Glyph didn't convince me. (And I'm biased; I've sat in multiple sessions given by his dad 20 years ago about creating good code.)
Code can be pretty, but I don't consider it art. Just like I don't consider technical documentation art.
I much prefer being able to talk to my computer and tell it to make a Stream Deck plugin to control the lights in my office and create a weather app (and show the precipitation forecast on the buttons) in 10 minutes, rather than me futzing around for two days to create something similar. Or to create the world's first(?) Telemark ski game from my Strava dumps in 30 minutes.
Admittedly, my book and course sales are down quite a bit (which I would mostly blame on AI). I'm not sure what the future holds, but change and adaptation have been pretty constant in my 25+ year career.
I do have concerns about what the minimal knowledge requirements will be for product development in the future.
The problem with talking is it lacks rigor. It's fun to get you from 0 to 90. Then the last 10% sucks and you don't really understand how you got there.
It's an engineering problem humans and AI will solve together. The same discussion is happening in math. It's a question of meaning. In this case it's even more obvious. Fairly soon, your weather apps and your ski games are not going to be interesting. I predict within a year.
Everyone will be able to ride their own version of whatever app. They will increasingly all be good enough, so why does anyone care about your "first"? Who cares about a first when cloning is cheap?
The last 10% is always the differentiator and talking is not going to be enough. Even if it doesn't make it better for someone else, for you to care we will see people return to something like programming.
The rest of the vibe coding people will slowly drift back to being consumers, because they don't care enough and because when everything can be good enough the only thing that matters is what you want, not anyone else. If you care about something talking about it isn't good enough. You will talk to your weather app because really you don't care that much. It's rather transparently for most people a mere question of utility (do I need an umbrella today vs meteorological phenomena are so fascinating).
The optimistic part of your story isn't what you've accomplished it's that your "accomplishment" will be duplicated to the point of such eventual economy shattering irrelevancy (post bubble, post UBI) that no one will have to care.
What this synthesis of recentering human intent for activities a human wants to pursue for no good other than the intrinsic, "thing in itself" with LLM capabilities ends up looking like I'm not sure. But I am certain it will not simply be talking. Likely it will be something multi-modal, but one where users also invent their own programming languages from the axiomatic principles that most agree with their interests/intuitions.
> Everyone will be able to ride their own version of whatever app. They will increasingly all be good enough, so why does anyone care about your "first"? Who cares about a first when cloning is cheap?
Everyone is probably already able to make their own version of whatever app from 1995.
Even before AI coding tools, the number of people who could write a 1995-functionality app (let alone a 1985-functionality) one was massively higher in 2020 than 1995.
So "what's next" probably isn't "everyone just clones what already exists" - instead, it's probably going to be determined by what's hard to clone that someone clever spots and then sells. What's the "I could write Facebook in a weekend" of the future, where the ability to clone isn't sufficient? I think there's likely something beyond "network effects" that we haven't found yet.
One probably-wrong guess I've had would be software that itself configures itself so well around the user that the user could only trivially clone it for themselves, not for anyone else. The "agentic" version of "I could write the 20% of Excel that's useful easily, with a better interface!" -> "oh, not everyone uses the same 20%".
You may be right. For my part I don't agree. I think the scale will be so much greater than the jump from 1995 to 2020. The moat will be data, or anything derived from hardware/the physical world. That is the tie back into "the ability to clone is not sufficient". You can clone Amazon app but how will you ship a package? Especially as their dominance in that area rests on their monopoly power?
Eventually they will fall too. It will just take longer. I think really the reason I disagree is simply that AI will get good enough where necessarily no human will have an edge. The super advanced ones will be on the whole time ambiently and likely adversarially self prompting. Your great idea already exists somewhere out in latent space. They will either get their first or just copy you. You will not care about anyone's edge. When you care about something you will necessarily engage with it manually. I mean this in the definitional sense. How did we know Alice cares about task x? Because she did it manually. She doesn't really care about the weather, so she just asked the computer in enough ways to make her own version of that. In the same way you wouldn't think Alice cared about Bob if she delegated correspondence to him to someone else. You know she likes him because with all her needs met and all her knowledge of the weather, she still chooses to spend time with him. For everything else compute/intelligence as utility will have you covered.
> What's the "I could write Facebook in a weekend" of the future, where the ability to clone isn't sufficient? I think there's likely something beyond "network effects" that we haven't found yet.
The every other part of the business is what's hard. Oh, and doubly so if access to markets will be controlled by a few giant incumbent gatekeepers. We're heading straight into a new gilded age.
This is the same kind of nonsensical sentiment that people say when they go to a modern art museum and "don't get it" - they say, "I don't consider this art"
Art is more than your tiny brain considers art. Maybe you should ask AI if X is art and see what it says. AI Delusion Syndrome is rotting people's goddammmmmm brrrrrraaaaaiiiiiiins.
That's the great thing about art. Reasonable folks can disagree about what art is...
My career and actions are proof to me that although I appreciate code and spent almost half of my career teaching folks the "correct" way to write Python, I (personally) don't view code as art.
On the other hand, just because a group of people agree something is art doesn't make it so. Or heck, maybe it does. I guess I could spend years perfecting my unique "prison wall fecal impact technique" and as long as lots of people can identify the nuances of my unique technique, and as long as it makes them feel emotions, I'm sure it can be considered art and can find a home in a modern art museum. But then to paraphrase Syndrome... "[When everything is art, nothing is]"
...the whole point of Syndrome's speech (at the metatextual level) is that everyone is super and he was wrong for not seeing it. (Mr Incredible was also wrong for not seeing it in Buddy, though.)
Anything can be art. It doesn't have to be appraised as "art" to be art. An audience doesn't have to derive the intended emotions for it to be art. Art is not truly a product; it is a component of the human condition, and we simply box it up for ease of sharing sometimes.
Code can be pretty, but I don't consider it art. Just like I don't consider technical documentation art.
I much prefer being able to talk to my computer and tell it to make a Stream Deck plugin to control the lights in my office and create a weather app (and show the precipitation forecast on the buttons) in 10 minutes, rather than me futzing around for two days to create something similar. Or to create the world's first(?) Telemark ski game from my Strava dumps in 30 minutes.
Admittedly, my book and course sales are down quite a bit (which I would mostly blame on AI). I'm not sure what the future holds, but change and adaptation have been pretty constant in my 25+ year career.
I do have concerns about what the minimal knowledge requirements will be for product development in the future.