DantesKite

I like:

"Explain in 10 sentences or fewer."

"Give me the gist of what you just said."

"ELI5"

Wonderfully captivating article about a loquacious man that like the rest of us, wrestled with his demons and even sometimes machine elves.

I do find vibe coding fulfilling. Very much so. But not in the way someone who gets into a flow state with coding typically does.

Why? Because I can finally make rapid progress on ideas I've had that would have taken me years to develop by myself. I enjoy the outputs. Only the outputs.

There is genuinely no flow state for me. It is a grinding list of frustrations I handle whenever the agent falls short of my expectations.

It is assembly work in that sense. I am an order of magnitude more productive, but the work itself is not intrinsically enjoyable the way coding used to be.

I am a factory worker churning out features as I monitor the LLM agent and its outputs. There is no craft. Only productivity.

I do not mind because I was never a very competent programmer to begin with, so there is no major loss on my part. This black box produces what I truly want, but I will not pretend it has any meaning for me. It provides no more meaning to me than cranking a lever for hours on end.

But I love seeing my ideas come to life. These distant dreams I had only in my head are suddenly walking and breathing after years of only thinking about them.

About four right now.

I used to have far more, but I learned the habit of bookmarking tabs that I've kept open for a long period of time.

Actually, now that I think about it, this would make for a useful browser extension: have an LLM that automatically asks to bookmark tabs that have been inactive but present for X number of days.

If you could do calculus in high school, you likely have more than enough neural horsepower to build a wide array of commercial products with AI, since it can breakdown, suggest, recommend, and educate you on the conceptual gaps that remain.

The hard part is that you'll have to inefficiently learn about these concepts as you encounter them, as opposed to a more skilled engineer who can already anticipate and use them ahead of time.

In general, increasing GDP per capita has all sorts of positive second and third order effects on net, likely improving the collective welfare of society.

It's a bit far removed from what you might envision as "doing good" though. It's a bit too abstract to feel on an emotional level.

And true, not all acts are equal. The emotional valence of improving ad rates for Meta can't really be compared to saving a child from cancer on a moral scale that most humans intuitively understand. Nor would I demand that you try to do so.

I often find that "making the world a better place" is synonymous with "looking for meaning in my own life".

Having children is usually a good start, if you don't have any already. Making the world a better place for them does wonders for the soul.

Or so I've been told.

Alternatively, if neither children nor economics appeals to you, you can always just donate money to St. Jude's Hospital or another charity. The money you earn would then make the world a better place in a way you can directly feel.

You might like https://abooklike.foo/.

I think Wordle is instructive here in that it has many lessons for how to make a game go viral or become even mildly popular.

For starters, perhaps you should make it also available on the web, if at all possible. That makes it easier to share and discover.

And maybe invest in the ability for users to export the gameplay as video. I could see it spreading more if people post long combos that last for minutes, sped up with music, for example and posting on TikTok, Twitter, etc. Otherwise, even if it's a good game, most people aren't ever going to discover it, except after many months.

Obviously it always helps to make a game more fun and compelling, but I presume you're already aware of that.

It helps to clarify your goals and what it is you want to do with your life, otherwise you’ll always have this ambiguous sense that you’re not progressing in the way that you want.

What do you mean by “productive”? What would progress look like for you? And towards what?

Highly recommend MovieLens if you have eclectic or niche movie tastes. It can be a bit of a nerd-snipe though. One of my favorites activities is rating movies and watching the recommended ratings (what rating it thinks I'll give a movie) update overnight.

It's at the very least, better than average chance at predicting which movies you will like.

If true, sounds like a strong argument for removing minimum wage laws, so as to provide more economic opportunities for people and create a skill ladder they can climb.

The best mental model of MVP I have found is that it is in some sense a science experiment and you’re trying to test a specific hypothesis as efficiently as possible with the resources you have, because you ultimately don’t know what’s going to work.

Yes. It's basically a custom StackOverflow human available to you at all hours of the day.

I really enjoyed the Edward Herrman narration of the "At the Mountains of Madness" when I listened to it years ago.

What struck me is how it still feels haunting even decades later; it aged quite well. I couldn't help shake the creepy feeling there was something unearthly at the edges of our world.

Once your child grows a bit older (or is particularly precocious), I recommend Math Academy.

If you search for "Math Academy" on Twitter, you'll find all sorts of great stories about it. The creators have studied the science of pedagogy pretty well to optimize the learning process at scale to every user. Justin Skycak in particular on Twitter discusses what goes into creating the content, the scaffolding, interleaving methods they use, buildup of automaticity through memory, and how working memory affects rates of learning. Justin has even shared a textbook on his Twitter page that details a high overview of the various teaching methods that go into the curriculum (although from what I understand it's still a work in progress).

You can even try it yourself. A few mathematicians use it to brush up on old concepts they haven't used in a while or to add a little that they didn't know previously.

They're also constantly expanding the courses.

There was a great Reddit thread where the creators answered some questions about the program and why they made certain decisions (posts from thread below).

Highly recommend since it'll answer a lot of doubts you may have about it. I really do believe it's probably the most effective way to learn math today.

" '[Question 2] It's hard to believe that 5 hours a week for a year starting from basic multiplication tables (found on your homepage) will have me completely prepared for higher-level university courses. I'd prefer an explanation for people who are not familiar with the XP system.'

Yeah, I realize that this can be a bit shocking!

I think the most convincing way I can answer this question is to start by telling you that in our original in-school program, 6th graders start at various places in Prealgebra, and then do about 40-50 minutes of fully-focused work per school day for the next 3 years. This takes them all the way through Algebra 1, Geometry, Algebra 2, Precalculus, and AP Calculus BC, passing the AP exam by the end of 8th grade.

Of course, I realize that also may seem kind of shocking, but we have the AP scores to prove it, and there's been plenty of news coverage over the past decade: https://www.mathacademy.us/press (Let me know if you'd like me to elaborate more on this.)

Now, look at the numbers: 40-50 fully focused minutes/schoolday x 180 schooldays/year x 3 years comes out to about 24000 minutes or 400 hours, and our Foundations series is about two-thirds the size of that (since roughly a third of topics are not actually prerequisites for university math), which comes out to about 267 hours, and then divide by 52 weeks in a year, and you're at about 5 hours per week.

Anyway, if there are parts of this argument you want more clarity on, or any parts you're finding unconvincing, let me know and I'm happy to elaborate."

Source: https://www.reddit.com/r/learnmath/comments/1edeuw8/comment/...

One of the other reasons why I recommend Math Academy over textbooks is because fundamentally the process can be so much more efficient than a textbook could ever provide. Justin himself notes the same:

"...Yeah, MIT OpenCourseWare (OCW) is what I used to self-study a bunch of advanced math while growing up. OCW is a good resource and I came a long way with it, but for the amount of effort that I put into learning on OCW, I could have gone a lot further if my time were used more efficiently. (These issues are not OCW-specific; they apply to textbooks in general, Khan Academy, etc.) That's one reason why I’ve been so motivated to help build Math Academy. We take away as much of this learning friction as possible and maximize your learning efficiency.

By the way, I wrote more about my experience here: https://www.justinmath.com/why-not-just-learn-from-a-textboo..."

Source: https://www.reddit.com/r/learnmath/comments/1edeuw8/comment/...

Such a great demonstration of the concept.

I think it’s inevitable that we’ll start to see more sophisticated ways of organizing our social media feeds.

I don’t think it has to be this binary decision where we either abandon social media altogether or expose ourselves to the most emotionally draining content possible. There’s likely many different unexplored metas as it were.

I often joke that we should have a marketplace of algorithms we can subscribe to, where the sentiment “slider bar” can go from Hello Kitty Island Adventure positivity to 4Chan LiveLeak nihilism, if you so choose.

If you toss it into Google’s NotebookLM, it’ll probably produce a much more accessible version of the blog that’s succinct and easier to parse in a few minutes.

That does sound like a surprisingly useful and when I think about it, obvious application of LLM’s.

Makes me wonder how many other tasks involving category organization can be automated. Probably quite a few.

That’s a great website. I’ve been looking for alternative book recommendation websites for a while and it really has nailed it down.

It even recommended me a somewhat eclectic book I’ve recently been meaning to read.

Is there a reason you limit to only 6 favorite books? Is it due to computational restraints?

It’s interesting comparing the comments here to Twitter where there’s been a lot of discussion about the ethics of this (as in she shouldn’t have done this for some ethical reason that is never quite elucidated in detail, other than it’s wrong to do).

In contrast, many comments here are talking about the ethics of not having this approach more broadly available, which is fairly optimistic for Hacker News.

Or maybe it’s just that a subsection of bio-ethicists on Twitter have oddly shaped moral values.

According to this tweet, he used Math Academy.

“ Excited to see a @_MathAcademy_ student blowing up Hacker News with some insights they've realized about the process of learning math!”

https://x.com/justinskycak/status/1827397012758917622?s=46&t...

Anecdotally I've noticed adding electrolytes seems to help reduce the incidence of headaches and possibly migraines, at least according to what my family members say.

n=2

Trust essentially.

I don't trust random websites I've never used before, but I know Substack is a sure thing in terms of not excessively charging me money.

A few years ago, the PPM in my town's water was over 850 PPM, well above the recommended guidelines. Worse than that, it had a distinct sulfur-like smell.

So naturally we got a reverse osmosis water filter system and while the tap water has improved since then, I'm always reminded of the occasional accidents that can occur (e.g., lead, excessive chlorination, plain old entropy).

Plus it makes the water taste significantly better. Even if tap water was always perfectly safe to drink, I'd get one just for the taste alone.

Brondell Circle RO systems are my favorite because the filters are the easiest to change and when you have the same system for years, that ends up being the labor you repeat the most.

I’m curious if the breadth of the antigenic response increased as a result or significantly narrowed.

Perhaps I missed it but it seems like a large part of the focus was on whether or not that many vaccinations had damaged his immune response.

Steve Weis on Twitter described it best:

“This iMessage exploit is crazy. TrueType vulnerability that has existed since the 90s, 2 kernel exploits, a browser exploit, and an undocumented hardware feature that was not used in shipped software”

https://x.com/sweis/status/1740092722487361809?s=46&t=E3U2EI...

GPT-4.5

I want to get a feel for the trajectory of how much more these large language models can improve.

I do. I'm quite fond of it.

I think one thing I would recommend is actively muting accounts and words that you don't enjoy and creating lists.

Over time, you'll be able to fine-tune you feed so you get a really great "newsletter" and takes around current topics.

The lists are important because sometimes you don't necessarily want to hear about experts daily. Maybe once a week. Or when something breaks out in a part of the world you're interested in.

Muting words is a very powerful feature that's under utilized, because humans are mimetic and something in the zeitgeist may resonate for quite a long time. Especially if it's cruel or chaotic.

I mute any words or accounts that make me frustrated in a powerless way.

Anger is an important emotion, especially for overcoming fear. And expressing disdain for something you don't like is a healthy response. Most importantly, it can be fun. I'm not so naïve as to not acknowledge that.

But I do think if you live in a garden, it should be a place you enjoy. Whatever that may be.

For now, Twitter and many social media platforms lack the ability to fine-tune the "algorithm" to your specific needs. A feed that's only positive might take some time to get there, but as your interests expand, you'll find your niche.

I think eventually these books will end up becoming training data for LLM's but at the moment, a human plus a machine is far more powerful than either one in isolation.

You still have time to derive some utility from them. Additionally, you can use ChatGPT to help you learn the material faster.

It doesn't have to be mutually exclusive.

If we knew, the other LLM's would be better.

The correct answer is: Nobody outside of OpenAI technical staff currently knows.