There's already been multiple stories of AI deleting the code and the production database and the backups, but people are still YOLO-ing despite that. So you're right and it's not gonna stop people.
There's count down timer videos on YouTube with millions of views. Like, we live in such a time of digital excess that a video of a timer, thrown over the Internet, is just how some people do timers. I'm very confused by it all.
I mean, if you're living in the truck that you're driving across country, back and forth, and your parents have passed, where would you put as your address? You don't have an apartment you're renting, because that would just be a waste of money since you're never there.
Fully autonomous is a long ways off, but semi-automated can't get here fast enough. Drive a comma.ai equipped car over a long road trip or other advanced driving aid, and just feel how much less frazzled you are after driving for a few hours.
Painting the ceiling is tiresome and is one place an exoskeleton really helps. If someone can paint for 8 hours instead of 6, and be less physically bad off, the gain isn't measured in productivity, it's elsewhere.
What's changed since then is AI can handle that shit for you. As a nerd, sure, I can scp to a Linux server, but with AI I just need to give it control of my computer and it can deal with the details. Meta's Muse is amazing as a product, despite some teething issues.
In the long run, what matters is the maintainability of the code. And the security, and the performance. All that comes from the code. The most beautiful code with proper layers of abstraction is gonna turn to shit when business needs come into play and there's not enough time to rearchitect the system properly.
That's a really good question! The code to the old software, the first version of Adobe Photoshop, Microsoft Office, IdSoftware's Doom, are available on GitHub, and people have poured over them. Or more modernly, there's large open source projects out there currently. But most source isn't something anyone else would see outside of work because it's all proprietary software. Google's internal systems are a thing of beauty, but no one outside of Google can legally be shown them, so it goes unnoticed by the wider world. The closest thing we get to art appreciation thus is blogs like Raymond Chen's Old New Thing. He's a programmer with Microsoft that talks about what it was like to work on windows.
You can vibe code without knowing what a function or a variable is. What is an int vs a bool vs a string. You can absolutely build something useful with today's technology without knowing how any of it works underneath. If you know how it works underneath you'll have a leg up on someone who doesn't, but what's the opportunity cost of knowing how that all works underneath. What are you missing out on and not learning while you're learning and reasoning about the aforementioned 1 & 3?
Except, now that frontend design is now something an LLM can just shit out a reasonable looking webapp with buttons to click on your phone, why use a terminal? Sure, a GUI is never going to have the absolute composability of kubectl | grep that a shell has, but at the end of the day I want to click on the button and have it do the thing.
I expect that because smartphones are terrible for running ssh from, that, with our new codegen powers, we'll see some innovation in that particular space. I don't want to hunt and peck on my phone's keyboard to ls and cd, but a TUI over ssh that I can use my touchscreen emulating mouse clicks on may be the sweet spot.
It's the window manager, the dock; spaces. Why am I limited to 16 virtual desktops in a horizontal row? Why can't I have more than that in a different configuration? For people who want to tweak their workflow how they want, they need the ability to modify things as they see fit. Maybe we don't need the actual source, given how good LLMs are at decompiling, but it would be nice if it was source accessible so I could make it how I want it, and for everyone else to also have that ability, if they so choose.
Yes, obviously. How about your home town last or the last trip you took? What were your parents like? What was your childhood like? What did you do last week? AI could certainly generate convincing sounding answers to that, but ChatGPT isn't set up like that and would thus fail the Turing test.