Brian Eno's Quiet Decades of Founding Companies

Most profiles of Brian Eno open with a list of credits. Roxy Music. Bowie’s Berlin trilogy. Another Green World, Apollo, Music for Airports. The Talking Heads records. The U2 records. The ambient catalog. The list of artists he has produced is long enough that, in many of the obituaries the music press has rehearsed for him, the company-building thread does not appear at all.

That is a mistake. Eno has been on the company-building side of the practice for almost as long as he has been on the recording side. The companies are quieter than the records, on purpose. They have rarely tried to be growth businesses. They have, in some cases, never tried to be businesses at all in the conventional sense. But they are real, they have shipped real artifacts, and they are part of the body of work the rest of the cohort has been quietly imitating.

This essay is an attempt to write the founder-side biography. It is not exhaustive. It is the version we wish more critics of his catalog would notice.

Generative Music, Before It Was A Category

The clearest of the company-track moments is Generative Music 1 (1996), the floppy-disk release that came with the software needed to play it. Most listeners under thirty have never seen the artifact. It was distributed on three discs and required SSEYO’s Koan engine — a generative-music software package built by a small UK company that Eno had been collaborating with for several years. The release was small. The license fees the project generated were not the point. The point was the artifact, and the artifact was an early piece of software-as-record.

The phrase “generative music” was Eno’s. He had been developing it in interviews and lectures for at least a decade before the disc landed. What he was actually arguing was that the recorded artifact could be a system rather than a frozen performance — a piece of software that produced a different listening experience every time it ran, within constraints the composer set. The argument is now standard. In 1996 it was a fringe position.

The company side of Generative Music 1 matters because it is the version of the story most fans miss. Eno did not just publish the piece. He underwrote the company that built the engine, advocated for the underlying technology with a working artist’s credibility, and used his catalog as a marketing surface for what was, structurally, a software startup. The fact that the company did not become a household name is not the point. The point is that the founder-artist posture — using the artistic catalog as a credibility surface for a technical bet — is a posture Eno was running in the mid-90s, two decades before the cohort the publication covers caught up.

Headspace, Calm, and the Sleep-Music Category

The next chapter the company-side biography rarely names is Eno’s involvement, in various ways, with the sleep-music category that grew up around Headspace and Calm in the 2010s. He was not a founder of either company. He was, by any honest reading of the record, one of the longest-running advocates for the underlying premise — that recorded music can be designed to lower the listener’s nervous system rather than to capture the listener’s attention — and the catalog he produced through the 70s, 80s, and 90s is, in retrospect, the canonical pre-history of the sleep-music market.

The publication’s interest in this layer is in the company-side mechanics. Eno’s Music for Airports (1978) was, in commercial terms, a flop on release. It is now, by some metrics, one of the most-listened-to records in the history of recorded music, because it is the substrate of half the sleep apps in the world. The catalog he built in 1978 is, four decades later, generating revenue at a scale that requires platform economics to make legible. That is a founder-track outcome that took forty years to land. The fact that Eno did not himself capture the upside is part of the story. The fact that the artifact was correct is the other part.

The lesson the current cohort has been quietly absorbing is that the record can be a long-duration asset whose commercial relevance is decoupled from the release cycle. The founder-artist who makes a record that gets used as sleep substrate twenty years after release is, in commercial terms, a founder who has built an asset. The catalog is the product. The lesson generalizes badly — most catalogs do not turn into substrate — but the version where it works is worth studying.

The Generative-Art Years

The next layer the company-side biography rarely names is Eno’s generative-art practice, which has run as a parallel track to the recorded catalog since at least the early 1980s. The 77 Million Paintings installation (2006), distributed as software and shown as an installation in dozens of cities, is the most public artifact. The work is, again, a piece of software that produces a different output every time it runs.

The structural point is the same point that Generative Music 1 was making. The artifact is a system, not a frozen image. The system is sold and exhibited as a piece of art. The collector who acquires the work is, in the conventional sense, buying software. Eno was making this argument, in public, in interviews and gallery shows, while the conventional art market was still resisting it. The market eventually caught up. The artist who had been running the practice for two decades was, by the time the catch-up happened, the canonical reference.

Most of the founder-artists we cover are too young to have a forty-year arc. They are doing, on a compressed timeline, the same thing Eno did on a forty-year one. The compressed timeline is not necessarily a virtue. The forty-year arc is the version where the artifacts mature alongside the market. The shorter arc is the version where the market is younger than the artist and the artist’s bets are not yet legible. Both versions can work. The Eno version requires patience the cohort sometimes does not have.

What The Cohort Should Take From Him

A few things the current generation of founder-artists could take from the Eno biography that they often do not.

First, the company side does not have to scale. Most of Eno’s company-side bets did not become growth businesses, and the catalog has not suffered for that. The fact that the music economy now contains structural roles for the kinds of artifacts he was making is part of the story. The fact that he never tried to capture the platform economics of those roles is the other part. The current generation of founder-artists tends to over-index on capturing the upside of their bets. Eno’s record suggests an alternative posture: make the bet correctly, ship the artifact, and let the long arc decide whether the bet pays off in cash or in cultural relevance.

Second, the catalog can be the company. Eno’s working life looks, on close inspection, like a single project with two surfaces. The records and the companies are not separate practices that happened to share a person. They are the same practice expressed in two different artifact types. The recorded work and the software work are arguing for the same thing — the artifact-as-system, the recording-as-environment — and the consistency of the argument across artifact types is part of why the work has aged the way it has.

Third, the press will mis-cover the founder side for decades. The conventional music press has, until very recently, treated Eno’s company-side work as a curiosity in the margin of the recorded catalog. The conventional founder press has, when it has covered Eno at all, treated his recorded catalog as biographical color. Neither press has, in our reading, recognized the practice for what it is. The current generation of founder-artists should expect the same. The work has to be its own argument. The press will catch up when it catches up.

A Note On The Records

We have under-served the records in this essay because the records have been covered extensively elsewhere, and because the founder-side biography is the under-told story. We will say briefly: the run from Another Green World (1975) through Apollo (1983) is, in our reading, the most consequential run of records in the second half of the twentieth century, and the Discreet Music through Music for Airports arc is the moment the category we call ambient was operationally invented. If you have never sat with Music for Airports on a real listening session, do that before you read another essay about the founder track. The records are the substrate of the founder argument. They are also, on their own terms, the work.

The next installment of From The Studio To The Stack returns to a working artist closer to the publication’s contemporary beat. We are running the Eno piece because the cohort, in our reading, needs to be reminded that the practice it thinks it is inventing has a forty-year prior art it should be studying.

Linus Embry writes about culture and the people building tomorrow’s tools.