Daily Briefing — May 27, 2026

At graduation ceremonies across the US, students are booing their commencement speakers. The reason: executives keep standing at the podium and telling fresh graduates that the skills they just spent four years and tens of thousands of dollars building might already be obsolete. Scott Borchetta, CEO of Big Machine Records, told music students at Middle Tennessee State University that AI is rewriting production as they sat there, and when the boos started, his response was essentially "deal with it." As far as reading a room goes, that didn't go great.

Both sides have a point. The students aren't wrong to be frustrated, and the executives aren't entirely wrong about where the industry is heading. A 2025 Harvard poll found that a majority of young Americans see AI as a threat to their career prospects, and these aren't paranoid kids — they're people entering a job market where companies are already citing AI efficiency as cover for layoffs.

The boos aren't really aimed at AI. They're aimed at the casual dismissiveness of established people telling people just starting out to simply adapt, as if adaptation is free and frictionless and costs nothing. It isn't, and the graduates know it.

OpenAI, Anthropic, and Google are all doing the same thing right now: hiring what they call forward deployed engineers and sending them to sit inside customer organisations. OpenAI even stood up an entirely separate entity, the OpenAI Deployment Company, specifically built around this model. On paper it sounds like premium customer success. In practice it says something about where AI actually is that the marketing decks don't.

The "AI as a utility" pitch has been the industry's favourite story for the last two years. Plug it in, turn it on, watch it scale. That story is appealing because utilities do not require hand holding — you don't need a specialist from the water company camped out in your kitchen every time you want a glass of water. But that's roughly what is happening here. The most capable AI labs in the world cannot get their products to reliably stick inside large organisations without embedding their own people to make it work.

The gap between a demo and a durable system is still enormous. The models are impressive, but the tooling around them, the workflow redesign, the organisational change management, the part where someone has to sit with a frontline team and figure out where the thing actually fits, none of that is anywhere near automated. It's deeply human work, and right now the labs are absorbing that cost themselves because the alternative is a graveyard of stalled pilots.

Spotify just made a pretty significant move. The platform announced a feature letting premium users create AI generated remixes and covers using real artists' music, all under a deal with Universal Music Group. The CEO's framing is interesting: this isn't about replacing artists, it's about creating a licensed, consent driven lane before the unregulated stuff takes over completely. Whether you believe that or not probably depends on how cynical you are.

The "slop" comment is doing a lot of work here. Norström is essentially saying the AI music wave is coming regardless, and Spotify would rather be the adult in the room than watch piracy and rogue tools eat the market. It's a classic platform play: get ahead of the chaos, negotiate the terms, take a cut. Universal Music Group's willingness to sign on suggests the labels see some logic in it too.

For the broader industry, the "AI versus artists" framing is starting to break down. The real question now is who controls the creative pipeline and who gets paid when it runs. That negotiation is happening right now, and it will reshape how music, media, and creative work gets valued for the next decade.