You must have failed to notice that you were replying to a comment of mine where I gave a specific example of a book that I think wouldn't exist without Word (or similar WYSIWYG word processors), because you're asserting that I'm never going to see what I am telling you I am currently seeing.

Generally, when there's a new tool that actually opens up explosive changes and development of new products, at least some of the people doing the exploding will tell you about it, even if there's no direct way to track it, such as Darwen & Date's substandard typography. It's easy to find musicians who enthuse about the new possibilities opened up by digital audio workstations, and who are eager to show you the things they created with them. Similarly for video editors who enthused about the Video Toaster, for programmers who enthused about the 80386, and electrical engineers who enthused about FPGAs. There was an entire demo scene around the Amiga and another entire demo scene around the 80386.

Do people writing code with AI today have anything comparable? Something they can point to and say, "Look! I wrote this software because AI made it possible!"?

It's easy to answer that question for, for example, visual art made with AI.

I'm not sure what you mean about "accelerating technologies". WYSIWYG word processors today are about the same as Bravo in 01979. HTML is similar but both better and worse. AI may have a hard takeoff any day that leaves us without a planet, who knows, but I don't think that's something it has in common with Microsoft Word.

▲ spicyusername 3 days ago | parent [-]

I noticed.

Books written with WYSIWYG could have been written by hand just fine, it would have just been more painful and taken longer. What WYSIWYG unlocks is more books, not new kinds of books. And sure, you might argue that more books is new books, which is fair.

So it is with LLMs. We're going to get more code, more lesson plans, etc. Accelerating.

    Do people writing code with AI today have anything comparable?

Like every fourth post on here is someone talking about their workflow with LLMs, so... I think they do?

▲ kragen 2 days ago | parent [-]

People talking a lot about their workflow with LLMs are evidence of the cost. What I'm asking for is evidence of the benefit, the "explosive changes and development of new products".

Remember that old Wendy's commercial, "Where's the beef?"

Where's the meat? I want to see your meat, and you're showing me pastures.

▲ spicyusername 2 days ago | parent [-]

    explosive changes and development of new products

That's because that's not happening. Like I've been saying! The benefit is just that it makes coding (or brainstorming a vacation, or planning a diy project, or writing a dnd campaign, etc) a little faster... It's an accelerating technology.

I don't disagree that the hype around LLMs is overblown, but that doesn't mean the utility isn't tangible.

If you listen to the salesman, everything is always going to sound like it solves every problem, but that doesn't mean the products they are hawking solve no problems.

	▲	kragen 2 days ago \| parent [-]
		I see! Now I understand what you mean by "accelerating technology", and I understand that you're not in agreement with mmargenot's thesis. Thank you for having the patience to explain until I stopped misinterpreting you!