begemotz

My favorite 'mind training' work is Shantideva's Bodhisattvacaryāvatāra. Technically this is not a lojong text but it always struck me that in purpose it most certainly is. I also remember reading somewhere that HHDL said something to that effect as well.

As for lojong, Geshe Sonam Richen's Eight Verses for Training the Mind is a great little book (I really like his translations).

B. Alan Wallace's A Passage from Solitude is an exposition on another core lojong text, Geshe Chekawa's Seven Points for Training the Mind.

need numbers to go up.

https://archive.is/ytvju

https://archive.is/IQeG6

Given that reddit threads on Claude's performance over the past couple of months are making their way to HN as 'news', I thought this post was interesting enough to add (yes, I see the irony).

If you are willing to rule out mass hysteria (which may or may not be reasonable) - the question is whether continual internal tweaks - for whatever reason - affects performance to a degree that whether or not you are able to use the platform for 'real work' (define it as you wish) can fluctuate from month to month or even week to week. - should you off-load key work to an industry that has not yet found a success business plan? I'm just an AI hobbyist, but this seems insane to me. Of course, if you are on an enterprise plan there is some buffer. But there is no guarantee.

So how do you mitigate this risk?

claude told me to say it.

I am not a professional coder. However, I have typically seen people respond more favorably to Code compared to other systems including Codex (percentage-wise). I'm curious to know what exactly you feel is inferior?

On the other hand, recently for me, the usage limits in Claude have been inconsistent and frustrating. I seem to get a lot less out of it than I have in the past and am considering trying one of the other big 3 sub plans to see whether it suits my use case more.

Interested in hearing whether you think this accurately captures the state of affairs.

It makes sense for companies to show Wall Street that they understand the assignment: adapt or die. It matters less whether a company knows how to deploy A.I. and more whether investors believe it is on track to do so. In that respect, it doesn’t matter whether a company’s stated rationale is sincere or not. Once it announces it is cutting jobs for A.I., the remaining workers have no choice but to buy into that vision.

this is how it works sadly. Short term optics for mid-term pain (most likely)

Full title: Google's 'Project Genie' Is Basically A Huge Plagiarism Tool, So When Will Nintendo's Legal Ninjas Strike?

It's not just the misleading "average" that is problematic. USA Today "reporting" on a "report" by Empower where you find this:

The numbers below reflect Empower Personal DashboardTM users as of October 2025, with the average and median net worth of these individuals, broken down by age. *(These figures are not identical to, nor directly comparable with, representative national data from the Survey of Consumer Finances from the Fed.)*

To elaborate on comments below; I agree that the correct term is "dark patterns". And, one common form of dark pattern is to increase "friction" - to slow down or make more difficult what the user is trying to do (or in other cases, reduce friction "buy now, pay later").

if you are interested in such theories, you might also enjoy Bruno Snell's The Discovery of the Mind.

might want to add percentage dice (d100)

I would echo the sentiment of taking some time. The way to think of this is not as "time off" - but acknowledging that you have the time and space (and resources) to thoughtfully consider your next move. Don't make important decisions under a time-pressure that is self-imposed. Instead, take a month to consider what you want to do during the next state of life - and be grateful that you have the opportunity to consider your next move from this point of view.

https://archive.is/wLnJE

Perhaps there is something interesting here; but what is the comparison group? I am not convinced that earlier generations had values that better matched corporate interests. We were just more likely to show up. It can also be the case that both corporate and individual interests have walked further away from each other as individuals become more hyper-individualistic and corporations become more hyper-capitalist.

like the answer to any other similar question, with a prompt?

Anecdotally I have found that the quality of perplexity.ai has gone down significantly over the past year. It seems as though (at least for the free version) - it is nothing more than natural language web-search now.

not theonion

I figured this would show up on HN. I'm really puzzled why the WSJ would publish this as an 'editorial'. But here is my cynical take -- companies want you to buy into the hustle at 20 so they can exploit you until 30 and spit you out for the next generation of gullible workers.

"There’s no sugarcoating the mental-health struggles, the physical deterioration or the social isolation that came with this intensity...I plan to become a billionaire by age 30. Then I will have the time and resources to tackle problems close to my heart like climate change, species extinction and economic inequality."

I hope this doesn't come across as obnoxious, but *golly* -- the simple, unaware naivete of youth! First, the assumption that he will become a billionaire, second the assumption that he will live long enough to then turn to the things 'close to his heart'. This really does seem like he is trying to convince himself that all the struggles he outlines above is "worth it". In all honesty, I hope he makes it and I hope that, in 40 years, he writes another "editorial" that reads "It was worth it".

Same self-promotion posted 2 days ago: https://news.ycombinator.com/item?id=44364635

and posted 3 days ago: https://news.ycombinator.com/item?id=44351431

For those interested, I would highly recommend `W.E.B. Du Bois's Data Portraits: Visualizing Black America`.

And Anthony Stark's (decksh) recreations of the plates here: https://github.com/ajstarks/dubois-data-portraits

- At a macro-level? I honestly do not know if it will at least in my life-time. And, if it does, it won't be due to higher education being proactive. What I can say is that I have a number of colleagues who are doing all paper and pen, in-class assessment. But this will not work for all disciplines.

- I agree that those who want to learn will learn. And for these individuals, generic higher education isn't that valuable anyway. However, I am a bit pessimistic about the number that represents, on the whole.

And, short-term goals and constraints can undermine these individuals. As with most things, the problem isn't with the thing but with the humans who wield it.

Because that is not how p-values work. Each 'row' is a separate hypothesis test against some null hypothesis.

And frankly the conclusion that "B is winner" because it has the lowest p-value also demonstrates the author's limited understanding.

P-values tell you very little, and generally nothing about what you really want to know.

In this context, you would be better off looking at effect size estimates of key metrics that would signal "better" - not the p-values.

I would say that on the whole, Chat-based LLMs ("AI") have been a large negative impact on higher education. They are good enough now that mediocre to poor undergraduate students (or whole institutions with average students) will receive better grades on assignments generated by AI than if they did it themselves (at least in the disciplines I am familiar with). Or, if not better then comparable with no effort expended.

The short-sighted benefits are just too overwhelming that even for students who arguably have skin in the education game (most do not and merely are there for credentialing) , it is a difficult battle - akin to the generation(s) who recognize that the type and amount of their social media use is problematic but react with resignation.

There is a difference between 'knowing' and 'doing things' - the consequences of AI are different across these.

Some other thoughts:

- I do not think there are any viable solutions to the AI 'problem' in education given the current structure of higher education.

- The affordances, the motivations and practical, if not overt, purpose of 'education' cannot meet AI head on.

- A stronger more explicit distinction needs to be (re-)drawn between what used to be 'vocational' and 'liberal' or 'higher' education. And a narrowing of the purpose of higher education (which would, in itself, be very disruptive)

- There needs to be an explicit addressing of AI in both curricula -- in terms of both practical training on its use as well as the pitfalls and downsides-- in terms of self-interest (e.g. if you have machines do the work to set you 'free' then you run the risk of becoming slaves to the people who make the machines (paraphrased Dune).

- Complete change in approach and most importantly assessment -- I am not sure this can be done in a system where grades are still the litmus test for learning.

- This educational crisis has to be addresssed BEFORE college- otherwise it is probably too late.

Larrin Thomas (creator of MagnaCut) is truly a steel nerd.

Btw, Crucible Steel filed for Chapter 11 recently -- and I know the knifeworld was anxious about what this could mean for CPM steels.

Your hypothesis is: layout influences signup behavior.

This might be your hypothesis but this isn't the hypothesis that the p-value is related to.

Setting a p-value threshold of 0.05 is equivalent to saying: "I’m willing to accept a 5% chance of shipping something that only looked good by chance."

P-values don't provide any information about "chance occurrence" but rather they test the probability of observing a particular outcome assuming a particular state of the world (i.e. the null hypothesis).

Bonferonni

Besides being a very aggressive (i.e. conservative) correction, I would imagine that in industry, just as in science, the motivations to observe results will mean it wont be used. There are other way more reasonable corrections.

Avoid digging through metrics post hoc

The reasonable solution is not to ignore the results that you found but to interpret them appropriately. Maybe the result actually does signal improved intention - alternatively maybe it is noise. Treat it as exploratory data. If that improved retention was real, is this important? Important enough to appropriately retest for?

*PEBCAK

The problem is not the tool or the system; PEBAK.