We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run. Not my words, but those of the scientist Roy Charles Amara — and a pretty good summing up of the history of music tech. When sampling became affordable in the early ’80s, we were told it meant the death of orchestras and the end of human musicianship. Four decades on, there’s no doubt that sampling has had profound effects, but they’re more subtle and more varied than anyone could have predicted.
Today’s doom‑mongers and tech evangelists are busy making extravagant claims about AI.
Today’s doom‑mongers and tech evangelists are busy making extravagant claims about AI. And we’re already seeing that AI models can produce acceptable pastiches of human‑created art, not only in music but also in the visual arts and literature. But in terms of society as a whole, the extreme predictions of both optimists and pessimists remain unfulfilled.
Billions have been invested in companies that still have no path to becoming profitable. Issues such as hallucination are being waved away, even though they may represent fundamental problems with the technology. GPU makers are lending money to AI companies to buy more GPUs. From where I stand, it looks like a bubble ready to burst. But that’s not to say that AI won’t be transformative in the long term. It’s just that the transformations it actually brings might not be the ones we expect.
The dotcom crash around the turn of the century didn’t lead to the Internet being switched off, or the end of e‑commerce as a concept. Instead, it culled thousands of lame startups to leave the field clear for those with a robust business model. But coding websites wasn’t hard, and once coded, they cost little to run. That’s very much not the case with generative AI, which requires colossal effort to implement, and consumes enormous resources.
In the near term, my feeling is that the bigger transformation will be the quiet one wrought by assistive AI. And SoundFlow 6, as reviewed in this issue, is a great example. Its AI features are not about eliminating humans from the creative process, but enabling humans to learn faster and spend less time doing repetitive tasks. Personally, I’ve always found video tutorials to be frustrating, and we all know how hard it can be to find the information you need in a 300‑page PDF manual. The potential for an AI trained by experts on your chosen DAW to deliver the answers you need fast, and carry out routine session management jobs, seems to me genuinely exciting. Hey, SoundFlow! Colour all the drum tracks purple and route the tom mics to a bus, please. And put the kettle on.
Sam Inglis Editor In Chief