Anyone else noticed the increase of Vibe coded plugins flooding the market?

[chk071]

The boss of Germany's biggest software producer, SAP, announced that they will most likely switch 100% to vibe coding in the future.

Thought it's be interesting to put that here, to put the point of "professionality" and the alleged lack of, when it comes to vibe coding, into perspective.

That said, the biggest issue I see is that knowledge will be lost, when people rather learn how to use A.I. than to learn what the A.I. actually produces. What it also tells me though is that, obviously, there's already a big lack of knowledge, when vibe coding is such a thing, even in the most professional businesses.

[shonky]

That said, the biggest issue I see is that knowledge will be lost, when people rather learn how to use A.I. than to learn what the A.I. actually produces. What it also tells me though is that, obviously, there's already a big lack of knowledge, when vibe coding is such a thing, even in the most professional businesses.

This has been the marker of technological “progress” for years - the ongoing deskilling of the general population. Think I became suspect of this a few years back when there appeared a glut of beat/chord/melody creation software which are fine as idea starters but not great for understanding what works and why musically. If the creation is completely offloaded to machines then what exactly is the creator’s role in their art?

[Munin]

If the creation is completely offloaded to machines then what exactly is the creator’s role in their art?

I think it will be even more that which it already has been for some time: Taste. Kind of like that famous Rick Rubin clip:


After all it was never technical skill which made certain artists popular. But rather the intuition for a certain sound or type of songwriting that people want and enjoy. So I think this will become even more extreme. Maybe at some point Suno can create what is on the surface the perfect song, but it still requires human taste to distinguish if it genuinely would land with people. And the vast majority of people do not have the skill or intuition to judge this.

Of course, you can argue that at that point you not really "creating" anymore. Rather you are curating, and I think that's a fair assessment. However the world is already very much full of people who are respected merely for curating, not creating: In the music world, we call them DJs.

[jamcat]

That said, the biggest issue I see is that knowledge will be lost, when people rather learn how to use A.I. than to learn what the A.I. actually produces.

That knowledge was already lost when programmers got lazy and ditched writing assembly code for writing in high-level abstraction languages like C and C++.

[whyterabbyt]

That knowledge was already lost when programmers got lazy and ditched writing assembly code for writing in high-level abstraction languages like C and C++.

How many different processors have you written assembly code for?

[jamcat]

Just one

[billinder33]

I think this is true to some extent but IMO it is also due to there not being really any significant advancements or innovations for like, the last 10 years maybe? Like when's the last time a new plugin really blew you away or did something truly new.

The recent adoption of component modeling has created some releases that were mind blowing IMO - the Korneff Amplified Instrument Processor and the Relab 176 are (IMO) the first compressors I've heard that actually sound like hardware.

Krafter, TBPA clip, Lacuna, Minimal Evoke, and Noiseworks Voice Assist blew me away in terms of innovation.

So there's still incredible new stuff being released. The issue is that the percentage of great plugins is dropping against the overall market. I'm demoing new things almost weekly now, and deleting most of the things I demo. And not demoing tons of releases just because there are so many releases.

Point being that just like music, there are still great new offerings being released, but you've got to work a lot harder to separate the wheat from the chaff.

[billinder33]

Just one

Which is one more than probably 98%+ of all coders. I've been in the tech and tech adjacent industries my whole life and known of exactly one person that writes assembly - when I lived in San Francisco I played hockey with a guy who wrote assembly for hardware controllers in the defense industry.

[whyterabbyt]

Just one

Ah, so no assembly 'porting' experience then.

[ksandvik]

Writing assembly today for highly optimized cache lines and multi-core CPUs, not to speak of GPUs, is something left for compilers. Simple embedded systems are OK, but even there more and more code is moved to C.

PS: And yes, I got my engineering training writing Motorila 680x assembly .

[Re5etuk]

i dont think people got 'lazy' and moved to c / c++ etc , it was the hardware people writing for .
yes, c64/amiga/atari etc was 6502/68000/z80 assembly but those platforms died and the need to continue paying the mortgage pushed people towards languages suitable for ps1/ps2 / pc / xbox etc via megadrive/ n64/ saturn etc.
synth hardware seemed to move similarly .. with the need to reuse software from hardware platforms to easily transfer to vst's .. (though i've not worked in synth world.. just gaming) . processors got out of date and much more complex to code by hand.

[whyterabbyt]

Just one

Which is one more than probably 98%+ of all coders. I've been in the tech and tech adjacent industries my whole life and known of exactly one person that writes assembly - when I lived in San Francisco I played hockey with a guy who wrote assembly for hardware controllers in the defense industry.

I think it would depend a lot on what age they are. By the end of the 90s I think I'd used half a dozen or so.

You'll probably find that code portability, and easily reusable libraries are actually why high-level languages like C replaced most assembly code. . And I suspect most professional programmers who have ever had to adapt assembly code from one processor family to another (aka 'a total rewrite from scratch') would accept that.
The other real reason high-level languages took over, of course, was the increasing complexity of instruction sets; very few coders can actually optimise better than a good compiler. Its not 'forgotten' knowledge, it just became unnecessary and far more inefficient in terms of development resources.

You'd have remarkably little code and notably few coders in the world if the balance hadn't shifted to high-level languages.

(And if the goalposts ever move to encompass about modern software efficiency, one would be far more honest to point the finger at the tiers of code libraries, not high versus low programming languages. )

[Tiles]

I just loved non programming tools back in the days. RPG Maker 95 ftw !

[legendCNCD]

languages suitable for ps1/ps2 / pc / xbox etc via megadrive/ n64/ saturn etc.

Those platforms got assembly language written too, in some parts of the games. I remember SCEE programmer telling tales about PS2 also N64 got its microcode RSP/RDP and Megadrive is 68000.

[Bunny_boy]

I just loved non programming tools back in the days. RPG Maker 95 ftw !

1995? Youngun

https://en.wikipedia.org/wiki/Graphic_Adventure_Creator