yea but i acknowledge that in my post, and explicitly state I’ll come back once I’ve watched the video. And on top of that, I actually read the OP.

Anyhow, for the sake of those who do open threads on this site, I ask that you do please read the OP before posting. It’s very frustrating when you have an idea for a topic of discussion be completely derailed by users who have a knee-jerk reaction to the title.

Vibe coding is just as bad as GEN AI and I am not watching a video of a dipshit who uses AI on his own channel and bullshit clickbait thumbnails. They have nothing to offer. There is nothing to discuss. 

I've said it before: I'm out. It was a love affair that began as a toddler, playing my dad's Atari-2600 (or thinking I was), and then fully blossomed when my mother bought me my NES Action Set. There were a few dips, a view pits and valleys along the way, but it was an affair that lasted for decades. But now with AI, that love affair is over. At least with new games. I have amassed a small treasure trove of retro games and consoles, more than enough to fulfill a human lifespan. And there are also other consoles and hundreds of other new retro games that I have never tried that I look forward to adding to my collection in the future. I have a CRT, OEM hardware... and the most creative IP the industry has ever made. So it was a good ride, being a part of the modern industry. I thoroughly enjoyed it. But it's time for me to move on. I want no part of AI gaming. It's where I draw the line.

Well, Herbert Simon, Turing and Nobel prize laureate, did point out that artificial intelligence can evoke strong emotional reactions.

Video is more about shock that goes with "no software is safe from decompiling anymore" and potential ramifications of that, than anything else really. It's not very good video, it's first time I watched the guy (and he seems to be on...something), but it's interesting topic for actually someone much more informed and eloquent to make a study out of it.

I think you might just have sold it to me. I might just watch the whole video, I guess mainly to see how AI helps with decompilation, when decompilation was possible even before, at least to some extent (I'm not familiar to which extent exactly though).

I've already posted about this in a different thread I made so won't go into as much detail again here but man this stuff is getting wild. Closed source software in general is about to get hit by a nuclear bomb so I do wonder how companies like Adobe are gonna handle this new reality. The past few weeks have been crazy in terms of AI news and progress so it really is starting to feel like the singularity has begun with the slope beginning to trend upwards.

The progress this year has exceeded my expectations with everything going on like what's been going on in math lately and I've been optimistic on AI progress for a long time so I now fully expect 2027 to be a crazy year and likely when the public as a whole starts paying a lot of attention to this field. The current reality of rapid AI progress while society largely stays the same can't last much longer unless AI starts plateauing soon and everything indicates that progress has been speeding up lately so there's some real massive shocks ahead.

Yeah it's pretty notable how likely due to the title difference my thread on this got less responses but they were more reasonable. This thread so far is closer to the level of nuance one can normally expect over this topic.

Moving more code (compiled or not) out of the hands of customers seems like the obvious solution, the cloud probably being the most obvious way for that. It's definitely not a viable solution for all software, but software that can do so will probably more eagerly do just that.

That could help in the short term but long term I expect a huge transition towards open source. In general the cloud won't help when the data gets leaked after the company gets hacked into. The increasing hacking capabilities of these models could start causing some havoc in the years ahead.

I'm not believing security to be an issue until it's proven to be. Right now I have no reason to believe security to be unattainable, although there may well be some learning pains in the coming years while AI-powered attacks decimate organizations unprepared for it. After the initial shock, security should improve though.

Another alternative for the cloud that comes to mind though is extracting value from other things, like support and, uh, cloud services, and probably a lot else, but I think those two are some of the traditional ones. I'm sure there's some business opportunities there, even if you're doing open source. I don't think companies will be eager to jump into the open source bandwagon though, it's too much of a change of mindset. In the long term, this might change, but I just don't see companies doing open source without trying out all the other options first, so it'll take time, if it even ends up happening ever.

Anyway, exciting times for sure. I hope something good comes out of it too, because much of it also seems concerning.