AI and programming.

liquibyte · 2026-08-16 20:50:03

What's everyone's thoughts about this post by golinux? Should AI be allowed to help programmers with research or tricky bits of code or should it strictly be verboten? I find that it's a great research tool but apparently some people are dead set against it.

golinux · 2026-08-16 21:52:39

If you know anything of human history, the pattern is to create new tools/toys which eventually come to bite us badly. That is because humans are fatally flawed in their judgment and motivations. What is needed is a transformation of the human mind from never-ending pursuits of froth and foam that keeps us running in place and repeating the same blunders over and over . . .

IOW . . . we need to stop shooting ourselves in the foot!

Perhaps we are what we eat but we definitely become what we think. All the Sci-Fi I read in the 50s is now horrifyingly manifested in our daily life.

We need deep wisdom not a roller coaster of titillation with shiny new s**t to survive the future that is on our doorstep . . .

liquibyte · 2026-08-16 22:37:04

Bro, it's a tool. It's a really useful one too if you control it. I get the fear because for all of human history anything new and good has been used by those in power to control others. Did you say that very thing when the internet came about? Look what they've done to that. We could make amazing things with a computer program as powerful as that and release it to the masses for free for anyone and maybe even better than what exists in the walled gardens. Imagine Linux with a 90% share on the desktop. That's all I'm saying, it's a dead useful tool and if we don't do good with it bad and greedy people will control the world with it. I used it to figure out how to integrate Compton and WPEWebkit into my program and I'm rather pleased with how things worked out. Manually doing that research would have taken me another decade to work out. The code is mine but asking the right questions really steered me to faster results. Am I a bad person for asking for ideas and a little help with the integration path?

golinux · 2026-08-17 00:19:45

I am just attempting to make you aware that "sleeping with the enemy" is a dangerous gambit that could easily have catastrophic consequences for "freedom" and you, personally. Digital knowledge and skills are quickly lost if not used repeatedly. Imagine a future that due to reliance on AI solutions, humans lose their ability to effectively control their digital devices? Game over.

Altoid · 2026-08-17 01:20:10

Hello:

liquibyye wrote:

... a tool. It's a really useful one too if you control it.

Yes, it is indeed a tool.
A badly thought out / designed one.
And yours is also a huge "if".

I'd say it is already running amok.
And as no one bothers to control anything, soon there will be no control possible.
So it will get worse.

golinux wrote:

... deep wisdom not a roller coaster of titillation with shiny new s**t to survive the future ...

+1

Best,

A.

greenjeans · 2026-08-17 01:28:26

“There is nothing either good or bad, but thinking makes it so”
~Shakespeare

golinux · 2026-08-17 01:35:40

The mind is the source of all our woes . . .

greenjeans · 2026-08-17 01:39:09

The mind is the source of all our woes . . .

And the source of all our joys. I choose joy.

“We create most of our suffering, so it should be logical that we also have the ability to create more joy. It simply depends on the attitudes, the perspectives, and the reactions we bring to situations and to our relationships with other people. When it comes to personal happiness there is a lot that we as individuals can do.”

~His holiness the Dalai Lama the 14th in the Book of Joy

Last edited by greenjeans (2026-08-17 01:55:46)

golinux · 2026-08-17 02:06:14

The Dali Lama is right. All of our "suffering" is self-inflicted. Being able to maintain emotional "neutrality" toward whatever comes your way is quite liberating!

IamThatIam · 2026-08-17 02:17:17

I fully disagree with golinux on this. I think that if you are using the AI as a tool and are acting as a software designer or architect and keeping a schematic with blueprints of your software, intelligent questions for the AI, and with a specific design, and also keeping records of what you are doing then there is nothing wrong with using AI to make coding more efficient. I think alot of people will come up with niche clique ideologies to be an elitist and will turn everything to its extreme and try to say if it is bad for AI to be in control of all of our sotware then AI must be bad in all situations and circumstances and for any purpose. I disagree. I think this is just being an extremist and truly a software fascist as well as an elitist. There is nobody stopping people from deciding not to use AI. It isn't being forced upon you necessarily. There is nobody saying that using AI means letting it do everything and not having an intelligent designer in control of the AI. I think when we start making everything into its extreme polarity like that then we fall victim to the ideologies of the brown shirts in Germany way back when. People need to think for themselves and use common sense, and stop letting other people tell you how to think. It does not take a rocket scientist to know that using AI makes coding more efficient and faster and it also does not take a rocket scientist to know that you cannot let AI be in control when coding nor can you use AI and code without being the intelligent designer who keeps records, backup copies, and uses a schematic with intelligent prompts and questions. Everything must be intelligently thought out first and put on paper before you start using AI to code otherwise it is going to do things you do not want it to do. Also I would not use cloud based AI if I could help it. Claude and Grok are not so bad but localized is always better because you can control it and tell it what to do. I would say Enoch is a good way to go. Be the programmer and the designer and not the one just using someone else's service.

greenjeans · 2026-08-17 02:36:38

Oh yeah!

blackhole · 2026-08-17 07:51:40

What many of you fail to realise is that it's all for profit. "AI" is just the last product from Big Tech, it is being heavily invested in, it's a bubble and it will burst at some point. It exists to make money for shareholders. But the immense amount of hype surrounding it, means that your boss is slowly becoming convinced that it can do your job. It probably can't - or maybe it can do aspects of your job well enough, that you can be laid off. You can then run along and babble about "AI" geing a great "tool", while the billionaires laugh at the real "tools" who bought into it all.

As Altoid said, it will run amok, because as with everything else, the risks are all acceptable, and the "running amok" is actually part of the plan - we're now at the stage where the big players want it regulated, so that they can maintain ownership, control and exclude competition. To be seen as viable "AI", and boost share value, LLMs are going to need be "headline news".

Altoid · 2026-08-17 11:05:36

Hello:

blackhole wrote:

... many of you fail to realise is that it's all for profit.
... exists to make money for shareholders.
... while the billionaires laugh at the real "tools" who bought into it all.

+1

Have a read:
https://www.theregister.com/columnists/ … ay/5288275

Best,

A.

Last edited by Altoid (2026-08-17 11:11:13)

RedGreen925 · 2026-08-17 15:36:01

Yet again I see no one mention my biggest problem with the garbage AI. If you use it you are plainly and simply a thief. In the Free Software community we support copyright and those models were trained on stolen code from the entire internet. Now if a company trains an AI on its own self produced or licensed for that purpose code it is a different thing. Though they have lost any claim to copyright when doing so then using it as all code produced by a machine is not allowed that protection. As far as I am concerned all those AI companies should have their executives sitting in jail and all their money taken as that seems to be the solution when an individual does it and infringes copyright. As they claim that corporations are people for their purposes when it suits them and their justification of the parasite behaviour of their actions and its harms towards society.

sgage · 2026-08-17 15:58:21

The notion that a given technology is neutral, just a tool that can be used for good or ill, is not borne out by fact or history. Technologies have their own imperatives, their own tendencies, and 'unintended consequences'. This goes 100-fold for so-called 'AI'. Look at what it has already done...

Last edited by sgage (2026-08-17 15:59:04)

blackhole · 2026-08-17 17:01:56

RedGreen925, I mentioned the plagiarism aspect in another recent post and in the past in other posts here and on other sites.

The likes of Microsoft, the owner of github, are well placed to use the "tools" they are promoting to "steal" code from the massive array of private and public github repositories it hosts. This is in fact a means for proprietary vendors to effectively circumvent GPL licensing.

SS · 2026-08-17 17:06:30

Fascinating topic. I hated the way Microsoft was getting too overwhelming with all the features, forced AI and updates etc on my laptops. Then windows 10 was no longer supported. I would have had to landfill 2 laptops (one with 4k screen and 16gm ram,). I am non technical and busy with my own commitments but curious about this thing called Linux. I asked AI and Gemini suggested Ubuntu , Linux Mint and for my very old 2008 laptop (4gb ram but SSD drive), Devuan. I had never heard of these. It explained Devuan is a fork of Debian created by VUAs. It will ruin flawlessly etc. I took the plunge, not knowing what system d was or anything really and realised Devuan needed a load of config . Using AI , forum help and perseverance I got the old 2008 laptop set up in a few days and now implementing on the 4k laptop. The topic of "init freedom" kept coming up. I had to figure out what that meant. At the end of it I realise that building one thing to do something properly and connect to next thing is brilliant, instead of building a complicated mess with no transparency. I apply that philosophy where I can in every day work now eg the way I will construct a spreadsheet or think about solutions. Without AI making me aware of Devuan's existence, and learning how to do/ not to do things, I wouldn't have got here. But then with AI it's easy to get lazy or degrade quality of output or principles. My wife played around on my laptop and found it so transparent, and thought it was so cool that devuan was founded on values that stand against corporate greed, and a reminder of human ability to be more conscious in choices instead of illusionary freedom, she is switching to Devuan on her laptop. Maybe I am not a desired user of Devuan (not sure), given my use of AI in the story, but the principles behind the distro are what makes me use Devuan as my only operating system of choice (plus it runs absolutely brilliantly. I haven't had a single crash, screen freeze or forced restart. Laptops are running blazing fast. Bluetooth is flawless. WiFi connection is instant and way better. Performance is amazing, wireless printer connects and locks in, no cut out. Nothing undesired, nothing forced on my laptop). The excellence should never be compromised.

Last edited by SS (2026-08-17 18:01:04)

bai4Iej2need · 2026-08-28 17:13:44

https://youtu.be/QJUBKYuVXUc?t=1925

AI gives you a statistical view of a program of "how it could look like, but no function guaranteed."
You have to give a software functional specification so detailed, that you have the job done yourself.
Do A first
Do B second
use the result string for Processing in C
output C Result to a device / file / stream what ever you need.

And then you have anonymously transferred all your know how to an mulltibillion corp which is eating it not even thankfully.

I tested this on several occasions with very small snippets of specifications.

I will not give them my money and my know how.

Xenguy · 2026-08-29 01:34:59

Paragraphs are a thing; please use them.

chomwitt · Yesterday 14:17:28

I remember an older 70s 80s era 'hype'. Breaking away from big brother, big corporation controlled mainframe central computing. Stories of hippies and love making hackers that fueled the era of Personal liberating computing .., hackers that mocked Biil Gates legal tandrums and power ! .An era were hollywood was selling fantasies of a teenager hacker hacking Joshua (a military supercomputer playing '''games''') with his IMSAI home computer to save out his girlfriend and the ...planet from nukes (a threat that's still over our heads.. while 'we' still think as a good idea giving more 'control' of our lives to "AI' agents..) I wonder how many macintosh would apple had sold in the 80s if they saluted as nazis and were promising only access to the biggest and improved mainframes controled of more conservative war mongering states and the bigger corporations on earth How tides turn...

Back to our current times (that feels like coldwar 60s..).

@August 27, 2026 by Drew DeVault's post on SourceHut policy and it's community discussion on the matter . DeVault's express an ethical view of computing that is against not the the LLM tech but against the forces that control it currently that exhibit blatantly criminal & exploitative behavior against the open source community and workers on general . Also he thinks that enviromental speaking the way LLM is used is accelarating our's planet destabilization. DeVault's has in the past expressed similar concerns against crypto. Of course it goes without saying that once upon a time when RMS or even Linus voices were critical again 'sensitive' corporate forces that drive the LLM bandwagon they were silinced as pedophile and autistics ...

I like that DeVault's if framing in as LLM. It's not AI . Not by far far farf far.. . AI is an ultrahyped term hyped from the ultra-hype engines of our times .

I prefer to call 'AI' apps as Natural Language Interfaces of Semantically Indexed Public and stealed Data stored on Private Data Centers (NLI-ofSIPD-onPDC) (mainly data from what we used to call internet). That name highlights equally all important aspects of the tech we discuss. Not just marketing terms. How smart would LLM algorithms be without the massive hoarding of memory ? (*) .Or without a constant stream of massive public and stealed data. And why could not we achive similar results from distributed computing ? Should we call from the dead the napster era ? Why can i not lend my 2000euro PC's (128GB RAM and 20TB SSD , latest GPU) resources to a inference LLM farm as i find fit and i must access a centrally controlled datacenter from my 2000euro PC (with 16GB and 1TB ssd and ondie gpu)? We have fast internet to see 4K netflix but not to pull a distributed LLM ? So internet (DNS,TCPIP etc) is not a tool for collaborarion anymore ? Isnt distributed computing an expression of freely engaged collaboration ? Federated LLM anyone ? What if i want in my local community to run a LLM on our 200hundred archive of our newspaper archive ? Must i upload the community's data to the cloud (another person OWNed hardware) since they have taken all the comp hardware making it expensive for all of us?

So after social-apps we get another service that further breaks internet. Another walled garden proposing another Faustian bargain that goes like :

- Be part of my plunder on the cyber-commons , enjoy a piece of the cyber-spoils , enjoy a better intellectual power BUT.. DO not challenge me ethically or question my origin. DO NOT ponder HOW i got so powerfull. Believe and spread the news.. that it's an innovative ''AI' tech. Not a 70s tech that needed internet public data to tap on and millions of hoarded gaming gpus and ram to process it....

(and why not ? because if you do the house of cards will fall and the sloppiness will go again at the bottom)

Do you see the irony that that particular faustian bargain has ? Your borrowed intelligence has off-limits..chains..restrictions. You are AS SMART as you are allowed to be. . Is this intelligence?.. or a form of loaned power on a leash.. conditioned. And it a leash that can be pulled so easily as you become addicted to it. So there is another reason to resist and do not indulge in that convenience. It can be taken so easily from you as they took your PC from you . And then what ? How would you program or troubleshoot ? And how resilient is a programmer that is becoming more attached to the internet and to a data center not only for programming but also for troubleshooting . Do they have 24/7 internet in countries in war or under censorship ? Do all countries have good 24/7 internet ? AHA... distributed computing you say ?? Really ?? It's more resilient you say ?? (ohh boy...) You mean resilient as internet was meant to be ?? Where knowledge (and it's processes that process it) is/are everywhere and not localized in certain super fat nodes ?

Or the more sinister variation of the faoustian trade -"Enjoy my cyber-spoils because if you dont , you co-worker will do and eat you first" . That is faoustian trade similar to RMS's : Dont trade your freedom for a convenience.., Which is a variation basically of B.Franklin's quote . But in that case i think a better one is : Dont turn your back to your society in order to gain personal powers that they were offered to you freely from outside. When power is offered to us we grab it. We grab it to position us better than the others. But would you grab a lost wallet's money ? That is a test for the cohesion of a society and the whole 'scene' personaly gives me Roman Arena vibes where you are a gladiator and the Roman throws at you a better weapon . If i were in the Colosseum i would grab it instantly . BUT if i was a member of a society that i like to live in i wouldnt try to increase my survival chances using a mystical power that i found in my backyard.

One more reason then to access less internet in a p2p fashion but easily indulge to a piece of convenience (in RMS's wise words) : Lets access SIPD stored on PDC accessed from NLI . In simpler terms : (it's interesting the news on companies aquiring huge quantities of old books... apparently the interesting bits on internet is getting dry!!!).. So we consume more data extracted from an era were internet valued sharing ethics. So the PDCs emit a strong extractive and exploitation ethics that will keep moving until the flows it tapped onto run dry. They need vast digital data flows to train on.

So big money with the help of big gov build in essence huge databases that track and acquire every data that move on public wires (internet at most) and integrated it semantically rich data centers and make accessible that data from natural language interfaces.

The question is . Do we use code 'synthesized' code from big-corp big-money controlled databases-datacenters ? Is it bad ? If so more bad that using a search engine like google ? Dont search engines paved long ago that way to create huge data sets and semantic integration of it ? How different is to script google search presenting you a multidute of code snippets aligning with a certain goal from using a more advanced Natural Language controlled script to do that for you ?

Lets look beyond the hype. Memory and gaming gpus are hoarded in a spirit that evokes war-time economics. Where is the tech innovation here ? Even the Moore mantra of shrinink process nodes has faded away. Intel , samsung , tsmc struggles as 2nm , go chiplet and sell hype. No innovation . Just massive - planetary scale hoarding and control and ultra-rich narcissists running amoc being lured by tigthly-state-controlled economies like Russia and China !! (state-economies that supposedly west is preparing to fight!.Way to go.!! ) And massive data centers need massive flux of data. Since when 'AI' dominance 's semantics are equal to stealing and hoarding citizens data ?

(*) A significanf aspect of LLM research is on training on Large Data Sets.

During the 2000s, with the rise of widespread internet access, researchers began compiling massive text datasets from the web ("web as corpus"[6]) to train statistical language models.

(@)

Isnt that paradoxical ? Give me more data and i will be smart Strange. Being smart means access more data ?? Could it be that we'are entraped constantly in the info flows of a mind of a glutonous data piggy and we try to reason in his terms ? Chomsky one theorized on an internal language aquisition device that help even kids grasp language's hidden structure from the smallest data set possible.

refs:
Breaking the AI hype

Last edited by chomwitt (Today 04:06:58)

golinux · Yesterday 14:58:41

We are definitely lost down the rabbit hole . . .

The officially official Devuan Forum!

#1 2026-08-16 20:50:03

AI and programming.

#2 2026-08-16 21:52:39

Re: AI and programming.

#3 2026-08-16 22:37:04

Re: AI and programming.

#4 2026-08-17 00:19:45

Re: AI and programming.

#5 2026-08-17 01:20:10

Re: AI and programming.

#6 2026-08-17 01:28:26

Re: AI and programming.

#7 2026-08-17 01:35:40

Re: AI and programming.

#8 2026-08-17 01:39:09

Re: AI and programming.

#9 2026-08-17 02:06:14

Re: AI and programming.

#10 2026-08-17 02:17:17

Re: AI and programming.

#11 2026-08-17 02:36:38

Re: AI and programming.

#12 2026-08-17 07:51:40

Re: AI and programming.

#13 2026-08-17 11:05:36

Re: AI and programming.

#14 2026-08-17 15:36:01

Re: AI and programming.

#15 2026-08-17 15:58:21

Re: AI and programming.

#16 2026-08-17 17:01:56

Re: AI and programming.

#17 2026-08-17 17:06:30

Re: AI and programming.

#18 2026-08-28 17:13:44

Re: AI and programming.

#19 2026-08-29 01:34:59

Re: AI and programming.

#20 Yesterday 14:17:28

Re: AI and programming.

#21 Yesterday 14:58:41

Re: AI and programming.

Board footer