The summer of 2025 marks some interesting anniversaries for me. It was a little over ten years ago that I graduated from high school, and soon enough, it will have been ten years since I moved out to Colorado. Any day now will mark 20 years since my first drum lesson, which is more or less the moment I consider myself to have become a drummer.  

Naturally, it’s hard to be introspective these days, what with the messes mankind constantly seems to find itself in and is incapable of pulling itself out of. Suffice it to say that, ten years ago, this is not what I had pictured; I visited the Grand Canyon lodge after graduating, and it just burned to the ground. So, what better time than now to consider the existential threat of AI? 

I don’t have the energy to blast off into dissecting the digital disaster that seems to be unfolding in front of us. You can take your pick: The inability to tell what’s real and what isn’t; the havoc that AI is wreaking on the planet, and our brains, and the already imploding education system; how AI is clearly and completely at the mercy of its demented creators; the troubling, complicated rise of AI spouses; the bizare eagerness of AI companies to share chilling information about how their product will destroy us all; the fact that AI seems poised to steal everything; and, to that end AI-generated “art.” 

Again, I’m not prepared to launch an armchair psychoanalysis of what has motivated tech bros to upend the humanities. This infamous interview from the head of an AI music generator displays such a catastrophic misunderstanding of why art exists that it feels hopeless to even try to correct it.

So you know what, let’s roast some AI slop, because as disturbing as AI can be, some of this stuff has a long way to go. My most popular article continues to be my piece on “Aja,” so I asked ChatGPT to write an article about Steve Gadd’s drumming on the title track. Since it’s a pretty well-known song, I figured ChatGPT would have an easier time with it than most of the things I write about. There’s also a reasonable chance my own writing will be gobbled up and spat back in my face. I’ll put CGPT in block quotes because there is some lengthy commentary to provide. 

⁂

Sure! Here's a draft article about Steve Gadd’s iconic drumming on the Steely Dan song "Aja" — one of the most revered drum performances in modern music history.

Precision and Poetry: Steve Gadd’s Drumming Masterclass on Steely Dan’s “Aja”

When it comes to iconic drum performances in popular music, few stand as tall—or swing as hard—as Steve Gadd’s work on Steely Dan’s 1977 track “Aja.” A seamless blend of jazz sophistication and pop precision, Gadd’s contribution to the song isn’t just accompaniment—it’s architecture. In a piece filled with musical virtuosity, his drumming is the cornerstone.

Right off the bat, there’s a lot this single paragraph tells us. Well… it tells us a lot about ChatGPT, and less about Steve Gadd and “Aja.” At the heart of it all, this writing is both outrageously florid and very nondescript. It’s one thing to give sweeping comments like “Steve’s drumming is great,” but a lot of these adjectives are hard to make any sense of. What exactly does it mean for Gadd’s contribution to be “architecture?”

We also see the first of many instances of “it’s not x but y” phrasing. Then there’s, quite famously, heavy use of the em dash (—), which is a damn shame, because I love the em dash. Despite ChatGPT’s desire to show off its deep vocabulary, it uses “precision” twice, which feels kind of lame. The bit about the performance “swinging hard” is equal parts generic and incorrect — the song is not swung. At all. We will continue to see that the two types of analyses we get are either unhelpfully broad or simply inaccurate.  

The Setup: Aja and Steely Dan’s Pursuit of Perfection

By the time Aja (the album) was released, Steely Dan—helmed by Walter Becker and Donald Fagen—had developed a reputation for obsessive studio craftsmanship. They recruited only the best session musicians to realize their complex vision, often recording multiple takes with different players to find the perfect fit. For the title track, a sprawling jazz-rock epic, they turned to Steve Gadd, already a rising session star with a reputation for taste, control, and unshakable timing.

This is more or less true, and not really news to anyone familiar with Steely Dan’s recording philosophy. ChatGPT interestingly ignores the rumor that Jim Keltner was supposed to record this song. 

The Groove: Subtle Syncopation and Impeccable Feel

From the outset, Gadd establishes a slick but subdued groove. His hi-hat work is tight and articulate, serving as the rhythmic glue that holds the opening together. The early sections of the song demand restraint—an elegant pocket that supports the song’s flowing melody and shifting harmonic landscape. Gadd delivers this with almost metronomic accuracy, but with just enough human feel to keep it from sounding mechanical.

You might have to check both your blood sugar and cholesterol after we’re done here. This must be what the kids call glazing, although that feels inaccurate, since “glazing” implies that praise is being heaped onto something that might not deserve it. But how else do you describe this?

When ChatGPT tries to get particular, it’s quite uninteresting. What does it mean by “shifting harmonic landscape?” What, chord changes? And there’s the horrifically cliched preoccupation with drumming in time but with some metronomic leeway — you know, muh groove!

The Solo: Controlled Chaos in a Jazz-Rock Storm

About four minutes into the track, the energy shifts dramatically. Saxophonist Wayne Shorter unleashes a fiery solo, and Gadd answers with a drum performance that has become legend.

Now there’s some confusion, as “Aja” has two saxophone breaks and an outro drum solo, but ChatGPT seems to be lumping them all together. 

What makes this drum break so extraordinary isn’t just the technical difficulty—it’s the musicality. Gadd doesn't simply fill space with flashy licks; he composes on the kit. 

I’m only going to say this once, so pay attention:

w h a t

The typographically minded may have spotted that ChatGPT randomly switches between dumb quotes (that are "straight") and smart quotes (that are “curly”). Most WYSIWYG text editors are supposed to automatically turn dumb quotes into smart quotes, but they may not do so if dumb quotes are pasted in, so perhaps this is a clue for all you teachers out there to be on the lookout for.

He uses polyrhythms, linear phrasing, and metric modulation to create tension and release, threading his way through the changing chords and melodic contours.

Linear phrasing is approximately true, although it’s not the only phrasing Gadd uses in the solos. Claims of polyrhythms and metric modulations are cataclysmically off the mark. “Slop” doesn’t even begin to describe this — it’s pure, unadulterated channeling of all the blowhards who get online and think every cool thing a drummer does must be a polyrhythm or modulation.

“Changing chords” is also quite wrong, as all of the solos are over a Bm13 vamp, a vamp that ChatGPT does not mention.

One of the most jaw-dropping moments comes during a series of rapid-fire triplet fills, executed with machine-like precision yet brimming with improvisational flair. It’s drumming that sounds both spontaneous and meticulously sculpted—because with Gadd, it usually is.

More fluff about the balance between metronomic accuracy and le groove. The discussion about triplets is intriguing, but what exact fill is ChatGPT talking about? Is it the John Bonham LRK thing, or the “gaddamacue?” It’s anyone’s guess. Unfortunately, this is the closest we have come yet to specifics of what Gadd is actually doing, but it’s still unhelpful.

The Technique: Economy and Expression

Steve Gadd’s style on “Aja” is a masterclass in the economy of motion. He rarely overplays. Every stroke is deliberate, and every phrase feels inevitable in hindsight. His ability to make complex ideas feel natural is a key reason this performance is so celebrated.

There’s something very corporate and sterile about this language. “Economy of motion…” I mean, really, who’s getting excited about this? And “every stroke is deliberate…” there are literal mistakes on this track. “Masterclass” is also a buzzword that instantly makes me roll my eyes.

Gadd’s signature use of rudimental vocabulary—particularly Swiss triplets and paradiddle-based phrasing—adds another layer of sophistication. And his cymbal work, often overlooked, provides shimmer and dynamic shape without overwhelming the mix.

Once more, as soon as we try to get to the nitty-gritty, we land farther from the mark. I do not hear any Swiss triplets in the song. There are paradiddle-adjacent licks, but paradiddles are practically the punchline for Gadd’s rudimental knowledge.

The Legacy: A Benchmark for Drummers

Decades after its release, Gadd’s drumming on “Aja” is still studied by drummers around the world. It’s been transcribed, dissected, and emulated endlessly. Yet, like all great art, it remains slightly elusive. The notes can be copied, but the feel—the intuitive push and pull, the musical sensitivity—is uniquely Gadd.

Even Donald Fagen, famously demanding in the studio, was awed by Gadd’s performance. In interviews, he recalled the band watching in silence as Gadd laid down the explosive middle section in a single take, realizing they were witnessing something special.

To the aspiring writers out there, this is a great opportunity to link to another piece of work to at least give an impression of research.

I’ve actually tried to track down the exact source of Fagen’s incredulity, since Modern Drummer (November 1992) brought it up as well. Googling the exact quote gave AI another opportunity to show off its insight:

Radical. Tubular, even.

Final Thoughts

Steve Gadd’s performance on “Aja” isn’t just a standout moment in Steely Dan’s catalog—it’s a landmark in the history of modern drumming. It bridges the gap between jazz freedom and studio discipline, between technical mastery and emotional depth. In seven minutes, Gadd reminded the world that great drumming isn’t just about chops—it’s about conversation, creativity, and control.

Some additional “it’s not x but y” sentences. This writing has also gotten very repetitive; quite frankly, it seems like the last three sections were meant to be the conclusion, since each one is desperately trying to churn out some pointed remark about the song and its legacy. 

Alright, despite the word count, that wasn’t very insightful (although in ChatGPT’s defense, it was a first draft). Undeterred, I denied ChatGPT’s offer to write a podcast script for me (seriously); instead, I tried again, this time asking ChatGPT to “break down Steve Gadd’s drumming on the Steely Dan song ‘Aja’,” hoping that I would get some heartier information this time.

Steve Gadd’s drumming on Steely Dan’s “Aja” is often cited as one of the greatest drum performances ever recorded. It's a masterclass in musicality, technique, and groove. Here's a detailed breakdown of what makes it so remarkable:

🥁 1. The Intro & Main Groove (0:00 – ~1:45)

Feel: Relaxed yet precise, with a jazz-funk flavor.

• Hi-hat work: Subtle and articulate, played mostly closed with ghosted accents that shape the time feel.
• Kick and snare: Gadd plays a syncopated groove that blends elements of funk and straight-ahead jazz. The snare lands slightly behind the beat, giving it a laid-back pocket.
• Ride cymbal: Introduced later, played with a jazz touch—light, airy, with tasteful bell accents.

Takeaway: Gadd keeps it restrained, laying a smooth, confident groove. The complexity is in the nuance—not in flashiness.

Well, this is a bit different. In fact, it seems like this has come off an assembly line for maximum engagement. The article is now a listicle, with a distracting amount of bolded terms and goofy emojis to pin every item. We have also evolved to “it’s y not x” sentences, and, again, what sparse details that are present are inaccurate; I would not describe the groove of the intro section as syncopated. 

🔁 2. Transitional Fills & Textures (~1:45 – 4:00)

As the song moves through key changes and instrumental layering:

• Gadd uses linear phrasing (no two limbs hit at the same time) in fills to create flowing motion.
• He uses the toms melodically, weaving between rhythms rather than blasting through them.
• Fills act like bridges—they connect phrases and shape dynamics instead of drawing attention to themselves.

Signature move: Short bursts of 32nd-note double strokes on the snare and toms, creating rippling effects without interrupting the groove.

What can you even say? It’s nice to see something resembling actual detail, but this is still way off. Steve’s playing during this movement is relatively spartan, and there are no linear hand-foot fills, nor any of this five-stroke/six-stroke roll thing ChatGPT is trying to describe. Is it just throwing out random drumming concepts that are tenuously associated with Gadd?

It’s also pretty clear ChatGPT likes to use a lot of weird comparisons that are borderline nonsensical, e.g., “Steve Gadd doesn’t just have a deep relationship with his toms, he incubates them in his very soul.” This is only good if your audience has never read an article about anything ever!

🎷 3. The Saxophone Solo & Drum Break (~4:00 – 5:00)

This is the legendary part—the section that made drum history.

Sax solo (Wayne Shorter) kicks off a dynamic shift, and Gadd’s response is explosive:

• Triplet-based phrasing: Gadd employs fast triplet fills that cascade around the kit, sounding fluid and composed.
• Metric modulation: He hints at shifting time feels (e.g., 4 over 3) to build tension beneath the sax solo.
• Call and response: His drumming seems to converse with the sax—starting subtly, growing more aggressive.

ChatGPT is still confused about the exact solo. Claims of “call and response” are highly dubious. Worse still, metric modulations are now being described as polyrhythms. Its efforts to be more informative are, unfortunately, making it less so.

The iconic fill (~4:41):
A rapid series of triplet groupings using a snare-tom-snare combo that spirals rhythmically across the barline. It’s complex, but feels effortless. This fill alone has been analyzed and transcribed dozens of times—it’s a textbook example of phrasing across the beat.

I have to assume that ChatGPT is talking about the Gaddamacue. There is a timestamp, but it’s not for any particular fill, just the start of the first sax solo. Describing the Gaddamacue as spiraling over the barline is wrong, and off-beat phrasing is also debatable. If anything, the off-beat chord stabs make the beginning and end of the fills ambiguous. 

I would say that it’s difficult to imagine a human writer being this confident and this misinformed, but I am reminded a lot of Whiplash and the way the screenplay describes the title track as “this fast, with this many polyrhythms, it’s impossibly hard.” Or perhaps when Fletcher barks out crap like “we’re gonna sharp that 9th.” 

People are clearly impressed by this nonsense, but for this “Aja” article in particular, I really struggle to see who the audience would be. A drummer is more likely than not to ask something like this, and is also more likely than not to be the least impressed here.*

🧱 4. The Breakdown & Outro (~5:00 – End)

As the track cools down:

• Gadd returns to a more reserved groove, but continues to add delicate hi-hat embellishments and ghost notes.
• He avoids repetition—there are subtle variations in almost every bar, but they serve the music rather than showing off.
• His cymbal crashes and fills are placed with near-orchestral awareness, enhancing the song’s narrative arc.

If you’re wondering what that emoji is supposed to mean, I guess it’s a pile of bricks. It’s about as relevant as anything else we’ve seen so far. Is ChatGPT really enamored by construction analogies? Or perhaps it’s telling me Gadd’s drumming has left it feeling bricked up. Far out!

There’s not much more to comment on here, other than, despite talk of “enhancing the song’s narrative arc,” there’s still no mention of the giant drum solo that climaxes the track. 

🧠 Technical & Musical Highlights

Technique	Description
Linear Drumming	No simultaneous hits; creates fluid, flowing motion.
Triplet Groupings	Used in fills to create rhythmic tension and release.
Ghost Notes	Soft snare strokes that add texture and feel.
Polyrhythms	Implied use of 3 over 4 (and vice versa) to stretch phrases.
Dynamic Control	Extreme sensitivity—able to go from whisper-quiet ghost notes to thunderous tom fills seamlessly.

We get a cool little table now, although it’s much less cool if you bother to read it. Polyrhythms, by definition, do not stretch phrases, as the two rhythms are contained within each other. Perhaps ChatGPT is thinking of polymeters, but it still might as well be talking about a different song. And how exactly do triplets create moments of tension and release?

🎧 Listening Tips

If you're revisiting “Aja,” listen closely with headphones to appreciate:

• The internal dynamics of Gadd’s hi-hat—how he “feathers” it with micro-accenting.
• The way his fills lead the music instead of reacting to it.
• The subtle push and pull he uses around the beat without ever losing the pocket.

Micro-accent??? Check, please!

🏁 Final Thoughts

Steve Gadd’s performance on “Aja” is the gold standard for musical drumming. It combines:

• Technical command
• Rhythmic creativity
• Deep musical empathy

…all while serving the song, not just the drum part. It’s no surprise that even the most elite drummers—like Vinnie Colaiuta, Dave Weckl, and Neil Peart—have pointed to this track as a pinnacle of modern drumming.

Look, the song is great, but as far as a “pinnacle of modern drumming” goes… it is almost 50 years old at this point.

I was about ready to call it at this point, but ChatGPT made me an offer I couldn’t refuse:

Would you like me to include a transcription of part of the drum solo or a breakdown of Gadd’s kit setup for this session?

I couldn’t help myself, I just had to see what it would spit out. So I said yes. 

Great! Let’s go deeper. Here's a more detailed look at Steve Gadd’s kit setup during the “Aja” session, along with a partial transcription and breakdown of one of the iconic drum fills from the saxophone solo section.

🥁 Steve Gadd’s Drum Kit on “Aja”

While exact gear can vary slightly from session to session, during the Aja recording (circa 1977), Gadd’s setup likely looked something like this:

Drum Kit Configuration

• Drums (Yamaha or Ludwig custom kits)
  • 22" kick
  • 12" rack tom
  • 14" and/or 16" floor toms
  • 14" steel snare (Ludwig Supraphonic or similar)
• Cymbals (Zildjian)
  • 14" hi-hats (New Beat)
  • 18" crash
  • 20" ride (possibly a K Custom or A Medium)
  • Small splash or china (possibly for texture)

Another typographic oddity: Aja is now written in both italics and quotes, as if ChatGPT can’t make up its mind whether we’re talking about the album or the song.

As far as the gear goes, the AI is trying its best here, but this isn’t much of an answer, although no one might know the truth — Gadd has long been associated with Yamaha, but there are a lot of mildly credible rumors online that the kit for this session was a backlined Ludwig. 

There are no splash or China cymbals on this track, but the cymbals that would have been used were very likely Zildjian. Out of the two options (As or Ks), ChatGPT has it narrowed down to two. They were very likely Ks, although whether they were the North American Ks or the original Turkish ones is a topic beyond the scope of this post.†

Heads & Tuning

• Coated Ambassador or Emperor heads

I have to butt in now; they were more likely clear Pinstripes or clear Black Dot/Controlled Sound heads.

• Tuned for a balanced jazz-rock tone—resonant but tight enough for articulate fills
• Toms slightly muffled (possibly with tape or rings) for focused tone

Other Gear

• Wooden sticks, possibly Vic Firth SD4 or similar jazz-style sticks
• Likely played with traditional grip (which he favors often)

I don’t know if the exact grip falls under the category of gear. But it’s interesting to hear that the sticks were wooden. I wonder if microphones were used. 🤔

I actually was curious to learn when SD4s were introduced, so I tried to Google it, and…

Hmmm, so the Vic Firth SD4 drumsticks were not made before the Vic Firth company came into existence? Bitchin’!

📝 Partial Drum Transcription: Triplet Fill at 4:41 (Iconic Drum Break)

This fill occurs right as the sax solo reaches peak intensity. It’s a rolling, melodic triplet phrase that moves across the toms.

Here's a simplified notation-style transcription (in 12/8 time feel):

Notation (starts at ~4:41) — Triplet Fill
(each line = a beat; played over two bars)

|  R   L   R  |  L   R   L  |  R   L   R  |  L   R   L  |

| Sn  T1  T2 | T1  T2  T3 | Sn  T1  T2 | T1  Sn  BD |

Well, the timestamp is still wrong, but hey, this notation-style transcription will be a great complement to my text-style article! Don’t be perturbed by the fact that it’s apparently in 12/8 and written with triplets. 

I’m sorry, I can’t — nobody should be excited to see the return of early-2000s drum forum notation that leaves you more confused than before. Also, note how ChatGPT is telling you to play the kick drum with your left hand.

I mean, I guess it’s mildly impressive that ChatGPT cooked up something resembling drumming, but I feel like I’m talking about software in the early stages of development. Maybe ChatGPT (relatively) is, but many people are convinced that this is already enough to topple entire industries. What I’m sharing today is comparable to a recipe that tells you to boil rocks. 

(Triplets moving across snare → toms → snare + bass drum)

Breakdown:

• This is a classic Gadd linear triplet fill: no simultaneous notes, all singles.
• The melody flows downward in pitch, creating a "tumbling" effect.
• The final bar incorporates bass drum with snare to punctuate the line and drop back into the groove.

The fill as described is strictly speaking linear, but not in the more conventional sense of kick drum notes that are mixed in with notes from the hands. Ultimately, though, it doesn’t matter, since Gadd doesn’t play anything like this on the song. Moreover, there’s not much of a “groove” afterward, as there is hardly any strong time at all. Again, I have to wonder if we are listening to the same song. 

🧠 Technique Tip:
The magic lies in the even spacing and dynamic control. Gadd often ghosts the earlier strokes, then crescendos into the end of the fill—giving it a rolling wave feel.

I’ll be honest, this tidbit is actually pretty interesting, and I’m surprised ChatGPT brought it up.

You see, at the time, Gadd was based out of New York, so he had to fly out to LA to record “Aja” with the rest of the band. While tracking the song at Village Recorders, Gadd traveled to nearby Santa Monica beach and was taken aback by the motion of the Pacific compared to the sights seen at the shores of the East Coast, inspiring much of his playing on “Aja,” and I’m just making all of this up. 

🔍 Additional Iconic Gadd Moves in “Aja”

Time Stamp	Technique	Description
0:20–1:30	Groove pocket	Subtle hi-hat variations and behind-the-beat snare create deep groove.
3:40	Linear fills	Snare and tom combinations with no overlapping strokes.
4:10	Ride cymbal jazz phrasing	Soft but swinging ride hits that float above the groove.
4:41	Triplet break	Famous fill discussed above.
4:50–5:10	Orchestral phrasing	Drums follow the contour of the sax solo dynamically.

I wonder if I should ask ChatGPT for resources to help me learn the techniques of “groove pocket” and “triplet break.”

📌 Summary

Steve Gadd’s playing on “Aja” is a masterclass in:

• Linear phrasing
• Triplet-based fills
• Dynamic control
• Serving the song with compositional awareness

Even the way he lands back into the pocket after that wild drum break shows superhuman control. It’s not just what he plays—it’s how and when.

There’s nothing more to add, other than noting another use of “it’s not x but y” sentence structure. 

ChatGPT then presented me with one last offer:

Would you like a PDF of the full drum transcription for “Aja” or a video breakdown recommendation for studying the solo?

I just had to see what I would get, so I said “yes,” only to get an error that no such PDF could be generated. By then, concerned about whatever ecosystem I may have destroyed during my little experiment, I decided to end the thread there.

⁂

Alright, we’ve all surely come to the same conclusion about this — the AI writing was not very informative, either do to vagaries or tidbits that were simply wrong. The writing did have personality… way too much in fact. 

Not to get too off topic here, but have you ever tried reading a speech transcript from the current commander in chief of these United States? Here’s a good one:

Gettysburg. What an unbelievable battle that was, the Battle of Gettysburg. What an unbelievable, I mean, it was so much, and so interesting and so vicious and horrible, and so beautiful and so many different ways. It represented such a big portion of the success of this country; Gettysburg. Wow. I go to Gettysburg, Pennsylvania to look and to watch. And the statement of Robert E. Lee, who’s no longer in favor. Did you ever notice that? No longer in favor. “Never fight uphill, me boys, never fight uphill.” They were fighting uphill. He said, “Wow, that was a big mistake.” He lost his great general and they were fighting. “Never fight uphill, me boys.” But it was too late.

You see what I’m getting at? The aureate avalanche of adjectives, clunky phrases that are difficult to process (“such a big portion of the success of this country,” “no longer in favor,” etc.), information that’s probably wrong…

Whatever, I’ll rein it in. But this raises an important question, one that might be the most disturbing of them all: are people, in their heart of hearts, truly impressed by this?

Now, I have to be a little circumspect here, because I don’t want to belittle people who are enamored by AI, and writing is a chore for a lot of people — I didn’t enjoy it at all until I was about 16, and I remain unsure if I’m any good at this.

School didn’t help much. God bless my freshman year English teacher, but reading Fahrenheit 451 as a 15-year-old and having to write about how beautifully Ray Bradbury complains about TV and loud music and fucking sports is not going to make anyone excited about anything (it doesn’t help that the author himself can’t even seem to decide what his book is about).‡

Erm, sorry, I think I just said something about reining it in, so that’s enough criticism of preachy beloved stories.

A lot of my commentary today has probably come across as polemic, and maybe it’s not AI’s fault that a bunch of windbags won’t shut up about how this technology will put so many people out of work… but I’m not very impressed, not at the technology nor the wierdos who feel a need to brag about such things.

When ChatGPT wasn’t wrong, it was pretty banal, almost stereotypically so; ideas of groove, pocket, feel, taste, and musicianship have become punchlines in the drumming world because you can prattle on about how they are abstractly important, but you don’t have to bother explaining what they truly mean, if you even can. 

It stands to reason that LLMs will only get better, but with the amount of finessing I would need to do to get ChatGPT to write something more coherent, I might as well just write my own damn article from scratch. How would you even begin to take this in a better direction? It’s one of the many things I would like to posit to the minds behind AI…

Perhaps it was a bad demonstration of ChatGPT’s abilities, since the details of drumming are pretty niche, and many drummers aren’t even up to speed on them. Still, I’m not so sure — someone on the website formerly known as Twitter shared a ChatGPT “review” of The Rachel Papers by Martin Amis, and the results were pretty similar: a firehose of verbose language that seemingly exists to distract from the fact that nothing is happening. The only difference is that one of these requesters was actually quite enthralled.

Subterfuge seems to be a big part of the AI world. Many people can probably share how a certain grossness washes over them when they realize they’ve been had by AI, especially by some photorealistic image. Because why else would someone share such a photo and not be honest about its origins if not to deceive on a large, effortless scale?

There are entire YouTube channels out there now that are entirely powered by AI, and none of the people running these channels would ever admit that they need a computer to do all the work for them.

And then, of course, there’s AI “music” (there do not exist enough scare quotes to write such a phrase). I could write an entire article about this topic specifically, but I don’t know if that would be good for my mental health. This stuff seems to be popular, but such numbers could be driven by bots for all I know. 

Although I did get an education in psychology, it still feels beyond my grasp to speculate on the motivations behind some of these AI businesses. It couldn’t possibly be jealousy; tech is much more lucrative than music, as most of us are constantly disrespected by both streaming sites and venues that, by and large, don’t want to pay us anything more than chump change. Most of us simply feel forced to participate in this industry as is, hoping for the best.

(Nevertheless, it still feels like creatives have been getting under the skin of certain people for some time now…)

I like to think that music is a particularly personal thing, but maybe that’s a naive thought in a world where everything needs to be consumed first and foremost. What I do know is that I’m not sure how I would react to someone insisting that they have such a connection to AI “music.” How so? There’s no emotion to process, no deeper meaning behind the lyrics or any of the musical choices. The computer doesn’t care. 

What’s even worse are the folks who legitimately think they’re spinning gold by making a computer spit out something for them that only has the veneer of being a work of art. I can’t help but be reminded of those goofy ads for MIDI chord packs that were everywhere at the start of the pandemic. 

I don’t know what kind of producers were giving testimonials for these adverts, and maybe they just needed the money more than they believed in the product. But one of these talking heads stops just short of saying “if you don’t know anything about how music works, this is great!” How could someone be attracted to songwriting (or any art form, craft, or discipline, for that matter) while not wanting to learn anything about it? Such an audience might need to do some soul searching, because someone (or something) might be lying to you about how this stuff works.

Surely, nobody thinks using something like Google Translate to turn English phrases into French ones means that they now speak French. I don’t see how AI “art” is any different. Although, come to think of it, this analogy isn’t even a great one, since Google Translate can at least tell you what a particular word is in another language. You’re not learning anything from AI “art.”

One line you may have heard is that AI is making art more “accessible,” which might be true in the sense that you don’t need a large amount of time and money to “film” something, or perhaps create a video game. But if your game isn’t good, you might not know how to fix it as someone who doesn’t know anything about game design, and is not prepared to learn.

In an attempt to curry favor with our (potential) AI overlords, maybe AI isn’t all bad. Those stem splitter functions are cool, even though they still have a way to go. AI image upscaling is sensible; much like stem separation, it’s something that seems incredibly tedious to do by hand. AI optical musical recognition is another intriguing development, although one that hasn’t gone far, unfortunately. 

There’s a very good chance ChatGPT would have had a better time writing about D‑Day or something, but that really cuts to the heart of the matter — regurgitating WWII facts from Wikipedia to write your project for history class is not the same job as analyzing a piece of art. This whole experience should make it undeniable that AI does not understand art, and accordingly, its “art” is not real. 

Well, this is starting to spiral, and for one final time, I’m not here to draw conclusions about what AI makers and users think they are contributing to this universe. I was only (almost morbidly) curious to see if AI could write my articles, and the results are a profound “no.” It would be neat if I could tell ChatGPT to transcribe me a song in seconds, but it cannot, and I really don’t know if it ever will. Maybe it’s for the best.

Edit (September 2025): I had to reset my Google Search Console verification the other month, so I’ve been Googling my most recent articles to make sure they’re getting indexed. Upon searching for “Obverse I,” Google’s AI really came through:

Surely, things are looking up now!

* ^ Here’s another anniversary: it’s been a little over ten years since I first saw Whiplash, and despite my best efforts to pretend the movie doesn’t exist, I still (apparently) feel a need to complain about it when the opportunity presents itself. I certainly miss the days when Whiplash was the biggest thing I had to stress about…

† ^ Here’s a good summary of Zildjian’s (un?)surprisingly complicated history.

‡ ^ Before any nutcases see this as an endorsement for blowing up the US Department of Education, I’ll have you know that I was at a private school before college.