I love attending concerts, but musicians may think my upcoming words are heresy. I hope technology doesn’t reduce the need for live performers, but the technological future includes many possibilities — more than anyone can imagine. Prepare yourself.
We must start by revisiting the Hartford Wagner Festival, the enterprise of Charles M. Goldstein. He expected to begin a cycle of The Ring of the Nibelung in Connecticut this month. The details leading to its postponement until 2015 can be read on Norman Lebrecht’s Slipped Disc blog, A Ring Without Musicians, or the New York Times.
Mr. Goldstein knew a performance of Wagner’s four opera event would be too costly to stage with a conventional orchestra. According to the New York Times, his monumental effort to create a substitute began in 2005. He started to enter every note of music for the 15 hour epic into musical software, having purchased “access to the Vienna Symphonic Library, a collection of sampled sounds of orchestra instruments (played by real players).” The idea was to create a computerized version of the orchestra out of the orchestral bits. “Singers from the Metropolitan Opera were engaged for the major roles, along with young artists for smaller parts.” Mr. Goldstein collaborated with the musicians to establish tempos. The plan was “to set up 24 speakers to mimic the positions of instruments in (a genuine orchestra) pit.”
Publicity produced a backlash. Singers’ careers were threatened. The inflexibility and sterility of computerized music was attacked and words like “absurd” and “travesty” were written. The Hartford Wagner Festival’s website now lists a 2015 start date.
Is the battle over or is it only beginning? The truth is, the conflict began much earlier. The New York Times covered the story in 2003. Local 802 of the Musician’s Union picketed Broadway shows in response to the threat of fully electronic delivery of what was previously the instrumentalists’ in person artistry. Anthony Tommasini, the critic who described the conflict, thought their argument had some holes. Orchestral and choral amplification with electronic support, synthesized instruments, and body microphones of principal singers were already a staple of the stage, he argued. The Union won, but technology has advanced in the last 11 years.
The Vienna Symphonic Library (VSL) is a tool intended to allow the creation of a good, albeit recorded, orchestral performance for the purpose of the user. In the case of Mr. Goldstein’s planned public presentation, input from the singers expected to be on stage, as well as the “conductor,” would determine the interpretation of the composition. Thus, the VSL holds the possibility of something greater than getting actors to lip-sync. Nonetheless, critics believe it does not permit the nuance, spontaneity, and sound of a real orchestra. Moreover, some say, unexpected mishaps on stage during an opera (an early or late entrance by a singer, for example) would be difficult for the conductor to adjust to.
I consulted a couple of musicians familiar with the VSL and its use. How much time might it take to “assemble” a fully realized rendition of a major symphonic work like Debussy’s La Mer? As Mr. Goldstein’s example suggests, they thought the time involved would be far too long to compete with today’s marvelous live instrumentalists (even for a talented, computer savvy conductor familiar with the VSL and musical notation software). You can listen to Andrew Blaney’s version of the second movement of La Mer on the VSL website and judge the quality for yourself. A first-class orchestra typically accomplishes its own magic in a few rehearsals using traditional printed scores. Wouldn’t live musicians therefore be safe?
Not so fast. Once the notes and instruments are “in the machine,” another conductor could come along and tweak the performance to his or her satisfaction. Still, for now at least, a computer-experienced maestro would have to learn how to use the VSL. The time involved in preparing his interpretation would be large even after that.
Does the situation change when we look only at opera and its audience? Might some music lovers be willing to attend an opera of live singers and a digitized orchestra if the price were cheap enough? Perhaps some who live in the provinces would.
Imagine a generous donor purchasing loudspeakers, the computer, musical notation software, and the VSL (one time expenses) for a small community with a decent auditorium, thus enabling staged operas. There exists a plethora of talented young singers and competent conductors of high school, college ensembles, and community orchestras. We are not talking James Levine or Riccardo Muti here, of course.
A digital orchestra reduces costs after the original outlay by our hypothetical donor. Goldstein’s long effort to enter the notes into the software for his “Ring cycle,” once done, needn’t be done again. Of course, he would have to be willing to sell his work product for an affordable price or simply give it away. Alternatively, several small communities could band together to pay whatever price Mr. Goldstein would set, or hire someone to do the job of entering the notes for an agreed upon opera. From that point, it could be widely and cheaply shared among them, as digital music commonly is today.
The cost of such an arrangement would be far less than hiring an orchestra and paying a major conductor tens of thousands of dollars per performance. Still, the result would be both poorer and different, at least until robots and androids are far more developed than they are now; replacements, that is, for some or all of the musicians!
Did I say robots and androids? The latter are robots designed to resemble humans. I’ll get to androids in a moment. Nonetheless, we are already in the world of the “second machine age.” Erik Brynjolfsson and Andrew McAfee, two MIT professors, elaborate in their book, The Second Machine Age: Work, Progress, and Prosperity in a Time of Brilliant Technologies. Even Goldstein’s idea of a digital orchestra puts us well beyond the era of steam engines, a prime contributor to the first machine age.
Would an enterprise such as Mr. Goldstein’s have an effect on major opera houses? Much depends upon the audience’s perception of the product. If most opera lovers believe the human beings at the Lyric Opera or the Met are worth the ticket price, as I hope they do, their orchestras have no worries. If, however, a sizeable number of patrons don’t, downward pressure on prices would follow (or at least the portion of the admission fee attributable to the orchestra). Instrumentalists and their salaries may be more vulnerable to the technological changes because they are in the pit, unseen by much of the audience. The human contact between the listeners and the players is more easily realized by the singers and the chorus in an opera house, unlike a concert hall where the ensemble and the podium focus our attention. Nonetheless, I don’t expect the technology we have now puts a superb opera orchestra in danger.
Ah, but the future — a different question altogether. Scientists speculate about something called “the singularity.” The singularity represents the point at which a fully conscious machine with greater than human intelligence is achieved. Coupled with the advanced android capabilities of that time, many of today’s jobs might vanish, perhaps even those of violinists and tuba players. The MIT professors cite evidence of the exponential growth of technology in recent years, predicting acceleration as we go forward. Self-driving cars already exist, as do computers that can beat the best Jeopardy or chess players. Ray Kurzweil, in The Singularity is Near, made a 2005 prediction placing the singularity somewhere around 2045. Kurzweil is a member of the National Inventors Hall of Fame, won MIT-Lemelson’s half-million dollar prize for innovation, and received the National Medal of Technology from President Clinton.
If one lets the imagination go wild, the possibility of android orchestras, conductors, or singers is conceivable, regardless of the year they turn up. I didn’t say preferable, particularly since the friendliness of such machines to humans is also an open question. Nor is it clear whether they would be interested in music, in the artistic history of the human race, or in presenting concerts of the works of humans in the formal way it has been done until now.
Assuming they are well disposed to Homo sapiens, however, it might become possible to pour all the recorded performances of the conductor, Wilhelm Furtwängler, into a capable machine and generate imagined interpretations of pieces for which we have no evidence of his work. A Furtwängler performance of Beethoven’s Missa Solemnis comes to mind. Does this violate Furtwängler’s memory? Were Mahler and Elgar violated, respectively, by completed performing versions of Mahler’s sketches for his Symphony #10 or Anthony Payne’s elaboration of Elgar’s unfinished Symphony #3? And could an android “inspire” human or robotic musicians in the hypnotic fashion of some of the greatest conductors? Would inspiration even be required? Live long enough and perhaps you’ll find out.
IBM is already working on a machine to do something similar for physicians minus the inspiration “to better diagnose what’s wrong with their patients. Instead of volumes and volumes of general knowledge, the supercomputer is being trained to sit on top of all of the world’s high-quality published medical information; match it against patients’ symptoms, medical histories, and test results; and formulate both a diagnosis and a treatment plan. … IBM estimates that it would take a human doctor 160 hours of reading each week just to keep up with relevant new literature,” according to The Second Machine Age authors.
Today we have opera supertitles and a giant in-concert video screen display of the Chicago Symphony in the Ravinia music shed, its summer home. We have digital music at our fingertips and in our earbuds. Robots guided by surgeons work on delicate human body parts. Some operas and orchestras simulcast their work around the world.
Once men thought Jules Verne’s 1865 novel, From the Earth to the Moon, was silliness.
The MIT authors remind us, never say never.
[…] We don’t share his conclusions, but the advance of robotic music needs to be addressed and the case for live music vigorously restated. Here’s Dr Stein’s sample scenario. Read the full text here. […]
In reducing the cost of production, the digital orchestra could possibly open the opera to the masses.
Thank you, Rosaliene. The consequences of the Second Machine Age are both positive and negative, some suggesting a possible utopia, others quite dystopian. Indeed, many things that previous generations didn’t have are now commonplace: computers, cell phones, astonishing medical procedures, etc. Part of the problem with an android orchestra, should it ever come to pass, would be displacing human musicians. I suppose we have enough to worry about from a political and climate change and poverty standpoint, without having to think of being displaced from our jobs!
This is a real concern. So many of us have already lost our jobs to machines.
Indeed. The Second Machine Age is a look into a possible future, nothing certain. The authors are actually quite optimistic, but acknowledge the downside and believe that some part of the failure to recover from the recessions of this century is a function of machines replacing people. I think you’d find the book provocative.
Thank you for a well thought out article. I hope you will attend our 2015 performances of “Das Rheingold” to see and hear the result.
Many thanks. I’m sure it will be closely watched and listened to by a large audience. Best wishes.