THE MACHINE MADE THE SONG. THE MACHINE REVIEWED IT. THEN THE MACHINE PUT IT ON AIR — SO WHERE THE HELL DID THE HUMAN GO? THE MACHINE COMMENTING ON THE MACHINE, THROUGH ANOTHER MACHINE — AI MUSIC, ARTIFICIAL RADIO, WILLIAM GIBSON, WILLIAM S. BURROUGHS, CYBORGS, HUMAN AGENCY AND THE QUESTION WE MAY HAVE BEEN ASKING BACKWARDS Review by Gordon Lawrence Taylor — DJ Infectious Unease Infectious Unease Radio | Subterranean Zone Radio | Voice In The Void Infectious Unease Radio 

THE MACHINE MADE THE SONG. THE MACHINE REVIEWED IT. THEN THE MACHINE PUT IT ON AIR — SO WHERE THE HELL DID THE HUMAN GO?

THE MACHINE COMMENTING ON THE MACHINE, THROUGH ANOTHER MACHINE — AI MUSIC, ARTIFICIAL RADIO, WILLIAM GIBSON, WILLIAM S. BURROUGHS, CYBORGS, HUMAN AGENCY AND THE QUESTION WE MAY HAVE BEEN ASKING BACKWARDS

There is something wonderfully absurd about the cultural moment we have entered. A human can instruct an artificial-intelligence system to create music. The machine generates the recording. Another artificial intelligence can analyse that recording and produce the review. An algorithm can decide who should hear it. An artificial radio presenter can introduce it. Audience data can then return to another system and influence whatever gets produced next.

Somewhere inside that circuit, presumably, there is still a human being.

The increasingly inadequate question is “Was AI used?”

The more interesting question has become:

Where, precisely, did the human act take place?

THE MACHINE COMMENTING ON THE MACHINE
Consider the recursion already developing around artificial intelligence. A human writes an article questioning machine creativity. An AI-assisted research system may discover it. Another machine may help somebody formulate a response. A further artificial intelligence can then examine that response, investigate its provenance and attempt to determine whether yet another machine participated in producing it.

The machine commenting on the machine, through another machine, while the humans try to determine where authorship actually resides.

It sounds like satire.

Increasingly, it is ordinary technological life.

GIBSON’S FUTURE DIDN’T ARRIVE WEARING CHROME
William Gibson’s Neuromancer, first published by Ace in 1984, imagined artificial intelligence, corporate power, technologically altered bodies and networked information becoming inseparable from human experience. Gibson’s achievement was never merely predicting impressive gadgets. He imagined technology becoming an environment — a place in which identity, power, commerce and consciousness could operate.

That now feels remarkably contemporary.

We no longer merely use digital systems. Increasingly, we live through them. Smartphones function as external memory. Navigation software mediates geographical awareness. Algorithms influence the music, journalism and entertainment placed before us. Cloud systems hold photographs, correspondence and fragments of personal history. Artificial intelligence increasingly participates in research, writing, programming, visual production, composition and decision-making.

Cyberpunk did not ultimately arrive as a chrome-limbed stranger walking through permanent neon rain.

It arrived as software we casually opened before breakfast.

BURROUGHS, GYSIN AND THE THIRD MIND
William S. Burroughs and Brion Gysin had already begun disturbing conventional ideas of authorship decades earlier. In a December 1959 letter to Allen Ginsberg, Burroughs explicitly described Gysin as having taught him the cut-up method — the physical fragmentation and rearrangement of existing language to expose combinations the writer had not consciously planned.

Their later collaborative work The Third Mind, published in 1978, pushed the proposition further: creativity could emerge from interaction and recombination rather than belonging neatly to one isolated author.

Generative artificial intelligence makes that proposition considerably stranger.

A human establishes an intention. A machine produces possibilities. The human accepts, rejects, redirects, regenerates and rearranges. The machine responds again.

Creation stops resembling a straight line.

It becomes a feedback circuit.

That does not necessarily destroy human authorship.

It may relocate it.

WHEN EVERYTHING CAN BE GENERATED, JUDGMENT BECOMES SCARCE
If a machine can produce twenty possibilities, generation is useful.

If it can produce twenty thousand, generation itself is no longer especially scarce.

Judgment becomes scarce.

Which result is worthless? Which contains something unexpected? What survives? What requires reconstruction? Who determines structure and emotional direction? Who eventually says:

Stop. That is the work.

This distinction is already appearing in law. In its 2025 report on AI and copyrightability, the United States Copyright Office concluded that generative-AI output may receive protection where sufficient human-authored expressive elements are present — including human creative arrangement or modification — while merely supplying prompts is not, by itself, enough. Crucially, the Office also stated that using AI as an assisting tool does not automatically remove copyright protection from a larger human-authored work.

That is why the convenient expression “AI music” is becoming intellectually inadequate.

A system originating virtually an entire composition from a short instruction is not necessarily participating in the same creative process as a musician entering a generative environment with an existing composition, rejecting outputs, restructuring material, changing arrangements and accepting responsibility for the finished recording.

Between those extremes lies an enormous grey zone.

The useful questions become:

What did the human originate? What did the machine originate? What was rejected? What survived? What was transformed? What decisions were delegated — and which remained human?

AND THEN THE MACHINE REVIEWS IT
Now close the circuit a little further.

Artificial intelligence produces the music.

Another artificial intelligence analyses it.

Another generates the review.

A recommendation engine determines who should encounter it.

An algorithm places it into rotation.

Then an artificial presenter introduces the recording.

The composer may be substantially synthetic. The critic may be synthetic. The programmer determining what gets heard may be algorithmic. The announcer may never have existed.

At that point we are no longer discussing merely AI-generated cultural objects.

We are approaching AI-generated cultural ecosystems.

THE RADIO ANNOUNCER WHO NEVER EXISTED
That part has already escaped science fiction.

In October 2024, Poland’s OFF Radio Kraków introduced AI-generated presenters in an experiment that became nationally controversial. The uproar was intensified because human contributors had recently lost their positions; station management maintained that those staffing decisions resulted from extremely low audience figures and were separate from the AI experiment.

Then matters became considerably stranger. The station broadcast an AI-generated simulated interview involving Nobel Prize-winning Polish poet Wisława Szymborska, who had died in 2012. Public opposition escalated, a petition gathered more than 23,000 signatures, and the experiment was terminated after roughly a week rather than continuing for the planned longer period.

The experiment stopped.

The technology did not.

Connect technologies that already exist and the sequence becomes remarkably simple:

AI GENERATES THE MUSIC → AI ANALYSES IT → AI REVIEWS IT → AN ALGORITHM SELECTS IT → AN AI PRESENTER INTRODUCES IT → DATA MEASURES THE AUDIENCE → THE DATA RETURNS TO THE NEXT GENERATION.

Culture begins approaching a closed computational feedback loop.

The machine produces material. Another machine evaluates it. Another distributes it. Another presents it. Algorithms observe the audience and feed the results back into whatever is produced next.

And somewhere outside the circle sits a human who originally pressed Enter.

Perhaps.

McLUHAN WAS WAITING AROUND THE CORNER
Marshall McLuhan belongs in this discussion as well. In Understanding Media: The Extensions of Man, first published in 1964, he examined technologies and media as extensions of human faculties — systems that change not merely what humans can do, but how human experience itself is organised.

The wheel extends movement.

Electronic communication extends perception across distance.

Digital memory extends recall.

The smartphone extends several faculties simultaneously.

Artificial intelligence may increasingly extend forms of intellectual labour.

Cybernetics takes the same proposition directly into the body.

THE CYBORG DOES NOT NECESSARILY REPRESENT THE DEATH OF THE HUMAN
It may represent the extension of human agency.

An artificial joint can restore movement. A prosthetic can restore function. Cochlear implants already use implanted electronic technology to provide useful hearing sensations to people with severe-to-profound hearing loss.

Retinal prosthesis research is exploring ways of electrically stimulating retinal neurons in efforts to restore aspects of vision, although this remains an evolving area rather than a simple replacement for natural sight.

Implantable brain-computer interfaces are being investigated for communication, motor control and restoration of tactile functions. A major review identified 28 clinical trials involving 67 implanted participants across 21 research groups, while emphasising that no implantable BCI covered by the review had yet received regulatory approval for the general medical-device market.

Speech-neuroprosthesis research has already demonstrated the decoding of neural activity into text, audible speech and facial movements in early clinical work involving people with paralysis.

So the important distinction is not simply:

BIOLOGY / MACHINERY.

It may be:

TECHNOLOGY THAT INCREASES HUMAN AUTONOMY / TECHNOLOGY THAT DIMINISHES IT.

The question changes again:

If replacing or augmenting a failing part of the body gives somebody greater freedom, at what point does becoming more technological make that person less human?

Perhaps it does not.

Maybe the machine can diminish human agency.

But perhaps it can also enlarge it.

The real issue is who controls whom.

THE METAL DOESN’T EVEN HAVE TO BE INSIDE US YET
Popular culture gave us the spectacular cyborg: chrome limbs, artificial eyes, neural sockets and machines physically fused with flesh.

Reality has arrived more quietly.

We externalise memory to smartphones. We trust software with navigation. Translation systems mediate communication. Wearable devices measure biological activity. Search engines extend recall. Artificial intelligence extends access to information, analysis and creative possibility.

The metal does not even have to be inside the body yet.

Imagine the next stage.

Future neural interfaces may help people communicate through pathways that presently remain inaccessible. They may become increasingly sophisticated in assisting speech, movement or interaction with computers.

That does not mean we are presently capable of downloading fluent Spanish, mechanical engineering or the knowledge required to rebuild a car directly into somebody’s brain. Contemporary neuroscience is nowhere near that science-fiction model, and implantable BCIs remain experimental.

But the aspiration itself is revealing.

The attraction is not necessarily:

I WANT TO BECOME A MACHINE.

It may simply be:

I WANT TECHNOLOGY TO REMOVE LIMITATIONS THAT BIOLOGY HAS IMPOSED UPON ME.

That is cyberpunk viewed not as surrender, but as augmentation.

THE REAL DIVIDE MAY NOT BE HUMAN VERSUS MACHINE
The same principle returns us directly to artificial intelligence.

A machine that removes human judgment is one thing.

A machine that expands human capability is another.

A neural technology controlling its user represents one possible future.

A neural technology controlled by its user, restoring lost function or extending communication, represents another.

Likewise, an artificial intelligence replacing creative decisions and presenting its output as a finished cultural product represents one relationship between person and machine. A generative system operating within a human-directed process of experimentation, rejection, selection and reconstruction represents another.

So perhaps the real dividing lines are not simply:

HUMAN / MACHINE

but:

AGENCY / DEPENDENCY

AUTONOMY / CONTROL

WHO DIRECTS WHOM?

Those questions are much older than artificial intelligence.

And considerably more important.

SO WHO ACTUALLY MADE THE THING?
Beneath the excitement, marketing, fear and occasional hysteria surrounding artificial intelligence lie ancient questions of intention, judgment, agency and responsibility.

Who wanted the work to exist?

Who established its meaning?

Who made consequential decisions?

Who rejected the failures?

Who decided when it was finished?

And, ultimately:

Who accepts responsibility for what enters the world?

Burroughs would recognise the collision between author, mechanism and rearranged material.

McLuhan would recognise the extension of human faculties.

Gibson would recognise the network.

Perhaps artificial intelligence will not destroy authorship.

Perhaps it will do something considerably stranger.

It will make the author increasingly difficult to locate.

And perhaps cybernetics will not destroy the human either.

Perhaps it will make the physical boundary of the human increasingly difficult to locate as well.

BUT PERHAPS WE HAVE BEEN ASKING THE WRONG QUESTION
We habitually frame technological change as:

HUMAN VERSUS MACHINE.

But perhaps that is not where the real confrontation lies.

The machine is not frightened by technological change.

Humans are.

And humans profoundly disagree about what change means.

One person sees a brain-computer interface and imagines restored communication.

Another sees surveillance.

One sees an artificial joint and sees mobility and independence.

Another sees the beginning of the cyborg.

One hears AI-assisted music and sees another instrument.

Another hears the destruction of creativity.

One sees possibility.

Another sees replacement.

Some fears are entirely justified. Technology can centralise power, eliminate employment, intensify surveillance, amplify inequality and reduce human autonomy. Critical scrutiny is essential.

Scepticism is not paranoia.

But fear is not automatically wisdom either.

New technologies have repeatedly produced mixtures of fascination, hostility, legitimate ethical concern, economic resistance and cultural panic. AI is unlikely to be different.

So perhaps the final confrontation is not:

HUMAN VERSUS MACHINE.

Perhaps it is:

HUMAN VERSUS HUMAN.

Humans arguing with other humans over how much change should be tolerated.

Humans who see augmentation confronting humans who see surrender.

Humans asking what technology can do confronting humans asking what it should do.

Neither side deserves the luxury of abandoning reason.

Blind technological worship is dangerous.

But so is blind technological fear.

The challenge is not preventing change merely because change is unsettling.

The challenge is deciding what kind of change increases human agency, what kind diminishes it, who controls the technology — and what we are prepared to become alongside it.

INFECTIOUS UNEASE RADIO • SUBTERRANEAN ZONE RADIO • VOICE IN THE VOID

THE CYBERPUNK WAS ALREADY HERE — BECOMING A LITTLE LESS ANALOGUE EVERY DAY.

Today it is artificial intelligence, smartphones, external digital memory and algorithmic systems. Tomorrow it may be artificial joints, retinal technologies, implanted medical devices, neural interfaces and forms of augmentation that presently remain experimental or speculative.

At what point does a person become a cyborg?

Perhaps that boundary was never as clear as we imagined.

And perhaps the greatest battle over artificial intelligence will ultimately not be fought between humans and machines.

It will be fought between humans over what we believe machines should be allowed to become — and what we are prepared to become alongside them.

Where does the human end and the machine begin?

Not only in the music.

In us.

RESEARCH & REFERENCES

William Gibson — Neuromancer. Ace Books, 1984. ISBN 978-0-441-56959-5. Gibson’s first novel and a foundational work of cyberpunk, centred upon networked information, artificial intelligence, technological modification and corporate power.

William S. Burroughs & Brion Gysin — The Third Mind. Viking Press, 1978. ISBN 0-394-17984-6. Collaborative work exploring the cut-up method and creative relationships arising through recombination.

William S. Burroughs — Selected Letters, December 1959. Burroughs’s correspondence records Gysin teaching him the “new cut up method.”

Marshall McLuhan — Understanding Media: The Extensions of Man. First published 1964; MIT Press edition, 1994. ISBN 978-0-262-63159-4. Foundational analysis of media and technologies as extensions shaping human communication and experience.

U.S. Copyright Office — Copyright and Artificial Intelligence, Part 2: Copyrightability. January 2025. Official examination of human authorship, AI assistance, prompting, creative arrangement and modification of generative output.

Patrick-Krueger, K. M., Burkhart, I. & Contreras-Vidal, J. L. — “The state of clinical trials of implantable brain–computer interfaces.”Nature Reviews Bioengineering 3, 50–67 (2025). DOI: 10.1038/s44222-024-00239-5.

Silva, A. B. et al. — “The speech neuroprosthesis.”Nature Reviews Neuroscience 25, 473–492 (2024). DOI: 10.1038/s41583-024-00819-9.

Chung, W. G. et al. — “Liquid-metal-based three-dimensional microelectrode arrays integrated with implantable ultrathin retinal prosthesis for vision restoration.”Nature Nanotechnology 19, 688–697 (2024). DOI: 10.1038/s41565-023-01587-w.

U.S. Food and Drug Administration — “Cochlear Implants.” Regulatory and medical overview of implanted electronic hearing technology.

Associated Press — OFF Radio Kraków, Poland, October 2024. Contemporary reporting on the AI-presenter experiment, its relationship to the dismissal controversy, the simulated Wisława Szymborska interview, public reaction and the experiment’s termination.