Will AI replace musicians? 7 Things robots can never do better

I have a love/hate relationship with artificial intelligence.

The idea of having robots do the things I hate doing is the most exciting thing since Heavy Metal.

But the prospect of having robots do the things I enjoy doing faster and better than I can is unsettling.

Hearing Peter Gabriel express his concerns about AI taking his job makes you wonder if you squandered your life practicing an instrument.

So I gave the issue some thought.

AI is a real threat to almost everyone’s job. The robots can already compose music, and the AI is still in its infancy.  

Yet, when you think about it, we musicians are among the luckiest because music is the language of emotion. A language AI will never understand (unless it becomes sentient, but let’s not even get there).

I would be way more concerned if I were an accountant, because while artificial intelligence can learn anything an accountant can do (there is no emotion in accounting. Numbers and letters, elements the machines are fluent in), there are things AI can never beat musicians in.

These are a few of them:

  1. Connect emotionally.

Have you ever listened to a song and been so moved by it that you felt a connection with the singer or the band?

Then you looked them up to find out who the people who sparked your emotions were?

It happened to me several times. It happens to a lot of music lovers.

Will music listeners have the same emotional response if the exact same sound came from machines with very little human input?

Some will, and AI music will almost certainly carve out its own niche.

However, there will always be people who want to communicate with emotions that have actually been felt by the source expressing them.

  1. Build a legacy

I’m a fan of bands like Rush, and I proudly wear it as a badge of honor.

I like Rush because of the music (which the robots may eventually create, though I believe it will take them a long time to write the equivalent of 2112).

But I’ve grown to love Rush and everything it stands for – the musicians, the fans, the concepts, the artwork, the lyrics – the whole package.

Even more subtle things, like the relaxed way Geddy Lee’s fingers move over the bass, and the late Neil Peart’s poetic way of speaking.

This applies not only to huge bands like Rush, but also to a small local band with a loyal following of 50 people.

AI will never, ever be able to compete with that band on those 50 people.

We don’t just love listening to music. We also love what the musicians represent.

  1. Have a history

Every band, every musician, and every song have a history.   

AI may come to be able to write beautiful lyrics like this:

Ragged lines of ragged grey
Skeletons, they shuffle away
Shouting guards and smoking guns
Will cut down the unlucky ones

I clutch the wire fence until my fingers bleed
A wound that will not heal, a heart that cannot feel
Hoping that the horror will recede
Hoping that tomorrow, we’ll all be freed

Sickness to insanity
Prayer to profanity
Days and weeks and months go by
Don’t feel the hunger, too weak to cry

I hear the sound of gunfire at the prison gate
Are the liberators here, do I hope or do I fear?
For my father and my brother, it’s too late

But I must help my mother stand up straightRush (Red Sector A)

But those lyrics take a higher level of significance when you understand how they were inspired by Geddy Lee’s mother experience as a Holocaust survivor.

Now it’s not just beautiful poetry. It’s a story of a human being with emotions like us.

Likes songs, musicians and bands have a story behind them. And many of us are interested in knowing it.

  1. Make Happy Mistakes

I had a conversation with Aditya Shukla of Cognition Today about his concept of “the human premium” in art. He says the errors make art “art”. Summarizing the conversation:

“Some of the most exciting aspects of musical “feel” are minor deviations from perfection. We are talking about a bend note on a guitar that is just shy of reaching its intended note. Then there is the drumming that has a mild change in tempo here and there. Then there is the vocalist whose pitch isn’t perfect and you see a note adjust itself in the music. All of these are happy accidents. Most of them occur subconsciously and circumstantially. For example, a vocalist gets mildly distracted, a guitarist’s leg gets caught in a cable, and the note wavers. A lot of these mistakes also come from imperfect skills. And let’s be real – hardly any musician has perfect skills in all domains, errors happen, errors get recorded, and errors enter our fond memory of a song.

AI can’t yet predict these errors and can’t recreate them because they just don’t have a predictable foundation. But the bottom line here is not that AI can’t predict the errors, it’s that these errors make the music feel “organic” and “natural”. The human body itself isn’t a precise machine so when it plays one thing twice or thrice, there will be subtle differences. Listeners subconsciously notice these differences and remark how a song felt different in a live concert vs. a studio demonstration vs. a Spotify release. These minor errors are not errors – they are variations caused by human biology, mental state, and circumstantial aspects, and they add to music’s overall feel.

Human emotions can create errors – a chord can sound aggressively dirty when played in an angry state vs. a calm state by changing how the hand and body muscles move. That change in the chord’s dynamics, and its impact make that musical variation exciting, and that’s just one unexpected random moment an audience can appreciate in human art, not AI art.

Another amazing aspect of musical errors is that while composing songs, a wrong note somewhere can spark musical creativity, and then the composing journey changes. Without those errors, a song could’ve turned out very differently. Again, these are errors caused by lack of attention, imperfect technique, circumstances, mood, etc.

For AI do to this exactly, the AI will have to be modeled as a human first and then train it to make music, there’s a lot of time before that happens!”

  • Pull faces
pull faces

A friend once put a tasteless colorant in my coffee and turned it green.

Did it taste different?


Objectively the taste was the same, but since I associate the color green with anything but coffee my subjective experience of the coffee was different. This is called influences in perception and Aditya explores this subject in detail in this article

I believe a similar principle applies when musicians express emotions with their faces (guitarists are the most famous for this, but in reality, you can see emotion on every musician’s face) and other parts of their body.

Do you imagine BB King with a poker face while playing his heart out? It feels like the green coffee, doesn’t it?

I don’t think the robots will be pulling this one anytime soon.

  1. Get arrested
AI getting arrested

Some musicians got arrested, charged and convicted for horrible acts like child abuse and murder. Rightly so this also ruined their musical career.

Other musicians however, got arrested for mischievous things like smoking weed and peeing in Police cars.

The arrest becomes part of their story, their legend. We enjoy reading about it and gossip with our friends what the wild men of Rock n Roll have been doing.

Will a robot be able to do that?

Surely not anytime soon, but the moment I see a robot smoking weed, or peeing in a Police car, I would be worrying about things way more serious than losing my job!

Note: This is not to suggest that getting arrested was an enjoyable experience for these musicians, or that they deliberately got arrested to sell records. A nod here goes to musicians who were falsely accused of crimes they wouldn’t even think of doing, like influencing their fans to take their own lives. In such cases, not only weren’t the arrests and the trials bad publicity for the musicians involved, but also a traumatic experience they didn’t deserve to go through.

  1. Be successful

Successful people are admired because success indicates hard work, ambition, patience and persistence. Traits we consider positive.

A robot can easily dish out notes twice as fast as Yngwie Malmsteen or Joe Satriani.

But while the robot can put its ability to play fast into practical use, such as building intensity before an accented note, it can never gain the admirations of millions who stand in awe at the fact a human being can do what seems beyond human.

What this means for musicians

Since I wouldn’t give up on music even if the machines took over the whole industry (the joy of playing the guitar itself is priceless), I have given some thought about adapting to a world where a machine can do the same things we do better and faster.

And I believe the answer lies in beating the machines on the one thing they will never have, human emotion.

And while the above shouldn’t be taken literally (such as by getting arrested, or making deliberately stupid faces when bending strings) it should get you thinking on adapting to a world where the machines are getting really smart.

How can you give more of yourself to the listener, musically and non-musically?

What is the story behind your songs?

What is your story?

You may consider giving a donation, by which you will be helping a songwriter achieve his dreams. Each contribution, no matter how small, will make a difference.

Leave a Comment

Your email address will not be published. Required fields are marked *