Seems like a decent balance to me. They note that there's no substitute for experiential learning. The harder you work, the more you get out of it. But there's a balance to be struck there with time spent.

What I do worry about is that all senior developers got that experiential education working hard, and they're applying it to their AI usage. How are juniors going to get that same education?

This is also what I often wonder.

Imho, AI is a multiplier and it compounds more as seniority grows and you know how to leverage it as a tool.

But in case of juniors, what does it compound exactly?

Sure, I see juniors being more independent and productive. But I also see them being stuck with little growth. Few years ago in an year, they would've tremendously grow at least on the technical side, what do they get better at now? Glueing APIs via prompting while never even getting intimate with the coding aspect?

TFA> By regenerable, I mean: if you delete a component, you can recreate it from stored intent (requirements, constraints, and decisions) with the same behavior and integration guarantees.

The only way to do this is with a mathematically precise and unambiguous stored intent, isn't it? And then aren't we just taking source code?

> I _explicitly_ said that there will be changes that require adaptation.

I think this understates how incredibly expensive, violent, and deadly that adaptation will be.

What are some of the problems with eradication of fiat currency?

Gold is heavy and difficult to verify it's purity and authenticity

This model makes the most sense to me. If you just model it as: Trump wants to do things that make Trump look good, everything he's done fits into it quite nicely. If you want to predict his next move, think to yourself, what does Trump think will make him look the best to his adoring supporters?

Logically, conquering Greenland makes zero sense and is only damaging to the United States. But to his supporters, it will make Trump look powerful and good. Which is why he's talking about it, and why I think there's a decent chance that he's going to do it. I just hope there are enough sensible people left in his idiocracy cabinet to stop him.

This isn't a product, but on the topic of unplugging, we decided to replace our TV with a bookshelf. One of the best results of this decision was it allowed us to rearrange the living room in such a way that wasn't focused on the TV and was instead focused on being social.

I call it the “TV altar” when the whole living room is set up to face the TV. People don’t like that

Brilliant term. I actually don't have a TV in my apartment either, which is also uncommon here in Japan. My son makes lots of noise in any case.

I use Markdown for all my books, currently. It used to be a custom XML format, but that was really annoying to type even with custom Vim keybindings.

I do wish Markdown were more capable, but it's a good lowest common denominator for HTML and PDF. Also Pandoc-flavored markdown is pretty decent.

My current flow is:

Markdown -> preprocess -> pandoc -> HTML

Markdown -> preprocess -> pandoc -> HTML -> page-splitter -> split HTML

Markdown -> preprocess -> pandoc -> LaTeX -> PDF

That last one is slow, and I'm hoping to replace it with Typst, probably:

Markdown -> preprocess -> pandoc -> docbook -> xlstproc -> typst -> PDF

I've tried other things like Sphinx and it's tough to find something that checks all the boxes I need.

In general, though, I'm pretty impressed with Typst. I wrote a test program that takes the XML output from cmark-gfm and converts it to Typst with xsltproc. It produces PDFs in orders of magnitude less time than Pandoc/LaTeX. I use that now for all my casual PDF documents. https://github.com/beejjorgensen/xml2typ

> I'm hoping to replace it with Typst

This might be stupid but is there a reason you want to go through docbook and xsltproc instead of setting --pdf-engine=typst?

Looks like it's getting there, but still needs some work. It's not obeying the book output or number sections. I can probably get the indexing going, though.

Measured at 10x faster than LaTeX.

Yes. Two reasons:

1) More control over the typst output... 2) I didn't know you could do that. :D :D :D

I'll check it out--thanks for the pointer!

Repo is nice, thanks. Small nit: it would be nice to have included the final PDF output of your tool on `test.md` in the repo.

That's not a "nit", that's a request, and for someone with the same process, it would be really annoying.

There's a reason you don't want generated data under version control.

Is a link in the README to an example PDF really too much to ask?

I don't think so, but again, it's a nit.

I can tell you the current thinking of most of the instructors I know: teach the same fundamentals as always, and carefully add a bit of LLM use.

To use LLMs effectively, you have to be an excellent problem-solver with complex technical problems. And developing those skills has always been the goal of CS education.

Or, more bluntly, are you going to hire the junior with excellent LLM skills, or are you going to hire the junior with excellent LLM skills and excellent technical problem-solving skills?

But they do have to be able to use these tools in the modern workplace so we do cover some of that kind of usage. Believe me, though, they are pretty damned good at it without our help. The catch is when students use it in a cheating way and don't develop those problem-solving skills and then are screwed when it comes time to get hired.

So our current thinking is there's no real shortcut other than busting your ass like always. The best thing LLMs offer here is the ability to act as a tutor, which does really increase the speed of learning.

Thanks for the response, I appreciate it. I absolutely agree with you about CS education. I went to school to learn how to learn. So, the best-case outcome is everyone has A Young Lady's Illustrated Primer available to them. At that point I suppose to really does live with the individual as to whether they want to see how much potential they really have.

>>I get it, you’re too good to vibe code. You’re a senior developer who has been doing this for 20 years and knows the system like the back of your hand.

>> [...]

>>No, you’re not too good to vibe code. In fact, you’re the only person who should be vibe coding.

All we have to do is produce more devs with 20 years of experience and we'll be set. :)

Original author of the post here. Just to give credit where it's due, that was a quote from this other excellent article written by someone else:

https://www.stochasticlifestyle.com/a-guide-to-gen-ai-llm-vi...

I did double quote it, but you're right—I should have made it clear.

I have an old point and shoot (20x optical zoom) and a SIM-free phone that has never been used in my name for anything.

> SIM-free phone

Is the modem completely disabled? Does it still show the "SOS" option that allows you to call 911 without a SIM? If so, and if it's ever been turned on in your residence, there's a decent chance the IMEI could be traced back to your house just based on pattern-of-life movement.

That's why you also need to enable airplane mode.

Which you can just engage on your current phone.

Airplane mode often leaves bluetooth on, with all the tracking that enables.

On Android at least for now, you can use systemui tuner to pick what gets toggled for airplane mode: Cell, Bluetooth, Wi-Fi, NFC, WiMAX.

No root needed

Analog recording and viewing may need to become the state of the art given how low the bar has become to manipulate digital media

great news! this old shoulder mount with a strap mounted VHS recorder will finally be able to come out of storage and find a new life! /s

not sure why you feel analog recording is necessary. just need a camera that isn't part of a phone. any DSLR, MFT, Mirrorless cameras would be just as good.

however, there's something to be said about live streaming so that even if the camera is confiscated the images are already publicly available.

I don't disagree with you about the encumbrance and impractically, but the live streaming providers could eventually become unreliable or compromised in a number of ways (genAI, political pressure, advocacy for an agenda of its leadership)