I didn’t say they have no knowledge, quite the opposite. Here a quote from the comment you answered:

LLMs are extremely knowledgeable (as in they “know” a lot) but are completely dumb.

There is a subtle difference between intelligent and knowledgeable. LLM know a lot in that sense that they can remember a lot of things, but they are dumb in that sense that they are completely unable to draw conclusions and put that knowledge into action in any other means besides spitting out again what they once learned.

That’s why LLMs can tell you a lot about about all different kinds of game theory about tic tac toe but can’t draw/win that game consistently.

So knowing a lot and still being dumb is not a contradiction.

But wouldn’t you point still be true today that the best AI video models today would be the onces that are not available for consumers?

(sorry for the long answer)

Besides the vendor-lock-in and “enshitification” things (which are also a direct consequence of capitalism) just compare the Microsoft Office Suite against LibreOffice. It’s not even close.

Don’t get me wrong it works good enough for typical use-cases but the difference in depth, quality and polish is huge. Same with Photoshop vs Gimp, After Effects and other Adobe stuff.

Tons of software doesn’t even have proper Open Source equivalents because they are so nieche (compared to office software or VLC for example).

Unity (vs Godot), IDA Pro (vs Ghidra), EnCase, SolidWorks, etc.

Check out how Mozilla Firefox makes its money. They are over 80% funded by Google and make all kinds of shitty decision. Not because they are bad people but somebody has to pay development because good software doesn’t just happen because a hand full of junior devs have some free time.

Look at the shitty Ubuntu decisions, also done because somebody needs to pay for development.

Look at audacity, redis, FluentAssertions, OpenOffice, MongoDB, ElasticSearch, MySql, IdentityServer4 and many more OpenSource projects that went commercial or changed their license because it was unsustainable.

You can of course always fork, but the forked project than has the same issues. Development works as long there are skilled people motivated to practically donate their free time in exchange for nothing. That already greatly limits the people that even can work on open source and if their situation changes history repeats.

It’s not even OpenSource only. Look at YouTube creators especially science educators. A lot of them with great content and talent explaining various topics. But also trying to sell you stuff like squarespace, Brilliant and NordVPN because it would be unsustainable otherwise.

So yes great (by that I mean big polished good quality “unenshitified”) FOSS won’t happen under capitalism because writing software is time-consuming.

Why do you think people pirate commercial software like AutoCAD, Photoshop, Lightroom, InDesign, MS Office, SketchUp, SolidWorks, etc.? Because they are polished in a way no open source project could be, because hundreds of engineers worked for decades on those.

All that said, FOSS is great. I use and rely on a tons of OpenSource software, self-host a lot of services, regularly contribute to quite a few projects and also am the maintainer of some libraries other people use (even commercially) and I still stand by what I wrote before:

Open Source can’t be as polished and high quality as big commercial software because it doesn’t have the funds to do that. Five motivated people are not enough to write an After Effects competitor.

Probably won’t happen under capitalism. It’s way too expensive (time consuming) to write good software/make good products.

Coding isn’t special you are right, but it’s a thinking task and LLMs (including reasoning models) don’t know how to think. LLMs are knowledgeable because they remembered a lot of the data and patterns of the training data, but they didn’t learn to think from that. That’s why LLMs can’t replace humans.

That does certainly not mean that software can’t be smarter than humans. It will and it’s just a matter of time, but to get there we likely have AGI first.

To show you that LLMs can’t think, try to play ASCII tic tac toe (XXO) against all those models. They are completely dumb even though it “saw” the entire Wikipedia article on how xxo works during training, that it’s a solved game, different strategies and how to consistently draw - but still it can’t do it. It loses most games against my four year old niece and she doesn’t even play good/perfect xxo.

I wouldn’t trust anything, which is claimed to do thinking tasks, that can’t even beat my niece in xxo, with writing firmware for cars or airplanes.

LLMs are great if used like search engines or interactive versions of Wikipedia/Stack overflow. But they certainly can’t think. For now, but likely we’ll need different architectures for real thinking models than LLMs have.

I don’t see how that follows because I did point out in another comment that they are very useful if used like search engines or interactive stack overflow or Wikipedia.

LLMs are extremely knowledgeable (as in they “know” a lot) but are completely dumb.

If you want to anthropomorphise it, current LLMs are like a person that read the entire internet, remembered a lot of it, but still is too stupid to win/draw tic tac toe.

So there is value in LLMs, if you use them for their knowledge.

I can’t speak for Lemmy but I’m personally not against LLMs and also use them on a regular basis. As Pennomi said (and I totally agree with that) LLMs are a tool and we should use that tool for things it’s good for. But “thinking” is not one of the things LLMs are good at. And software engineering requires a ton of thinking. Of course there are things (boilerplate, etc.) where no real thinking is required, but non-AI tools like code completion/intellisense, macros, code snippets/templates can help with that and never was I bottle-necked by my typing speed when writing software.

It was always the time I needed to plan the structure of the software, design good and correct abstractions and the overall architecture. Exactly the things LLMs can’t do.

Copilot even fails to stick to coding style from the same file, just because it saw a different style more often during training.

There actually isn’t really any doubt that AI (especially AGI) will surpass humans on all thinking tasks unless we have a mass extinction event first. But current LLMs are nowhere close to actual human intelligence.

Text that’s not code might also work.

A drill press (or the inventors) don’t claim that it can do that, but with LLMs they claim to replace humans on a lot of thinking tasks. They even brag with test benchmarks, claim Bachelor, Master and Phd level intelligence, call them “reasoning” models, but still fail to beat my niece in tic tac toe, which by the way doesn’t have a PhD in anything 🤣

LLMs are typically good in things that happened a lot during training. If you are writing software there certainly are things which the LLM saw a lot of during training. But this actually is the biggest problem, it will happily generate code that might look ok, even during PR review but might blow up in your face a few weeks later.

If they can’t handle things they even saw during training (but sparsely, like tic tac toe) it wouldn’t be able to produce code you should use in production. I wouldn’t trust any junior dev that doesn’t set their O right next to the two Xs.

I don’t think it’s cherry picking. Why would I trust a tool with way more complex logic, when it can’t even prevent three crosses in a row? Writing pretty much any software that does more than render a few buttons typically requires a lot of planning and thinking and those models clearly don’t have the capability to plan and think when they lose tic tac toe games.

Play ASCII tic tac toe against 4o a few times. A model that can’t even draw a tic tac toe game consistently shouldn’t write production code.

Practically all LLMs aren’t good for any logic. Try to play ASCII tic tac toe against it. All GPT models lost against my four year old niece and I wouldn’t trust her writing production code 🤣

Once a single model (doesn’t have to be a LLM) can beat Stockfish in chess, AlphaGo in Go, my niece in tic tac toe and can one-shot (on the surface, scratch-pad allowed) a Rust program that compiles and works, than we can start thinking about replacing engineers.

Just take a look at the dotnet runtime source code where Microsoft employees currently try to work with copilot, which writes PRs with errors like forgetting to add files to projects. Write code that doesn’t compile, fix symptoms instead of underlying problems, etc. (just take a look yourself).

I don’t say that AI (especially AGI) can’t replace humans. It definitely can and will, it’s just a matter of time, but state of the Art LLMs are basically just extremely good “search engines” or interactive versions of “stack overflow” but not good enough to do real “thinking tasks”.

Best thing is, they introduced some settings to turn that auto-conversion off and they don’t work 🤣 can’t make that stuff up.

Hetzner ❤️

Does that get translated if someone else with a different language opens that file?

Floating point numbers and arithmetic is not inaccurate. They are actually very accurate but a lot of developers have inaccurate assumptions about them. They can’t exactly represent base 10 decimals. That’s the only inaccuracy. If you have two floating point numbers and you let’s say add or multiply them the result is always the closest floating point representation of the real result.

The list of misconceptions wouldn’t reasonably fit in a comment, but if you are really interested and have a few minutes you could give that a read: https://zeta.one/floats-are-not-inaccurate/

Selling your data would be stupid, because they make money with the fact that they have data about you nobody else has. Selling it would completely break their business model.

I don’t think that’s how it works. If it exactly looks like something protected by laws like copyright or whatever your country uses, I highly doubt that any court would say that it’s fine just because it was created by AI.

Ok I also try it one last time 🤣

Go to Google images and search for “Desktop”. What you see is Desktop machines amd setups and how I and the vast majority of the world use the word “Desktop”.

Now search for “handheld game console”. It’s very likely that one of the first few results is literally a SteamDeck.

Now back to the stats. As I already said. SteamDeck will be tracked as a Desktop because stat tracking sites just use Browser User Agents and try to detect what the device actually is, but that’s very hard if not right out impossible because clients (including the SteamDeck) intentionally (for privacy and compatibility reasons) lie about what they are all the time!

If you take your mobile browser and enable “Desktop site” or “Desktop mode” it will lie(!) and make the server think it’s a Desktop - even though it is really not. A smartphone doesn’t magically become a Desktop PC. If I browse the web with my typical mobile browser - every site will track my activity as smartphone. If I switch to Desktop mode most sites will track me as a Linux Desktop Machine. But my device has not changed.

So you are right that the SteamDeck is tracked as a Desktop PC. But that’s because the Server has has either no better category for the device or can’t determine what the device really is because it lies about what it is.

https://webaim.org/blog/user-agent-string-history/

Stat tracking always had (and will have) two big issues (which can’t really be fixed).

Devices which lie about what they are (see link above) and the problem that they have to come up with some categories and there will always be some devices which fall between the categories (Think fridge, microwaves, sex toys, etc.).

If your SteamDeck is currently actually connected to a monitor a mouse and a keyboard than you are actually using it as a Desktop PC. But if you use it like most people - even though the SteamDeck lies about it - it’s not a Desktop, because the word “Desktop” really is about the form factor - it’s not just my definition. Give any of your friends a piece of paper and a pencil and ask them to draw a Desktop PC - I would actually be amazed if anybody in the world (even you! outside the context of this discussion) would draw anything even remotely resembling a Steam Deck.

👋