You must log in or register to comment.
…Verily?
Honestly, if I can use an AI agent that is not coming from an evil corporation that’s destroying the environment, I would use it to help me solve problems like figuring out why I have a build error, why I’m getting certain error messages, why a piece of code isn’t working as I expected, find potential bugs, etc.
It can be a great tool to help you grow, like having a tutor explain where and why you made a mistake so you can learn from it.
NOT if you ask it to do the job for you and generate your code.
I used a local model today to write a long form regex for me today. it would have taken me 30 minutes. instead it took 3 seconds and worked out of the box.
I think this is a perfect example of what LLMs should be used for.
Ok. But did you understand the regex? That’s what people are missing out on. The learning aspect.
absolutely. I know regex, better than most devs I know personally.
I also don’t use it all the time. I happened to have it up and just said, “hey fucker, give me the pcre regex to find all instances of these keywords in a string like this”
reviewed it, threw what I had into my code, ran my tests and everything was golden.
That’s the neat part, you can.
If I had to pick one, it’d be Olmo 3 from Ai2: Non-profit. Ai2 is a research institute, not a company that needs to sell you a subscription or harvest your prompts. Fully open, not just “open weights” Ai2 publishes the training data, the training code, every checkpoint and the logs. Anyone can check exactly what went in, and in principle rebuild it from scratch. You don’t have to take their word for anything.
Runs locally. The 7B version runs on an ordinary gaming laptop or a recent MacBook. Your code never leaves your machine, there’s no data centre involved when you use it, and the energy cost is your laptop being on. Apache 2.0 licence. No usage restrictions and no strings attached.
Good enough for the tutor use case. Explaining errors, walking through why code misbehaves, and pointing out likely bugs are all well within what it can do.
To be fair about it: the training data is largely scraped from the web, like nearly every model. The difference is that Olmo is honest and auditable about it.
If strict consent matters most to you, StarCoder2 is cleaner (only permissively licensed code, with an opt-out)
Devstral is very good
Ah, yes, “when used responsibly.” Like how we use drugs, alcohol, guns, cars, money, power, and everything else responsibly? Humans can’t be trusted with things that do more harm than good when used irresponsibly.
There’s an example used in regulatory circles: electricity would never be permitted in homes under today’s regulatory standards - we only accept it and its hazards because it is so incredibly common and useful in daily life and we’ve been accepting it since before we cared so deeply about safety.
So, have today’s goalposts moved so much that “AI” no longer passes modern society’s definitions of acceptable? I feel like such a technology would have been cheerfully embraced in the 1960s.
AI is the car of modern times. It’s convenient, ruinous for the environment, and a danger to anyone in proximity of one. Hell, we even wasted tons of land on highways kinda like what we’re doing with datacenters. But boy, does it get you from point A to point B quick.
boy, does it get you from point A to point B quick.
Not so long ago, face to face interaction was the only way business was transacted, people were employed and paid.
In today’s world, if we could get past the baseless RTO mandates and let the 60% or whatever it is of modern workers work from home or wherever is convenient for them to work, get the commuters off the roads, restructure the food distribution systems to eliminate that need for cars, we really could make a significant improvement in the environmental impact of cars and trucks.
In the world of AI, I suspect a huge chunk of current AI usage, energy burned in data centers, etc. is “wasteful” - investigational, experimental, learning usage. If we can get past that kind of usage and eventually use it for what it’s good for and leave it alone for counterproductive wastes of time, it should also be a clear net-benefit for “getting stuff done” not only faster, but more correctly and valuably. Which software is more valuable: one coded by humans over the course of a year, with a half dozen undiscovered security vulnerabilities, or one coded by those same humans over the course of a month with those vulnerabilities exposed and fixed?
If you’re concerned about the environment, those humans commuting to work for an extra 11 months add up to far more environmental impact than the AI they used to accelerate their work. Good luck getting their employer to give them 10 months off with pay for delivering the same valuable product faster and better.
There is one key difference though, drivers and car manufacturers are sometimes held accountable for their actions
sometimes held accountable for their actions
Very weak theater.
“Knives cause lots of stabbings, so let’s ban knives”
All those things, including knives, are heavily regulated, yes.
I’m sorry, who or what exactly regulates knives? Knife manufacture? Knife ownership? Even knife sales… If you’re referring to ballistic knives or switchblades, sure - those are like guns, but the 99.999% of knives in the world? The ones anybody with a piece of metal can hammer out and sharpen for themselves?
In plenty of countries carrying knives in public is not allowed. You can own one for use in your home or work but don’t brandish it out in the open.
The reason few people know this is because it’s rarely ever needed, and the people violating that law are generally the people you want arrested anyway (e.g. because they’re trying to stab someone).
I see nothing special about knives in that respect, though I have no doubt there are laws.
Making credible threats of killing or seriously injuring people, with whatever tools, would seem to be a good basis for arrest.
Nice try to equating a knife to fucking drugs. AI is even worse than drugs, is like the digital version of nuclear weaponry: it shouldn’t even exist.
I swear, some people still think Prohibition was a great idea.
I mean, it is how ice cream became so popular… https://www.nationalgeographic.com/history/article/ice-cream-prohibition-era-craze
However, the end of Prohibition was a much better idea than the start of it.
It was.
If the goal was to reduce the amount the average American drank alcohol year over year (and that was the goal) it was an overwhelming success; not only during prohibition but thereafter as well.
Banning stuff and then enforcing the ban works really well and we have tons of evidence of it working; not just in the US but literally world-wide.
You seem to have missed the 1001 negative side effects of Prohibition, that eclipsed the tiny ideological victory of reducing drinking.
https://www.pbs.org/kenburns/prohibition/unintended-consequences
Additionally, there is no consensus on whether Prohibition was even successful at that, owing to a lack of hard data.
https://en.wikipedia.org/wiki/Prohibition_in_the_United_States
This is quite aside from the fact that today we’ve figured out that criminalising drugs and users is the wrong approach and does a lot more harm than good. But we STILL have people like you.
Then just let AI use AI. I’ll collect my Nobel prize now thx.
You might want to rethink that position: https://www.anthropic.com/research/self-replicating-ai
Sorry, I don’t really feel like reading anything from Anthropic atm (I might later, it sounds interesting), but my actual, non-sarcastic position on the matter is fuck LLMs; they should not be doing any sort of policymaking.
Long story short: AI using AI can find ways to run itself on TV set top boxes and similar common IoT devices, as part of a bot-net. It can also find ways to self-propogate such bot-nets across many surprisingly low powered and widely distributed devices in our world.
And then yesterday he said something along the lines of “we need to make it harder to contribute to the Linux kernel.” Which I believe he means to say “oh god there’s so much being submitted all of a sudden and it’s all slop.”
The fucking point that Linus misses is, using AI by itself is intrinsically the more irresponsible thing to do. Change my mind.
How do? Does this apply even if it’s a locally run LLM? Even if it’s done using green energy like Mistral? Even if all of the AI written code is carefully reviewed by professionals?
Is it a problem of making it too easy for people? If so, what is the correct amount of easiness for people? Shall we start taking clothing down to the river to scrub instead of using washing machines too?
attack it with ethics!
You are claiming it’s irresponsible, why do you say so?
AI is just a technology, how you use it makes it bad or not.
Nuclear is the cleanest safest energy we have, but you can also make bombs with it.
The “no such thing as a bad tool” argument falls apart if you start thinking about it for longer than a few seconds. If you have a hammer made out of styrofoam, that’s a bad hammer.
A tool is thus any kind of thing that lets you do something you wouldn’t otherwise have been able to.
Immediately goes on to using a fork as an example. Wtf are you doing eating with a fork that you couldn’t be doing with your bare fingers?
Argument is stupid, and falls apart at the first hint of critical thinking.
Wtf are you doing eating with a fork that you couldn’t be doing with your bare fingers?
Preach. There was a lot of fuss made when I was in school about “humans are the only animals who use tools” - of course by then all kinds of animals had been documented clearly using tools, but they were still teaching the kiddos (self included) that “humans are special.”
I have been making and using software tools for 45 years, a tool is something you use to make a job easier, faster, cleaner, safer, a tool might make the impossible possible, but that’s far from a requirement for “tool-ness.”
I’ve read this article before. Found it very reductionist. For example, it says that weapons might be good for Ukraine, but they’re bad in general. You can’t just “in general” something like weapons or AI. They have thousands of different contexts and use cases. Sure, a weapon can make somebody homocidal in one context, but it might make somebody a militia in another context. AI can be a slop generator for a vibe coder, and a bug detection tool for Linus Torvalds.
I’m not a fan of how modern politics and social justice is so reductionist nowadays. “One aspect of X is bad, so we must reject X altogether”. Can we add even a little nuance to our discussions? This mindset is brainrot.
In other words, you found an article whose headline seems to make your point for you - which you probably didn’t even read or you might have an actual answer. I did read it, and found the reasoning spurious, like jumping through a series of philosophical hoops to prove all taxes are illegal or that birds don’t exist. The author’s example of a Chinese dish with an unusually large amount of chopped peppers being more easily eaten with chopsticks than with a fork, could just as well have been a bowl of broth that’s much easier to eat with a spoon than with chopsticks. All it demonstrates is that any tool can have appropriate and inappropriate use cases. They’re still just tools.
I read it
I read it
Did AI translate it for you first?
intrinsically the more irresponsible
That doesn’t claim it’s intrinsically irresponsible. You can use AI responsibly
deleted by creator
I’ve been thinkin:
We went from STEM to STEAM. But I think we really need SHTEAM. (The H is for Humanities.)
Philosophy, psychology, history, basically anything that would get nerds to think about the social consequences of the stuff they like to tinker with.
I’ve said it before, and I’ll say it again : eventually you just fucking end up with “be a complete person; here’s the entire dewey decimal system.”
STEM is meant to deal with logic based thinking and the problems unique to that industry, like for instance, a huge gender gap. Shoving arts and humanities in doesn’t fix the issue. Cross disciple collaboration is great, expanding your horizons and putting ethics into practices is great, but that’s not what STEM is.
There are already plans for these things in practice (hippocratic oath, iron[1] ring[2] ), which need to become more widespread. We don’t have to have STEAM toys and SHMEAT courses.
There’s many good reasons why even Bachelor of Science degrees have such a load of core courses across the academic spectrum.
Yes. That stuff is necessary. I am agreeing with that, while maintaining that “STEM specific initiatives” need to be STEM specific. I’m talking about conferences and toys, not courses.
Warning! very unpopular opinion
Title
I hate STEAM, preferred when it was just STEM.
very unpopular opinion
True! I think your opinion sucks lemming! 😎 /s
But seriously, while I do think STEAM sounds a little dumb compared to STEM (and I’m young enough where I saw the name changing happening in real time and despite being at the age where it’s easy to impress on kids, I never really cared for the name), I absolutely agree with the idea that we need to teach STEM kids and professionals more about the non-tech side of hard sciences and subjects.
SHTEAM sounds and looks stupid as shit, but we should absolutely take steps to properly educated up-and-coming STEM majors that they need to understand how their work and inventions affect humanity, along with actually having an appreciation for the work that they will do. Not teaching these things is how we create idiots that are proudly declaring “Mecha hitler” and “AI will kill us all, and that’s good!”
People complain about electives and gen ed classes in college like, “why do I, a science major, have to take an ethics class?” and then go on to become Sam Altman. As a STEM major, I fucking hate those people. I understand you want to get your money’s worth in college, but having to learn history and take one ethics or culture class should always be mandatory. I have learned so much about people and become more knowledgeable about humans and empathy through arts and ethics classes, and have gained an appreciation for them. It also affects how you do tech in general too. So often do I hear about coders, security analysts, data scientists, etc who make music, art, literature etc, and feel like it affects and improves how they work.
but having to learn history and take one ethics or culture class should always be mandatory.
Fuck it, one of each. Programmers had the opportunity to show they could handle not getting a humanities education with their diploma and they blew it.
So SHTEM?
That looks like it’s the name for an STD
stop trying to force a reality that fits your narrative
lol how many tokens did you burn today
Its been an hour, the chatbot hasn’t given them the perfect retort yet.
Perfect retort to name-calling? FM69’s comment is not exactly a rhetorical masterpiece.
I find this comment a kind of masterpiece, because there was both no actual name calling, and you refuse to type FartMaster’s actual name.
believe it or not i have the ability to still use my mind even when there are tools at my disposal! And I was at the local veggie market talking to people during that hour, so sadly again your narrative is wrong 😞 your response, however, pretty low effort… might need to throw a captcha at you
You won’t trick me into solving basic logic problems that you and your bot cannot!
let’s see, i woke up and did jiu jitsu. I went to the local vegetable market with my dog, bought some breakfast and got some coffee and practiced my Spanish with the vendor. I spoke with a nice lady about art residencies and her desire to live an alternative lifestyle even though shes 75. I had a conversation about iterative prisoner’s dilemma game theory. Then I came home and now I’m replying to your low dimensional attempt at an insult, while spending some tokens to fix a form validation bug across a web app 💙
Ew, a web sloper even.
how sad that’s what you decided to highlight…
I agree, but the question is, how to use responsibly. I like to use it as an assistant that can look things up. Googling has the tendency to bring up outdated information, a good AI model is able to cross-reference what it finds with documentation and release notes. I never let the AI write anything for me though.
This is exactly how I use it too. Google and other search engines have gone to complete shit in order to cater to LLMs. So that’s how I use the LLM. You are my search bitch, Gemini. No, you’re not allowed to actually create anything.
i came to this forum thinking i’d find interesting stuff and it’s mostly just people getting angry at reality
Why is it bad if people get angry at things that are happening in the world? Would you rather we get angry at fiction?
it’s fine, i’m just commenting that it’s getting tiring and I don’t see people doing something constructive with that anger on a forum where my perception is there are a lot of intelligent people who have power from technical understanding that carries a lot of power. I have a need to see progress and structure through collective wisdom as I think that is lacking.
When reality is shit one must rage at it.
Rage is uncontrolled anger… it’s not helping anyone and certainly not the person experiencing it. If you can find a way to channel that anger into something constructive, then you are doing something beneficial for all
Yeah, but if that’s all you do nobody wants to be around you.
We’re on !linux@programming.dev. You think people want to be around us?
There is more to life then !linux@programming.dev
Hmmm, prove it using a bash script.
Sudo rm -rf ~/bin/.*
Claude, write me a bash script that proves there is more to life than !linux@programming.dev
Blasphemy!!!
Oh no, you are one of those “if you don’t adapt you will be left behind” people, aren’t you?
AI is the new reality aparently.
or imagine a reality where AI just exists alongside other things. Stop catastrophizing
Ignoring the hype is not catastrophizing.
alright, i’m disengaging from this echo box… everyone should go outside and sit in nature for a while
Feel free to do so next to a data center. Feel (hear, breathe) some externalized costs for a while.
i’d be curious if you eat chicken from the supermarket? Have you ever lived next to a chicken factory, breathed and listened to that one? Do you drink Coca Cola? Sorry to break it to you, but we’re all complicit in things that are destroying nature and humanity and just because you’ve desensitized yourself to the other ones don’t mean they’re not happening. You can rage all you want at each one, but it’s not going to be constructive. At some point we just have to come to terms with reality that we’re also responsible for this reality and I don’t agree with AI’s effects on nature either, but I think if we’re smart we can leverage its capabilities in the window of opportunity available with the people who are most capable (good developers) to create pathways leading toward better things. Or just go yell and rage some more, as you wish.
low dimensional thinking
Yea I’m sure when Google came out “Real Programmers” were angry at all these people just looking up solutions to their coding issues, and using Stack Overflow instead of studying the API or books.
Only to be replaced by people that embraced the new technology.
im old enough to know this is exactly what happened
Only to be replaced by people that embraced the new technology.
I think you have to make a distinction here:
It makes a huge difference whether you know what you’re doing and can work more efficiently thanks to the new technology… or whether you have no idea and think the new technology is a magic wand that can solve everything (which you only think because you have no clue about the subject and therefore believe that whatever the new technology produces is good or desirable).
To stick with your example: Someone who couldn’t code before didn’t suddenly learn how to do it on their own just because of StackOverflow; it just made things easier for people who already had some knowledge but were simply lacking the specific know-how.
In this case, I’d cautiously assume that Torvald can judge what makes sense (within his area of expertise) and what doesn’t.
I think that is still true today with AI just to a lesser degree.
After google many people could copy templates, change a few things and boom they “Built a Website”
It’s pretty obvious when someone who has no clue, just used AI to slop out code or something. And it’s often riddled with bugs and issues.
Where as someone with say 10+ Years experience in coding, can use AI to do their job a lot faster and better, and denying that is delusional and you will fall behind - !RemindMe 5 years
Eg. I can just throw my error logs and repo at a local hosted AI, and it will find the bug significantly faster than I would have.
Presies.
Having an AI help you code is like getting your six-year-old to help you put a shelf on the wall. Undirected, they’re just going to make a huge mess for you to clean up later. Given specific tasks within their abilities, they can actually make it both go faster and be an altogether more pleasant experience.
With direction and an operator who knows its limits, AI can be used to semi-automate many boring tasks. One of the things I’ve got set up in some of my open source projects is an LLM that takes bug reports and turns them into regression tests. I don’t always end up committing the AI-generated test (sometimes it’s very slow and I can better pinpoint the behaviour and target a test at that), but it helps because I don’t have to do that scaffolding myself, and I can just immediately jump into a debugger.
Yeah, but therein lies the problem
If you don’t have new junior devs learning how it all works through trials by fire, we’re going to be the last generation of senior devs.
I like llms, they’re interesting and can be a real force multiplier. They let you skip the unimportant details and build bigger and faster. You design the system and focus on the architecture, and you get pretty great results
But you can also leave it all up to the ai and create things that no one understands. They’re also not going away
But I don’t know how we go forward from here. My best idea is that we make it illegal to provide llm access as a service - maybe everyone who uses one should have to set up their stack themselves. It’s not a perfect solution, but at least everyone who uses them will have to learn them like they would any other tool in their toolbox
I don’t fear AI in the slightest, but I do fear the ways we’re using it as a species
My best idea is that we make it illegal to provide llm access as a service - maybe everyone who uses one should have to set up their stack themselves.
Are we gonna set up a Great Firewall to block providers in other countries too? And VPNs, and all the other fun stuff.
Fall behind what
Every single one of you is behind the best coders in the world. Guess you should just give up and take on crochet or pickleball
You’re behind, why botherThere are thousands of professional football teams that are behind the UEFA top 10, why do they even bother?
Yes, exactly. That’s what i said.
I’m sure I’m also behind the best cyclists in the world, and having a basket on my bike might make it slower, but it’s way more enjoyable and functional for me that way.
And it better serves my purpose, sure I’m not going to win any races, but I never wanted that in the first place.
Except that a basket isn’t analogous of AI, a smoky two-stroke engine that pollutes the air with unburned fuel and oil mix is. And your reliance on the engine will atrophy your legs lol
You’re on the slow-roll group ride saying you need an engine to keep up and we’re looking at you with disgust because you’re so loud and we can’t breathe.
My analogy is fine, you waisted it, but Lol Funny enough, I love 2 strokes.
I’d rather rip it on my KTM down a single track any day of the week rather than suffering slowly on a road bike.
Both are a workout, only one is really a full body workout, way more fun and comes with adrenaline.
until he fell into the slop vat and came out sloppified
thank you meta!
So how do they gatekeep the vibecoders out then? FOSS need to change and develop new quality control systems. Stop complaining about vibecode, just review it before you use it. Why can’t you do it?
rationing:
Limit lines of code or diff size. Limit scope. Limit PR per user per month. No bots, obviously. New users don’t get to send anything, they can comment and vote, at best. And bring in negative ratings for contributors: low quality, suspicious code, bad PRs.
This is another way of saying deceleration, a word that accelerationists hate.
Stop complaining about vibecode, just review it
Hahaha. Not a developer, huh?
There is no way I could possibly review my coworkers’ 1000s of lines of code he posts daily. I just blindly approve it because fuggggit we don’t care about quality only velocity.
This is different from before because before he would normally post 2 or 3 small code changes per week. Also, he would have tested it himself so starting off it would already be in a good state. Also, since he tested himself, we could talk about the change since he still had the context.
Now it’s just ship, ship, ship. We literally can’t “review it” like we did before. That world is gone.
Because the pace of code production has officially overwhelmed the human capacity to review it. As it was discussed recently. Linus answer was to use more AI, with the implementation of code review tools that gobble up code and spew out summaries.
For the time being, humans reading the summaries count as human reviewed. But at some point in time this system will crack. The current zeitgeist is to restrict who can contribute. Which leads to the obvious follow up question of who decides based on what, and how to decide who deserves to contribute.
You just witnessed the first steps into the philosophical and political erosion of the open source movement. Not just on the kernel but it is also affecting almost all popular FOSS projects.
But of course, that doesnt matter because Linux is not and has never been political. /s
Anti-social hermit discovers that his job is more fun if he has someone to talk to.
You have to kind of feel sorry for him.
There is no responsible way to use AI
