jarjoura
Joined 15 years ago
Joined 15 years ago
I think it is an amazing example of how labs should take all of their fundamental research and turn into products entirely designed, limited and scoped only to empower artists instead of obsessed with trying to bypass them.
To say that it's because Microsoft overhired is a gross mis-remembering of the last 20 years.
For starters, I remember way back in the late 2000s meeting an engineer from the office team. Their entire job was focused on making buttons. Microsoft also used to employee engineers focused soley on testing software, though I've heard that has changed.
Anyway, this is a company with decades long cultural roots for enormously large teams where each person is hyper focused on tiny little pieces of a much larger product.
Also, let's not forget that they bought up 1/3 of the gaming industry.
I don't mind that Microsoft is hurting, because it makes mid products, that always seem to fail-up. However, I do mind that the people are okay that the answer is to lay people off instead of, maybe, breaking the company back up into smaller companies that preserve the stability of the workforce.
That car costs $200k in the US.
Outside of North America, the cars are really good and affordable. So it's out there. We are clearly doing something wrong here.
Everything is weird now.
There is this belief amongst software engineers that humans are always part of the process and that AI prompting is just the new IDE. Yet, every new model release is even more optimized to replace programmers and is pushing us one more step closer to that reality.
Right now (Fall 2026), we focus on empowering agents and LLMs and that's the current challenge facing software. However, once we solve that, the LLM really can kind of take it from there.
I know I'm being dramatic when I say this, but it really is how I'm starting to become jaded from all of this AI tech. It looks more and more like we're becoming a world full of really brilliant humans who are bored because there aren't any intellectual problems left to solve. That is not a utopia. That seems like a dystopian nightmare.
Looks like software is having its Napster moment now. Except, instead of one entity dictating the rules of engagement and deciding to sue Napster and its clones out of existence, we have 100s of entities with competing interests.
IP has been a good source of revenue generation, that social contract is now broken. I've got my popcorn bowl ready though, the industry that just fought Google and Apple for better revenue sharing is about to take all that legal infrastructure they acquired and make a big mess of everything. Here's to the next decade hah.
AI makes this so extremely efficient in all kinds of domains that, even if you try to continue to collaborate, I'm not even sure what the collaboration would be for.
At work, I've already noticed we aren't working together to solve hard ambiguous problems anymore. The whiteboards are filled with jokes and random notes people never bother to erase. Since everyone is tasked with their own project, unless you make a real physical effort to check-in, it's just way too easy to disassociate from everything else around you because you're able to just keep going.
Digital hobbyist projects of all shapes and sizes are becoming lonely little islands with people dreaming up all kinds of things, that everyone else can now also do, so there's now, no incentive to share or seek out experts.
I would be curious how real-world hobbies, that require patience and physical skill might start to see upticks in participation.
From where I'm sitting, it's just turning python functions into boxes and instead of write the function yourself, you drag from the output of one box to the input of another. For 2 or 3 boxes, this is cool, but I opened up a professional workflow and was taken into a view with 100s of boxes and wires all over the place. Uhh, ok?
For myself, I'd rather just create my own python environment, write some quick pytorch or mlx calls, wire up some cli to it and share that in a GitHub.
Rust is what you use when you want to maximize on the runtime performance. It will build software that requires significantly less memory and maximize the CPU that will save time and energy. It's a system language for writing software meant to access the hardware.
It was only in the last 10-20k years or so where farming emerged.
I don't know, but I think it's incredible how little time has passed relatively compared to our long history, of us being domesticated.
Apple and Meta are the only two big companies really pushing the frontier of this space, and everyone else is trying to figure out what to do with the technology. World Labs had the best marketing demos and showed the world what you could do with it, but it's all been vaporware from where I've been sitting.
I'm not super familiar with Nuke, but isn't ComfyUI the open-source version of that? If true, that's like saying Blender is making Maya obsolete, and maybe true, but Blender has its own community and so does Maya.