Good day . A Vancouver songwriter answered AI music by recording her new song live with friends, which made her vocals almost impossible to fix later. She meant to do that. On the same Friday, AI music held its first red-carpet awards show in Los Angeles. Also today: Gen Z uses AI every day and some of them feel guilty about it, Universal Music wants fans remixing its artists on approved terms, and Cloudflare gave away the browser it built for bots.
She Made Her Anti-AI Song Harder to Fix Atwood Magazine
What happened: Vancouver musician Marie Dresselhuis released “Leave the Music” on Friday, November 13. She wrote it after hearing that AI songs had reached the Billboard charts, then recorded it live with friends on drums and guitar instead of as the sad piano ballad some fans expected. She told Atwood Magazine’s Sophie Severs that recording live made it “really hard to tune your vocals,” and that this was “a good moment to give myself no excuses; to just let it be what it was.”
Why it matters: Most arguments about AI music happen in comment sections. This one changed how a record got made. The song came out of a challenge Dresselhuis set herself last November: write one song a day for a month and post it, finished or not. Her objection to AI isn’t only that a machine makes the music. It’s that the struggle is part of what music is.
What everyone’s saying: Dresselhuis is blunt: “I don’t trust and I’m not convinced by any work that’s created by Suno AI.” Her case is that a platform has never sung in a choir or had its voice shake at a first open mic. “It’s an innately rebellious thing to create,” she says. It’s one artist interview, not a movement, but it says out loud what a lot of working musicians feel.
My read between the lines: The interesting move isn’t “no machines.” She still used recording gear, and the internet to release it. She picked which hard part to keep. That’s the question every tool now puts to every one of us: which effort is the point, and which is just the chore? She decided the wobbly vocal was the point. Story 2 is the opposite answer, and it got a red carpet.
📖 Further reading: Every New Tool Was Going to Kill Art, with John Hardin — my conversation with historian and musician John Hardin about the tape machine, the 808 and sampling. Every one of them was supposed to kill real music, and every one ended up giving musicians like Dresselhuis new choices about what to keep.
Dresselhuis kept the hard part on purpose. You get to choose which hard parts you keep, too. Hand the rest to Viktor, an AI agent that lives in Slack and connects to more than 3,000 tools. It builds the Monday report, refreshes the dashboard, ships the code fix and drafts the campaign. It’s not a chatbot you babysit. It’s a coworker you assign work to. New readers get $50 off their first month. Hire Viktor →
AI Music Got Its Own Red Carpet The AIMAs
What happened: The Sonic Intelligence Academy, or SIQA, held the first AI Music Awards Friday night at The Mayan, a 1927 theater in downtown Los Angeles. Reuters reports the top three prizes, AI artist, song and album of the year, all went to Xania Monet, an AI-generated avatar known for “How Was I Supposed to Know?” Her creator, Telisha “Nikki” Jones, says she wrote all the lyrics and set them to music with AI’s help.
Why it matters: Yesterday we asked whether musicians will start writing for the AI detector. SIQA skips the detector entirely. Its awards page says verification works by “disclosure, not detection,” and EDM.com reports that every registered track gets sorted as fully AI-generated, AI-assisted, or human plus AI. Winners are picked by verified artists in SIQA’s registry, who can’t vote for their own work. The one exception is a spotlight award chosen by SIQA’s own team.
What everyone’s saying: At the ceremony, producer Drake Murphy put the room’s case plainly: “In a world where everybody calls AI fake, the human behind it is real.” The music press is not rolling out a welcome mat. EDM.com said the show honors “the grueling craft of describing a song to a website.” Zinstrel, which covers AI music, counters that the night is historic either way, and says the thing to watch is whether the artists’ vote matches the charts.
My read between the lines: Look at what the AI music scene borrowed for legitimacy: a red carpet, a historic venue, genre categories and peer voting. All of it is very human machinery. And “disclosure, not detection” is a polite way of saying the honor system. The hardest-to-judge categories are the ones that ask where the human stops. Those answers come from the entrants themselves.
📖 Further reading: The real AI slop isn't Suno — it's Spotify -- AI Brief June 8 — back in June we argued the real AI slop problem is undisclosed AI, not labeled AI. SIQA built an entire awards show on that same bet.
No red carpet here, just a confession: the free Brief is the headline, and it will always be free. Members get the paywalled deep-dives where I show you how to actually use the tools in these stories, plus the full archive behind them. Become a member →
Gen Z Uses AI Daily. Some Feel Guilty. Semafor via Yahoo
What happened: A new survey from the consulting firm Milltown Partners, shared first with Semafor, found 46% of Gen Z adults use AI every day, second only to millennials. It also found 30% of Gen Zers sometimes feel guilty using it, more than double the rate among people over 30. Milltown surveyed 6,429 adults in the US and UK, with an extra 3,092 Gen Z respondents and four focus groups.
Why it matters: Gen Z gets cast as the generation turning against AI. The numbers are messier. Half of Gen Z expects AI to harm society, but so do 50% of baby boomers and 48% of Gen Xers. Millennials are the cheeriest, at 39%. And 73% of Gen Z say they haven’t made up their minds.
What everyone’s saying: Semafor’s J.D. Capelouto calls it a paradox: many young people say AI makes them worse at thinking for themselves and makes jobs harder to find, yet they use it daily. They’re also less worried than older people about the end-of-the-world risk that dominates AI safety debates. Their worries are closer to home.
My read between the lines: “Adoption” is a thin word for this. It counts the behavior and skips the feeling. People use the tool, worry about it and feel a little guilty, all at once. That’s not a contradiction. It’s how most of us feel about our phones. One caution: the published summary doesn’t show whether the daily users and the guilty ones are the same people, so don’t picture one guilty kid. Picture a generation that hasn’t decided yet.
📖 Further reading: OpenAI Made Its New Agent Adorable. That's the Part That Worries Me. — my look at OpenAI’s adorable always-on agent, and why the cuter and more helpful a tool gets, the harder it is to leave. That’s the habit forming underneath the guilt.
Universal Wants You Remixing. On Its Terms. HITS Daily Double
What happened: In a HITS Daily Double interview, Universal Music Group’s chief digital officer Michael Nash said UMG sent “around 75,000 takedown notices for deepfakes that had nearly 500 million plays” in just the last few weeks. He also laid out the other half of the plan: licensed AI deals with Hook, Klay, Udio, Spotify and ElevenLabs that give fans approved places to remix the work of artists who opt in.
Why it matters: Fans want to play with the music they love, and UMG says its research across 13 markets backs that up. “But interest isn’t permission,” Nash said. His examples of artist-approved uses are translating a song to understand it, making an acoustic version or reimagining a video. His rule for any artist’s name, voice or face: “consent, compensation and control.” The 75,000 takedowns are the stick. This is the carrot.
What everyone’s saying: Nash wants Congress to pass a federal right of publicity, meaning a national law that would protect anyone’s name, voice and likeness, as a top priority. “Imagine a deepfake of you going viral, putting words in your mouth that violate your values,” he said. Keep in mind this is the rights-holder’s side of the story. The interview doesn’t say how many artists have opted in.
My read between the lines: Fan remixing used to be free, unofficial and a little illegal. Now it gets a front door, a fee and a guest list. That could be good for artists who never saw a dime from fan edits. But “approved” cuts both ways. The record label now decides where the fence goes around your creativity, too. Watch the fine print on who owns the remix.
📖 Further reading: I Make AI Versions of Myself for a Living. This One I Didn't Agree To. — my piece on Meta’s Muse Image and the twenty years we spent handing over our faces for free. Nash’s “consent, compensation and control” is the deal that was missing then.
Cloudflare Gave Away Its Browser for Bots Cloudflare Blog
What happened: Cloudflare open-sourced Kitesurf, the web browser it built for AI agents, under the Apache 2.0 license. Developers can read the code, run it on their own machines, change it and deploy it to their own Cloudflare account. Cloudflare first showed Kitesurf in August and promised from the start to open it up.
Why it matters: AI agents that shop, research or fill out forms need a browser, and the usual choice is Chromium, the heavy engine under Google Chrome. Kitesurf drops what only humans need, like tabs, themes, extensions and pixel-perfect rendering. Cloudflare says it uses less CPU and memory than Chromium for common agent jobs, at the cost of slightly slower load times. It still connects to standard automation tools like Puppeteer.
What everyone’s saying: Cloudflare is upfront about the limits. The changelog calls Kitesurf “experimental and under heavy development,” says it has no built-in login protection, and asks would-be contributors to get vouched for by a maintainer before sending code. Credit where it’s due: the small open-source browser Obscura says Cloudflare’s first prototype started as a port of its engine.
My read between the lines: Cloudflare sells the walls that keep bots out, and now it’s giving away a bot browser. The docs list what Kitesurf can’t do yet, and one item stands out: it can’t pass a bot-challenge check the way a real browser does. So the gatekeeper’s free browser is very polite at the gate. That’s probably the business model.
📖 Further reading: AI Agents Can't Find Your Business. I Tested Mine. — I ran Greg Isenberg’s agent test on my own business, and found that an agent is a customer with a credit card and no eyes. Kitesurf is what those customers will browse with, so it’s worth checking what they see.
That’s your AI Brief for Saturday.
—Nicholas from Artificially Intimidating
