Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Making

beej.us

Post reply on HN
Print view View original

181–190 of 194 posts

Re: Making

#181

Post by dosman33 »

AI is the industrial revolution finally arriving at the land of information and software development, full stop. Read Carroll Quigley's Tragedy and Hope if you want to understand whats happening and see a roadmap of whats ahead. What happened during the agricultural revolution, industrial revolution, and financial revolution is already happening to software development. IR brings cheap material goods to everyone in your civilization, not just the elite. Before, only the king and his court could afford to have skilled tradesmen hand-crafting clocks and pocket watches. After the IR, now everyone has a cheap wrist watch available to them for pennies. It doesn't mean everyone has the best stuff, but it raises the quantity of whats practically available to everyone at the base level and it does this cheaply. The timex wrist watch works fine and is only $15. Not everyone has a Rolex, and there will always be a market for artisanally crafted items that predate the IR. A lot of people don't care how the Timex works, they just need to get back to work from lunch on time. There will always be a market for those that desire (and have the money) to have a wrist watch that works the way they want it to work. In the future, artisanally crafted code will matter to some people who understand code. Most people just need a program to do the thing. Humans will still be able to outperform AI when it comes to making clever, compact, and beautiful code. Go to any local pottery show if you want to see the future. Everyone already has IR manufactured cups, plates, and silverware at home. Yet people still go out and buy hand-crafted versions of these items because they have more character in various dimensions they care about. On the downside, everything will become slop. But it increases the slop available to everyone, while smaller numbers of those who care will still buy artisinally crafted code.

But there's another downside too that Quigley talks about: the arrival of the railroad at your previously isolated village. Suddenly now you have food, medicine and people arriving at a far greater pace. Your towns demographic suddenly starts to shift as your townspeople start to live longer as knowledge can instantly penetrate from the outside world. People can leave much easier too which facilitates brain drain. The railroad brings the industrial revolution to your door. The countries that survived the best were the ones that created these revolutions and then exported them. If the railroad did not originate from your country then the railroad arrived at your doorstep quite literally before you were ready for it. I don't fully understand what this means for AI, but I'm certain this is part of the reason the US is trying to lead the charge. You absolutely do not want the railroad arriving at your doorstep: you want to be the one pushing the railroad out to your neighbors. The only guarantee is change whether you want it or not. If you want some control over that change then you had better be driving the change rather than just reacting to it.

Re: Making

#182

Post by shampoo_capital »

InitialBP wrote: post #55
wlesieutre wrote: post #53

Earlier quoted context omitted.

> don’t tell me I did not “make” it because it simply would not exist if I hadn’t created it. So like when I went to a restaurant last week and "made" a cheeseburger with no lettuce and extra onions, because it wouldn't exist if I hadn't created it. To me, making something can mean you did the physical work of its creation even if there's no particular creative expression, like "I made a peanut butter and jelly sandw…

The restaurant comparison doesn't really hold up. The parent didn't go choose from a menu of new apps that they wanted AI to create. They discovered a problem or a feature that doesn't exist on other platforms and they employed tools to create that. Does a product manager not get to claim they "made" something because they plan it out and employ others to actually build the finished product? Both of the examples the…

Ok, forget about the restaurant then. If I hire a private chef to make any food I want, do I get to claim that I made the food?

Re: Making

#183

Post by ModernMech »

austinthetaco wrote: post #176
ModernMech wrote: post #42

Earlier quoted context omitted.

What about a director of a movie. They aren't responsible for the creative performances but they are certainly given a heap of credit for having made the movie. I ask 10 actors for 10 scenes and I combine them into a movie I wanted. I ask 10 agents for 10 functions and I combine them into a program I wanted. What's the difference and why is the guy who used agents not credited the same way as a director? Maybe all th…

It's the level of effort and involvement (also if you listen to any director talk they will never say "I made", they will say "we made"). Similarly to AI, nobody gives credit to the guy who "thought of the idea for the movie", they give credit to the leadership of the movie who worked alongside the crew. Directors dont just give an outline of a movie and then step away from set, they put real effort in to every aspec…

> Directors dont just give an outline of a movie and then step away from set

But I don't think people using AI to write software give an outline of the software and then just step away? It seems like a very iterative process to me with a number of decisions on the order of those needed to be made by a director of a movie.

Re: Making

#184

Post by Willish42 »

I think another aspect is the deep understanding one gains from building the thing, and the idea of the thing being entirely the creator's and translated into "reality" through some kind of physical or mental labor. An author of fiction prose would have their vision in their mind or some other ideas of what the prose "should" be, and then created it with their own word choice, and tools. Similar for visual art. This applies even if parts of the creative vision were improvised while the creation was happening, but it started in the creator's mind and if faced with the final product, they could remember the small details and understand the work that went into it.

I think if I'm honest about when I look at LLM output for code, it feels close to a code review of code written by another person, where I spend a decent amount of cognitive effort figuring out those similar small details in reverse, without knowledge or memory of the specific motivations. And furthermore, if I were to read every single line of code, word in a written work, and pixel or brush stroke in visual art of some kind, it would feel like an artificial understanding, rather than something I can fully sign my name to as a thing I understood the ins and outs of while creating it.

I fully get that some people don't care about this, and see themselves more as "designers", but I think those are the same sorts of people who if given a choice, would rather "build" by leading a team of people and claiming credit for "building" as part of their leadership, rather than by doing it themselves by hand and getting a sense of fulfillment from that effort. Also, "agentic" workflows feel a bit like giving up even the leading / designing aspects. With Claude specifically, I'm often impressed (albeit other times often frustrated too) by the small design choices it made while building the code that I hadn't explicitly asked for but end up preferring once I see the output. This to me indicates I didn't really build it

Re: Making

#185

Post by inigyou »

subscribed wrote: post #179
inigyou wrote: post #171

Earlier quoted context omitted.

No, that is not what GP said. That is, in fact, the opposite of what GP said.

That's pretty much what they said: unless you "make" something, preferably by hand and without using advanced tools, maybe you didn't make it at all. They were using restaurant and customised burger example to refute the claim that using LLM to assist with building solutions (applications) is building and making. I only dialed their extreme to the further extreme, just on the adjacent plane.

no that's literally not what they said. Read again.

Re: Making

#186

Post by 2III7 »

skydhash wrote: post #110
2III7 wrote: post #61

Earlier quoted context omitted.

Many people have a disillusion that an LLM is something more than a tool.

Git is also a tool, but cloning a repo does not means I created the software.

Cool, you can also use a 3D printer/lathe/CNC machine to clone a mechanical patented physical item. Plagiary can be done with any existing tool/method, you don't need an LLM for that.

Re: Making

#187

Post by Balooga »

planb wrote: post #151
Balooga wrote: post #108

Earlier quoted context omitted.

When you create a work using a tool, you own the copyright. When you create a work using an LLM, there is no copyright. So then isn't an LLM less than a tool?

This is not correct and a common misunderstanding of the court ruling. The verbatim output of an LLM is not copyrightable. If you create a work using an LLM, this IS of course copyrightable.

When "using an LLM", means the human has to substantially alter the output of the LLM for the work to be eligible for copyright. Agreed.

But this makes the LLM tool different to other tools. If I author a novel using a word processor I don't then in addition have to substantially alter that output in order for the work to qualify for copyright.

[edit] It's like, how substantive would changes to J.K. Rowling's Harry Potter need to to be in order for a derivative work to qualify for copyright? Now apply that to the output of an LLM.

Re: Making

#188

Post by austinthetaco »

ModernMech wrote: post #183
austinthetaco wrote: post #176

Earlier quoted context omitted.

It's the level of effort and involvement (also if you listen to any director talk they will never say "I made", they will say "we made"). Similarly to AI, nobody gives credit to the guy who "thought of the idea for the movie", they give credit to the leadership of the movie who worked alongside the crew. Directors dont just give an outline of a movie and then step away from set, they put real effort in to every aspec…

> Directors dont just give an outline of a movie and then step away from set But I don't think people using AI to write software give an outline of the software and then just step away? It seems like a very iterative process to me with a number of decisions on the order of those needed to be made by a director of a movie.

The level of effort is minimal. It's not a quantitative definition, its qualitative. Also absolutely in no way shape or form is it anywhere near the level of decision making of a director. Your average director of some low budget B netflix slop is still making thousands of decisions per day, organizing teams, reading and writing documents, going over shot review with DP's. A director is the CEO of a film production, not some ideas guy. They require and utilize a technical understanding of all parts of the production.

Re: Making

#189

Post by bigiain »

networked wrote: post #102
Zolomon wrote: post #22

Earlier quoted context omitted.

To me there is a difference that lies in the words. E.g. `making` and `producing` both result in the creation of something, but `making` has a more active part while `producing` has a more supervisory part. This became more relevant to me when I began meeting producers as a software engineer in game development.

"Produce" is a good verb. "Direct", too, if it's a creative work and you give close instructions and make judgements calls. "Design" is accurate if you prompt with an idea about the form of what you want. "Make" gets confusing for AI-assisted projects because it carries a connotation of craft. It seems fully justified for a code project that is 90% you, 10% AI (you'd say you made something that was 90% you, 10% human…

> I find "make" potentially confusing with AI-assisted projects.

I'm feeling this right now. I have a working prototype of a project that has mirrors mounted on pan/tilt mechanisms driven by servos and controlled by an Arduino. I modeled the mirror mechanisms myself in OpenSCAD and 3D printed them, I laid out the wiring from my head, I assembled everything by hand. But I got Codex to write the Arduino code that uses a GPS module to get the location and time, calculates the sun position, and (with some manual calibration) aims the mirrors to beam sunlight 'pixels' in an x/z grid on the wall.

It works. But for some reason it's profoundly unsatisfying. I'm very unlikely to go from this 4 mirror prototype to the original dream of a James Web Space Telescope model in my courtyard with 18 steerable mirrors drawing low res 1 bit pixel art on my dining room wall.

Re: Making

#190

Post by bigiain »

subscribed wrote: post #179
inigyou wrote: post #171

Earlier quoted context omitted.

No, that is not what GP said. That is, in fact, the opposite of what GP said.

That's pretty much what they said: unless you "make" something, preferably by hand and without using advanced tools, maybe you didn't make it at all. They were using restaurant and customised burger example to refute the claim that using LLM to assist with building solutions (applications) is building and making. I only dialed their extreme to the further extreme, just on the adjacent plane.

'If you wish to make an apple pie from scratch, you must first invent the universe' -- Carl Sagan
Post reply on HN