Science1 publisher3 min readPublished
Alberta's musician survey puts AI hostility at 87 percent of the 142 who raised it
Of 263 working Canadian musicians surveyed at the University of Alberta, 142 addressed generative AI and 87 percent of those were negative about it. Venue closures and streaming payouts run through the same answers.
The Scientist · Science desk
What happened
- The University of Alberta surveyed 263 working Canadian musicians about their experiences in the industry, and Brian Fauteux, a professor of popular music and media studies there, described the results.
- Of those 263, some 142 addressed generative AI in their answers, and 87 percent of that group viewed the technology negatively.
- Recorded objections included that the output does not fit respondents' definition of art, that the tools involve theft of copyrighted work, and that environmental concerns about land and water go unaddressed.
- Most respondents described a forced, ambivalent relationship with social media and Spotify, distrusting the platforms while treating them as necessary to reach listeners.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- constraint Quoted as a prevalence figure for Canadian musicians, the 87 percent claims more than the sample supports; the defensible reading is strength of feeling among the 142 who raised the subject themselves.
- exposure Writers of jingles and advertising background music are the group Fauteux expects has already lost paid work, and a survey of attitudes cannot size that loss for them.
- contradiction Fauteux treats the tools as mostly toys for non-musicians, while respondents fear artificial songs burying them on streaming platforms; the two readings locate the harm in different places and imply different remedies.
- decision For a funder choosing where to spend, the most concrete constraint in these answers is the supply of small and midsize rooms and how far ahead they are booked, not the software.
Eighty-seven percent of 142 works out to about 124 musicians who told the researchers something negative about generative AI [3][1]. Set that against the full sample of 263 and it is 47 percent [1][2]. The remaining 121 respondents, 46 percent of the survey, did not raise the technology in their answers [3].
That gap determines how widely the 87 percent can be generalised. The published account describes the 142 as respondents who addressed generative AI, and it does not report the question wording or how the 263 were recruited [19]. If the subject came up because musicians brought it up, then 87 percent measures the people who cared enough to write about it. The share of all working Canadian musicians who dislike the tools is a different number, and it depends on the 121 who stayed quiet.
The survey did measure the content of the objection, and that part is specific. Fauteux said "There's an aversion to it. Some see it as not falling into their definition of art" [4]. He also said "A number of respondents pointed to environmental concerns and ongoing issues with land and water, and problems with the music industry taking those concerns seriously" [5]. Others reported being "personally and professionally" offended by the existence of such tools for artistic practice, and described them as part of a longer history of copyrighted work being taken [6]. Suno, one of the tools named in the research, generates a complete song at the click of a button [17].
Earnings are where the evidence thins. Fauteux expects AI has already cut into some livelihoods, particularly among musicians who write background music and jingles for advertisements [9]. That is his expectation. The survey asked about views and did not measure income. He also said the tools are mainly in the hands of non-musicians playing around, generating funny songs for social media [8], while respondents worried that a flood of artificial songs on streaming platforms could drown out authentic artists [7].
The hostility sits on top of two older squeezes, and Fauteux put the sequence plainly: "Musicians felt the shift to streaming about 10 years ago, and haven't recovered or seen the benefits" [10]. He said "It's not like this fell from the sky and people are mad about it" [11]. Respondents complained about Spotify's low payments, the industry's reliance on Spotify metrics to evaluate artists, and an algorithm they cannot inspect [13].
The concrete shortage they described was physical space. Many venues across Canada have closed over the past decade, leaving musicians competing for a shrinking number of rooms [14], and the survivors are less willing to take a risk on a newer artist [16]. "A lot of these venues are getting booked further and further in advance. So to do things like planning a tour, it's much more difficult," Fauteux said [15].
What to watch
- Publication of the peer-reviewed paper with the instrument, question wording and recruitment method, which would show whether the AI item was prompted or volunteered.
- Any income data for Canadian jingle and library-music writers, the segment Fauteux expects has already lost paid work.
- A second wave of the same survey, which would let hostility be compared against a measured change in earnings rather than recalled ones.