Hacker Newsnew | past | comments | ask | show | jobs | submit | shostack's commentslogin

There are so many things I wish I had time to learn about. I don't need my learning resources, I need a way to jack in and have them uploaded to my brain.

This is just an excuse for not learning and getting it done. You do have a way to upload to your brain, it is just a slow bandwidth connection that takes time. If you don't have time you have to make time by not reading the news, not reading social media, not watching stupid videos.

The book Make It Stick by Brown, Roediger and McDaniel is helpful. tl;dr of it is:

- lots of low-stakes quizzing and practice

- spaced repetition

- reflect on what you've learned and what you could do better next time, and apply these lessons in different contexts

- interleave practice of different but related topics

- try to solve a problem before being taught the solution

- distill the underlying principles to different problems

- remember that if learning is easy, you probably aren't engaging you brain very much

This will help streamline the process, but obviously there's just a limit to what you can take in.


All these things presume actual interest and savviness about the topic present in the student beforehand, which is precisely what most students that struggle with studies lack.

Actually, some of the research they base their advice on was performed on elementary school students, and college classrooms which had poor attendance; ie, not the most engaged students. Simple things like giving elementary students an ungraded quiz right before class (to force recall) two or three times a week raised grades substantially, and a college class that switched from midterms/finals to 9 quizzes plus a final not only had higher attendance, but also had much higher grades on their finals with basically none of the students falling behind. Another experiment had young kids practice throwing beans bags into a bucket, one group alternating practice between 2 and 4 feet, and another only practicing at 3 feet. After a month or two, they were tested on throwing the bag into the bucket at 3 feet and the kids who practiced at 2 and 4 feet performed significantly better than the kids who only practiced at 3. Anyway, my point is that small, simple changes to how you study can have big implications for retention, without too much extra effort.

Sorry, I'm still reading this book right now and it's super interesting.


Yes, if you have something like a classroom setting with a teacher who can just tell students to do things, that can serve as an external motivator for students who lack intrinsic motivation otherwise.

But when you just grab a pile of learning resources off the internet, the teacher doesn't come included. You need to be at least motivated enough to become your own teacher, or else find a way to have someone else supervise your self-study.


Not just a limit. Different people have different personalities. And different subjects have specific mechanism required for mastery(eg Surgery vs Philosophy). So different people fit different learning mechanisms. Then the problem is about awareness of where you fit and skill at coordination with others who fit elsewhere.

Very good tips.. I always mess up when doing spaced repetition since I don't take notes, I try to re-read the whole previous material in the book again and I get demotivated that I have to read all that so that I remember all the previous material. Do you know a way to get out of this habit?

Start taking notes.

interweave*

No, it is interleave in this case https://www.retrievalpractice.org/interleaving

Do you wish you had time to learn about them? Or do you wish you just knew them? Having them uploaded to your brain might make you know about them, but is much different from having time to learn them. This is important if for you, like for me, learning itself is a large part of the enjoyment

With a sufficiently sophisticated harness you can actually do quite a lot by just talking to your AI. I have regularly dictated to build things on my phone while walking to lunch for example.

Which of those providers are:

1. Keeping your data private on in the US

2. Not training on it

3. Not quantizing the model

4. Offer reasonable latency adds rate limits


OpenRouter has a list of providers, looks like NovitaAI would meet those criteria. Though not for $50/mth for 80/M tokens, which I assume is the Z.ai subscription pricing.

https://openrouter.ai/z-ai/glm-5.2

https://novita.ai/models/model-detail/zai-org-glm-5.2


This is literal whataboutism. Refute the comment if you have a valid point to make.

Erm, no? it is pointing out hypocrisy?

If you are against the thing China does, when US does the thing also be against it.

US has always done what China does, now trump is doing it vocally.

So it is easier to point out the hypocrisy in it.

Before there was plausible deniability.

Thank you for your attention to this matter


Well, it's more complicated than that.

Let's imagine that China and USA do some equally bad thing against my privacy, let's say intercept and store all the traffic between myself and my AI provider. I am inclined to excuse it from USA and not from China because it's much more likely that China would use it to undermine the security of my home country. USA on the other hand, is literally a military ally of my home country.

I don't naively think that USA is a 100% benign country because of the alliance and that my interests are aligned with them. Trump especially has made that part worse. They're just more aligned with me than China.


I wonder how many issues 5.1 could have caught if you ran it as multiple adversarial reviews against the original output it gave you.

Well I did run deepseek against the original. They all seem seem to spot different issues

And how does this compare to openai which is an ad company now? I am interested in trying codex but I want to know how my data will be used for ads or measurement and training data if I become a paid subscriber. And what happens to that data when I stop subscribing?


my comment was not an endorsement of codex. when I said "also" I was referring to openai as the other problematic company :)


10% off is practically nothing and also irrelevant because it is the total cost of the trip that matters and they can easily increase that over time behind the scenes in a way that makes up for that 10% and then some once they determine the price elasticity of these premium customers which I imagine is quite higher.


Also, a lot of people have dreams about a perfect idealized version of something instead of the reality. If they actually got to experience the reality they might find that they no longer hold that dream as they originally knew it.


Reading this somewhere in the Atlantic on a sailboat by myself. Nodding yes. A few days from now I will be at home after an almost 2 year journey of learning what sailboat ownership. I still can’t tell if I like the sailboat life. I’m going decide in the fall of 27 If I really want a sailboat. I had to try it or give up the dream. Maybe I’ll go back to my dream of running a small run manufacturer if facility.


This is so true. A few years ago I was volunteering at a running race that’s very hard to get into and many people have as a life dream. I found it kind of amusing how many people were having pity parties in the latter stages of the race despite having dreamed of getting into it for years, maybe decades.


More like incredibly lucky that the global hardware market dried up for compute capacity even as his AI product flopped. Right place at the right time.

I just dislike that it is now harder to avoid giving Musk money directly or indirectly.


Look up differential privacy. Done right, it is impossible to do what you said.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: