And then again if someone actually knew what they were talking about they would know that ~90% of the pieces of office suite were AQUIRED from their original developers (excel included) much as most of the ms "line up". MS has never actually designed much fully on its own. A great deal of their stuff to this day are acquisitions from smaller one/two man dev's looking to get rich selling their ware to the mighty corp. Didn't even use/need the "precious" ("AI" crap) to figure that out either...
Why using hydrazine if you have access to ammonium perchlorate? Trying to gain 14% in power by risking cancer isn’t the smartest move, especially since perchlorate is relatively non-toxic (95% of it is excreted in the urine within 48 to 72 hours; so the body gets rid of it completely in a few days) and, the cherry on the top, it is absolutely not carcinogenic. Furthermore, building a solid-propellant rocket engine is infinitely less complex than dealing with pressurized tanks, plumbing, valves and compressors…
that does not really apply to software... ...because of that. software usually consists of more than a single (a+b)/2 or sprintf(). and while it does not matter if you need 5 minutes or 5 years to create something to make a "work", if a computer created it, you did not. it is sometimes copied, it is sometimes bought, sometimes even stolen - but in all cases made by humans. for 40 years. then LLMs became intelligent enough to mimic that process. they are owned by billionairs and they steal 99% of what they know. about what recently wrote in a different thread. that one of the dangers with AI is that people who create might have to stop sharing their new invetions. the organised thievery of the LLM megacompanies will soon enable them to grab everything within inutes and release it. so it will lead to scientists sellings things more often directly to companies instead of publishing a patend, and to software engineers which offer their new stuff exclusively as SaaS, no demo versions, no free versions. Last edited by a moderator: Jul 7, 2026 at 1:01 PM
Your brain rots if you don't use it, but that has nothing to do with AI. Let's assume you want an AI to implement a specific feature in a piece of software. You could prompt an AI: 'implement feature X.' Okay, but: What problem is this feature supposed to solve? What technical methods exist to solve this problem? Which method is appropriate in this context? How should the feature handle the problem? What parameters does this feature have? How does this feature interact with other features? Are there limitations/exceptions, and if so, which ones? How should the user interact with this feature? Does the user always have access to this feature, or are there exceptions/exclusions? How should the feature present itself to the user? How should the interaction with the feature feel? etc, etc... Those are some basic questions for which you have to use your brain to answer. And there are a lot more detailed and complex questions that can be asked and answered. And you know what? You can just prompt an AI to ask you these questions. You don't have to put in the effort of making up the questions, just the effort of answering them. As you answerthe questions the AI came up with, new questions will pop up in your mind on their own. When you build a house, it's rare that all the work is planned and carried out by you as a single person. You have a designer, architect, structural engineer, bricklayer, roofer, electrician, heating technician, plumber, painter, etc., etc. Everyone has their task. And if you take on just one of these tasks, you've already done and learned a lot. If you can take on more than one of these tasks, even better. But nobody expects one person to be able to handle every single task. All of these tasks require you to use your brain. If you say, however: 'Build me a house. I don't care how, as long as it's big, looks good, and is cheap,' then 1. you shouldn't be surprised when the result is a piece of shit, and 2. you won't learn anything from it (except maybe that it's a bad idea not to use your brain) Well... If, alongside the idea, architecture, naming conventions, structuring, etc. (so the components that are uncopyrightable anyway), the prompter also objectively and specifically dictates the stylistic decisions and intervenes with iterative corrections and if these decisions are of a creative nature rather than purely rule-based, thereby dominantly shaping the output, meaning if human influence prevails, then according to previous court rulings, your statement is highly likely not correct, and authorship in this case goes 100% to the prompter. At least, assuming that the legal principles derived from previous court decisions on generative image creation can be applied 1:1 to generative code creation. We simply don't know. So far (to my knowledge), there have been no court decisions specifically addressing code generation.
It's truly disheartening to see how the course of human life is increasingly being shaped by economic and technological interests that take advantage of our lack of information. When the pursuit of profit outweighs concern for quality of life, we must practice the STT (stopping to think) and, from there, try to act in the best way possible. I recommend that everyone read La Fabrique du crétin digital (2020), a French bestseller by neuroscientist Michel Desmurget, which argues that digital natives (Generation Z) are the first generation in human history with IQs lower than their parents. On the other hand, many try to counterargue that factors such as pollution, stress, and social and financial inequality would directly influence this (which I do not deny!). However, more generally, cognitive scientists warn of a growing risk of digital content saturation, impacting human attention, memory, abstract and creative reasoning. Therefore, the discussion will develop beyond artificial intelligence and focus more on its impact on individual human intelligence and how this will shape future generations. I also have a personal account related to this reality. My little daughter, now 5 years old, was born at the worst time of the pandemic, when the first COVID vaccines were becoming available. My wife, who is a psychologist and an administrator, works for companies, while I, as a musician, work from home. Because of that, I had to take care of my daughter for half-days on some weekdays and, at times, full days. Out of unfamiliarity with screens, I ended up exposing my daughter to tablet content (cartoons and music) to be able to cook, wash, and work, for about 3–4 hours a day. This happened when she was about one year and eight months old. The result was a speech delay. Shortly before she turned three years old, a neurologist we consulted explained that she had been exposed to screens too early, and that this exposure made her passive, eagerly waiting for that low‑interaction content. Precisely at that stage, a baby needs to look at human expressions and map emotions such as fright, sadness, disgust, joy, anger, surprise, and other, more complex ones. She also mentioned that the first studies on the impacts of screen exposure during the pandemic were being made available, focusing on children who had been monitored for speech delays caused by the lack of social contact due to isolation, etc. At three years old, we removed all screens completely (including television, which we reintroduced gradually at age four-four and a half, but still without any exposure to phones or tablets). The improvement was striking. My daughter began to speak, and as soon as she entered pre‑literacy, she simply took off! Today she talks nonstop, loves drawing and writing, and happily spends hours listening to music and playing her toy xylophone and keyboard (fortunately, any issues associated with speech delay or autism were excluded!). To sum up: we must be very careful with this exact moment, in which the so‑called "Digital Cretin Factory" seems to find no clear limits, whether through misinformation, neglect, or the selfishness of those more concerned with filling their pockets than with the consequences. Last edited: Jul 7, 2026 at 4:15 PM
This can be easily observed when a composer/sound designer uses middleware to mediate musical creation and its implementation in a game engine (via the creation of a sound-implementation logistics). In 80–90% of cases, the implementation problems detected are related to logistical deficiencies... the old problem between "the chair and the keyboard".
A modular approach to synthesizers A synthesizer manufacturer would need to organize its expertise in such a way that components—like oscillators, filters, LFOs, MSEGs, sequencers, arpeggiators, reverb, etc.—could be fed individually into an AI; with the right programming, the AI could then assemble these building blocks. The ultimate question, however, isn't feasibility, but payment. Perhaps synthesizer manufacturers are already considering how to keep making money in a saturated market. I’d love a "synth-on-demand" model—10 euros per module. "I'll take 5 LFOs, 2 step sequencers, 10 oscillators, wavetable synthesis, reverb, and delay—okay AI, do it, I'm paying." Oh, and a selectable GUI costs an extra 20 euros.
After all that positivity towards AI in this thread, I thought I´d save me some time and ask Claude to just give me the drilling template for the faceplate of the midi controller I´m currently building. Simple 16 knobs midi controller evenly spaced in a 4x4 grid. Thanks Claude. I guess 1+1 could be 3. Or maybe it´s 5. Anyone here has theDAW running? How´s it going?
