tonnydourado

Particularly on GitHub, might not even be LLMs, just regular bots looking for committed secrets (AWS keypairs, passwords, etc.)

Maybe if he had read a better bible translation he would've reached a more moderate spiritual enlightenment =P

Ah, yes, the famously very effective sanctions against Brazilian judges. I'm sure it will work again.

/sarcasm

I'll my hat if they aren't.

I have a strategy for dealing with bosses that ignore logic, it's been foolproof so far, going on 14 years of experience.

I call it "quitting".

> there is probably a quite good reason for that.

Eeeehhh, might be overestimating executives a bit =P

But yeah, my first instinct is also to tell Gary to fuck off. That said, I would default to a process reason, so, the advice at the end wasn't totally useless for me.

I don't agree with the optimal path of The HiPPO War. The founder very explicitly said he thinks shipping the current user experience is Bad™, that he'd rather loose the 50k in ads. It doesn't make sense that he would accept shipping it anyway because "We can ship a 'Soul' update" later.

Also, you're commiting the team to deliver something you (probably) don't have the technical knowledge to estimate, so you might be adding another week of death March after just 1 weekend.

The lesson at the end is not wrong, but the characters don't seem realistic.

That said, I have always been IC for a reason, so what do I know.

I have been programming professionally (i.e., getting paid for it) for a much more modest 13 years, but unlike a quite large portion of my peers, I am actually interested in the history of our field.

So, yeah, I agree.

I won't deny OP learned something in this process, but I can't help but wonder: if they spent the same time and effort just porting the code themselves, how much more would they have learned?

Specially considering that the output would be essentially the same: a bunch of code that doesn't work.

The footnotes are top-tier ADHD. Particularly loved the footnote on a footnote, 10/10.

Hashicorp,an IBM company

Everyone in this comment session is now worse for having read this comment

This alternative is still storing the screenshots and the OCRed content unencrypted, and the web UI is also unauthenticated.

There's some docs on storing the data in an encrypted volume or external drive, but 99% of people ain't doing that.

The only real improvement over MS's version is keeping the data local, which ain't much, really.

Also worth noting that Mossad/CIA/etc. are not monoliths. Maybe you got a top agent assigned to you, but maybe your file is on the desk of the Mossad's version of Hitchcock and Scully from Brooklyn 99.

Of all the things I never wanted, a terminal implemented on JavaScript is easily on the top 10 that I never wanted the most.

For Windows, Windows Terminal is pretty ok.

Looking half the way there to fulfill my wishlist for a perfect kubernetes tool. Showing resources as a graph is a big plus but I'd like to also see:

1. Minimizing nodes into just a name, to get a view of just relationships between resources 2. Auto-organizing the graph. You're already warning about overlapping nodes, might as well give me a way to automatically fix it. 3. Reading existing resources from a cluster

3, in particular, is kind of a whole other type of tool, to view/debug existing infra, instead of creating it, but the UI has the right shape to do that extremely well, too.

I agree with other comments that adding a default example to the demo would be nice. I imported this manifest, and it worked pretty well: https://github.com/kubernetes/examples/blob/master/guestbook...

Second todo isn't a todo, should be a trade-off comment. Something like "Technically, this will raise an error if the user triple clicks the button, but most people ain't that fast, so screw it"

Anyone else read it as "Bitch at"?

Gotta say, 100/200 bucks monthly feels prohibitively expensive for even trying out something, particularly something as unproven as code-writing AI, even more particularly when other personal experiences with AI have been at the very least underwhelming, and extra particularly when the whole endeavor is so wrapped up in ethical concerns.

Sigh.

Firstly, and to be fair, this is more for the comments than OP, can we please stop kidding ourselves? People have been trying to make tools to allow "non-developers" to code without coding since COBOL, and they all either failed, or end up requiring that users learned how to code anyway. Except maybe Excel.

Secondly, this has some good ideas, but makes the same cardinal sin of, seemly, all visual programming tools: try to encode control flow visually. No one wants to write ifs, fors, and whiles as connected nodes. It sucks. Text is MUCH better for that. You probably are very, very rarely writing flowchart diagrams in real life. I know I don't. I write state machine, architecture, data flows, etc., unless I'm being forced by some braindead tool to write control flow logic with little boxes and arrows.

The thing that is actually interesting here is the boxes that provide higher level abstractions, like HTTP requests, LLM calls, etc. Give me one or two different nodes in which I can just write some code to process input and return output, and maybe some dispatch and map nodes, and focus on providing higher level abstractions out of the box, and allowing me to build my own abstractions too (i.e., parametrised graphs). Think more functional, less imperative.

As is, that doesn't look like a tool I'd be excited to use, but I won't deny it has some potential to be interesting.

Hopefully, not for long.

> It’s easy to fall into the trap of thinking that you get paid for work because it’s important

Is it? Maybe I'm out of touch, or used to a different work culture (non-US, consultant), but that's kinda obvious, you're paid because you make money.

Also, no one that knows a single teacher or nurse would ever fall into the trap of thinking that salary is proportional to job importance.

Of all the things I heard about deep seek that's the one that has motivated me the most to try it out XD

The studies weren't about gender in mice because, gender being a social construct, mice don't have it.

*Some* of the studies were about the potential physiological effects of hormone therapy. They used mice, but the point was to study how humans are affected by HRT.

The way the article is written makes it sound like they're putting mice in drag and asking them their pronouns.

Tangentially, fuck yeah, negative results, just as good as positive ones

I would add: a user's country of residency can be mapped 1-to-1 to the user's preferred language, or at least to a language the user understands.

So, SO many websites get this completely wrong.

> They're not evil, just out of touch

Kinda hard to see the difference. I, too, live in a completely different world than people with much less money than me, but I can still conceive that they can't have a cleaner twice a month, order food every other day, or use uber more often than public transportation. I wouldn't even consider making a decision that impacts people's lives without having at least an inkling of how they actually live.

> My name is Lars Wirzenius, and I was there when Linux started.

"I was there, Gandalf. I was there 3000 years ago" XD

I thought Australia sucked until I checked the latencies for South America and South Africa. Not a single "good" latency link =)