, what are the next moves for white here"? In some cases I've written scripts to create cards for all variations of a question I want to ask
ad 4: I think Cloze deletions can help to some extent here (I've basically made Cloze my new default card type), but you are probably running into the limitations of Anki there
ad 5: language learning, specifically vocab lists, has always baffled me as a use case for SRS. there is so much context that you need in order to use words proficiently (in what kind of medium was the word used? what register was used (formal/scientific/informal/...)? was it used ironically, empathetically,...?). the only way to learn language imho is to immerse youself as much as possible, through ways where it gets actually used, not such artificial environments
the one thing that I'd like to see changed about Anki would be to have more options for changing the scheduler, or making it easier to use custom schedulers on multiple device types. I simply don't like the logic of SM2/FSRS of hiding a card from you until a specified date and assuming that you'll be reading it then. if you don't open the app for a while, the review dates get completely messed up (I've had new cards scheduled for review sometime in the 2040's). I love the interface and that you can use HTML to enter cards, but I just want to put knowledge in there and get exposed to it from time to time. I wish there was a scheduler that just randomly shows you cards, with probability roughly proportional to how urgently you need to see them. and do not interpret too much into the fact if I haven't opened the app for half a year but still remember some of the cards well. I don't mind seeing those cards "too often", but I do mind if Anki hides the knowledge that I put in there from me, for years or decades even