I completely agree and appreciate sincerely that they released this. It’s unfortunate, the way the obviously nonsense claims made by the industry at large - “LLMs are AI and can do everything!” - have polluted a lot of devs’ ability or willingness to see the tools for what they are, and maybe official acknowledgements like these can help.

It also seems likely to me that the major players know a lot of negative truths about all this stuff, you’re probably right about hiring observations. I don’t follow any of their marketing really so I have to admit I’m even just assuming that releasing this is out of character.

If I’m being honest, I’m mostly just on the edge of my seat waiting for the hype bubble to burst, lol, and curious about how it’ll unfold. Probably just kind of hoping this marks a step toward that.

Interesting read and feels intuitively plausible. Also matches my growing personal sense that people are using these things wildly differently and having completely different outcomes as a result. Some other random disconnected thoughts:

1. I’m surprised they’re publishing this, it seems to me like a pretty stark condemnation of the technology. Like what are the benefits they anticipate that made them decide this should be published, vs. quietly kept aside “pending further research”? Obviously people knowing how to use the tools better is good for longevity, but that’s just not what our idiotic investment cycles prioritize.

2. I’m no scientist or expert in experimental design, but this seems like way too few people for the level of detail they’re bringing to the conclusions they’re drawing. That plus the way it all just feels intuitively plausible has a very “just so” feeling to the interpretation rather than true exploration. I mean, cmon - the behavioral buckets they are talking about range from 2-7 people apiece, most commonly just 4 individuals. “Four junior engineers behaved kinda like this and had that average outcome” MIGHT reflect a broader pattern but it sure doesn’t feel compelling or scientific.

Nonetheless I selfishly enjoyed having my own vague subconscious observations validated lol, would like to see more of this (and anything else that seems to work against the crazy bubble being inflated).

Neither this nor the comment you’re replying to help at all with providing v3.10 in particular. Unrelated.

(Just pointing it out for anyone later trying to solve it themselves)

If you’ve already got Docker available, testcontainers is awesome!!

I like little SQLite in-memory DBs for testing sometimes but just accepted the annoying friction of a slightly different SQL dialect (typically use PG). Testcontainers was a “have my cake and eat it too” situation, really like it.

Mind sharing any more details, like which model (or models and which to avoid), and what kind of front end technologies you’re involving?

I’d like to start doing what you’re describing for simple front ends on my own personal projects. I’ll need to learn front end properly someday I suppose, but not today lol.

Ah, the ever-elusive, mysterious stage in my process - the one I can’t ever seem to move much before the “planning” and at least “beginning to implement” stages, and sometimes stubbornly comes even later than that.

I think I agree with you, and I also think you probably know better than me, but - Python couldn’t become what Python became without doing this exact thing very deliberately, bordering on obnoxious at times. Fundamentals or “initial state” define the characteristic strengths and weaknesses for a language, but what to add and what not to, as well as “why” and “how”, over time determine the true shape and user experience (lacking a better word there) of a language.

Despite its reputation, in my view Python has always been far more opinionated about how to do things than most give it credit for.

Dotnet professionally and using lemmy.ml socially is hilarious to me and (sincerely) entirely consistent. Makes perfect sense, I just find it funny. (I’m not being sarcastic or attacking you, might not be clear lol)

Well, now, that’s useful, but we shouldn’t fail to mention good ol HCl, muriatic acid colloquially for this purpose, also great for cleaning oil stains from a driveway!

Happens a lot - my (quite small) shop was using NestJS for backends and my boss is way more experienced and wise than me. I unintentionally caused us to switch over to Python, which probably sounds as silly as JS to many, but - we deliver dope shit, on time and on budget 🤷‍♂️

FastAPI ftw, fight me! Lol jk Django is cool and useful and serves a different need, quite well from what I understand.

Dangit. That’s me too, I just saw your comment before posting one myself 😅

Rubbed me the wrong way too, dude is the exact opposite of AI. He IS clearly vibing in the photo however.

I haven’t bothered to actually search or troubleshoot yet, but since I’m here - have you had any problems with power management failing to automatically turn screens off when idle?

I don’t get consistent behavior there it seems (AKA it leaves them on when it shouldn’t), but that’s I think the only significant oddity I’ve found in the ~7 months or so I’ve been running Bazzite. And like I said I’ve done basically nothing yet to try to solve it, just wondering if you’ve seen it. I have the issue on a desktop and a laptop, using entirely different monitors (not even same brand) FWIW.

Yep I use KDE-flavored Bazzite and actually forgot GNOME was even offered! It works deliciously. Came over from Windows last winter finally and boy, the UI alone is just so much nicer.

Blech, this is my least favorite kind of testing. I’d much rather have some all-encompassing integration tests even if they’re confusing AF, than the “yep the language still works as advertised” nonsense that this approach often amounts to.

Just wanna add another Bazzite recommendation! If you’re on the fence and feeling like “but my games!” - Bazzite’s got your back. Shit works great.

For me it’s less about fear and more about having a limited budget of time and effort to spend on learning things, so CSS and front end generally gets deprioritized. But that’s cuz I’m a back end kinda dev in my soul, lol.

I’ve seen the good points you’ve made elsewhere in this thread - I would indeed react very poorly to willy-nilly back end changes and I think you’re right that people don’t give CSS and visual styling the same degree of professional respect when making changes. And that sucks.

Sounds like a certain lake troll is agitating to improve his lake-dwelling experience, if ya ask me.

So I think I was wrong, but you are too lmao.

10¹²⁰ is the number of valid game-trees, or valid ~80 move games.

The much smaller number I quoted above, though, IS the valid positions, I was thinking it was actually the trimmed down “truly valid” game-tree sequences.

Isn’t math fun? Limitless ways for us to be wrong!