It’s been a weird week. I’ve been spiraling down a Westworld themed rabbit-hole, on a quest for answers pertaining to the emergence and uses of AI (Artificial Intelligence) in the Audio industry. Needless to say, I’ve had a few ‘AHA’ revelations along the way, coupled with more than a few ‘WTF’ moments of innate panic.
I don’t need to rant on about the moral pitfalls of dehumanizing everything but I do need to express light concern for my own future as an audio engineer, because AI in audio is re-shaping industry workflow as we know it and has the potential to render some careers in audio redundant.
Let me say that again: AI in audio is re-shaping industry workflow as we know it and it has the potential to render some careers in audio redundant. Just let that sink in for a moment.
You may think I’m having a dramatic knee-jerk reaction to my rabbit-hole trip but let me substantiate…
AI as a standard mastering solution
There are quite a few AI driven mastering solutions available nowadays but for the purpose of getting my point across, let’s take a look at LANDR. Let’s be honest, when LANDR was released in 2014, it had a really gimmicky appeal. Fast forward just a year and Warner Bros was dropping a cool $6.2 million investment on it. According to Billboard, by November 2018, LANDR had mastered 10 million tracks by 2 million users. Today it’s endorsed by major artists, producers, engineers and major labels the likes of Warner Bros, Disney Music Group, Atlantic and Interscope, who are using it as a standard practice at mastering stage.
With a 24hr track turn-around promise for a low, subscription-based fee it makes total sense.
Like it or not, it has the attention of the suits and it’s here to stay – after all, time is money in this hyper-paced industry. But does the end result sound as good as a live mastering engineer? Well according to a comparative shootout conducted by the team at Pro-Tools Expert, LANDR only lost to the live human mastering engineer by a mere 4%. And that was in 2017. With the release of their new AI engine (Synapse) in April 2021, LANDR has once again broken new ground and is offering improved resolution, smoother compression and better loudness control – in other words, it’s reaching the same level of capability as an actual live mastering engineer working on fancy analogue gear.
In a world governed by digital societal norms, in a world where time is money, in a world where consumerist playlists reign, mastering platforms like LANDR have truly cemented their place in audio. Would I love to learn to be a sh*t hot mastering engineer? Absolutely. Do I think it’s a viable career path? Nope, sorry, not in this day and age. AI driven mastering platforms have and will continue to re-shape audio in the music realm as we know it. It’s a matter of time before traditional manual mastering becomes akin to cutting analogue tape with a razor blade.
AI in mastering is just the very teeny-tiny-tip of the dystopian iceberg that we as engineers are beginning to find ourselves up against. As we move forward, with further advancements in AI deep learning mechanisms, will we find our roles being reduced to mere button pushers (more so than the 8-fader bank on an Avid S1)? Can AI take on the emotive and creative decisions that we make when we mix or edit? It turns out it can. But that is a word-splurge for another day.