AI Music Revolution

Signal and Amplification and What That Means for AI Assisted Music Creators

• Josh • Episode 26

Share episode

Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.

0:00 | 23:15

Send us Fan Mail

The Washington Post ran a piece this month on what it called the AI content economy. Joel Waldfogel at the University of Minnesota has been documenting the AI book problem on Amazon for the better part of a year. Deezer reports 75,000 fully AI-generated tracks uploaded per day, which is 40 percent of all daily uploads. ArXiv tightened its submission policy in January 2026 because the moderation cost of fully AI-generated submissions exceeded the value of the contributions.

This is the noise economy. The total amount of content reaching every distribution channel exceeds the total attention available by orders of magnitude, and AI generation has dropped the marginal cost of production to nearly zero. The cap on production no longer exists. The cap on attention does.

Two responses are publicly visible right now. Both are dead ends. Position A is the AI content farm: manufacture volume, be the noise. This is failing to platform suppression (Spotify deplatformed 75 million spam-pattern tracks in 12 months) and competition from deeper-pocketed AI farms chasing the same suppressed shelf space. Position B is AI rejection: refuse the technology entirely. Moral clarity, structural disadvantages. Rejecting AI does not exempt the creator from the noise economy. It just removes the tools for navigating it.

The third position separates two things Positions A and B have collapsed. Signal is what gets created. Human-originated, taste-driven, quality-bound. AI cannot produce signal because AI does not have taste, judgment, or stakes. Amplification is what happens after the signal exists. The distribution, the production scaling, the social variants, the metadata, the workflow automation. Mechanical, repetitive, and well-suited to AI execution. AI in the signal layer produces slop. AI in the amplification layer produces leverage. The location of AI in the workflow is the entire question.

This is Lane 2 publishing, the direct parallel to Lane 2 production. The human directs the work. AI handles the scaling. The signal is human. The amplification is machine. The ratio is intact.

The second half of the episode lands on a specific example of disciplined amplification in practice: the Spotify paid stack. A friend texts you at 11 PM with $200 to spend and three Spotify paid options sitting side by side on the dashboard. Marquee, Showcase, Discovery Mode. Same minimum spend on two of them. Same vague promise of amplification. Different jobs entirely.

Marquee is a full-screen takeover during release week, with the highest intent rate of the three tools (29.6 percent vs Showcase's 9.6 percent per Brian Hazard's published campaign data). Showcase is a Home feed banner with lower intent and lower cost per click, useful for older catalog tracks or when the goal is conversion to owned channels. Discovery Mode is not an ad at all but a royalty trade: algorithmic amplification in exchange for a 30 percent reduction in royalty rate on the affected streams. The episode walks through what each tool actually does, when each one is the right instrument, and the four-question pre-spend diagnostic that determines whether paid Spotify is even available to the catalog before a single dollar gets spent.

The unifying argument across both halves of the episode is the through-line of the whole show. The discipline doesn't end with "use AI in amplification." The discipline runs all the way down. Every choice in the workflow is the same choice. Where am I putting the human work? Where am I putting the AI work? Where am I putting the money? The Lane 2 creator wins the noise economy by getting that question right consistently. Not just once. Every release.

This episode is the spoken companion to two pieces from the JG BeatsLab blog: the May 25 Monday Manifesto "I Use AI Heavily. I Don't Use AI to Replace the Signal." and the May 29 Friday Informative Bits "The Spotify Paid Stack and Why Most Artists Spend on the Wrong One."

If you want the complete amplification methodology, Unlock Music Promotion is the book. The full Spotify Paid Stack breakdown, the campaign math, the third-party playlist comparison, and the Plan, Run, Review framework for measuring algorithmic uplift all live in Chapter 5. Available for $9.99 on the JG BeatsLab site and on Amazon. The Premium Starter Kit at jgbeatslab.com/store for $67 includes Unlock Music Promotion alongside Unlock Suno and Unlock Music Rights and Registration, plus the Blueprints, reference cards, 3-Song Sprint course, and Fader (the custom AI Studio Manager). Everything Josh publishes lives inside Red Lab Access for $117 lifetime at jgbeatslab.com.

Subscribe so you don't miss next Friday's episode. Find more at jgbeatslab.com.


The Unlock System is JG BeatsLab's methodology for serious musicians working with AI tools. Lane 2 work: human-authored, AI-assisted music creation.

Visit JG BeatsLab: https://www.jgbeatslab.com

Every Thursday morning, I send out a newsletter. One email a week. It's where I share what I'm working on in the studio, the research running in the lab, and the methodology I'm refining in real time. New books, new Blueprints, new findings from the Red Lab Protocol research. If that sounds useful, you can sign up at jgbeatslab.com/newsletter. Thanks for listening, and I'll see you next Friday.

AI Music Revolution is produced by JG BeatsLab LLC, an AI music education company building the methodology, research, and community for serious creators working in Lane 2.

Get more from JG BeatsLab LLC:

Connect:

Contact: josh@jgbeatslab.com

Stop gambling. Start directing.

SPEAKER_00

Hello and welcome to the AI Music Revolution. I am your host, Josh Gedoland, the founder of JGB Slab. Today's episode is about a question every working creator using AI has to answer in 2026, even if they don't realize that they're answering it. Where does AI belong in your workflow? Not whether to use it, but where to put it. Because the location of AI in the workflow is the entire question. And most creators are getting that question wrong in one of two ways. We're going to talk about the noise economy, what it is, what's actually happening inside it, and the two failed responses that almost every working creator has settled into. Then I'm going to walk you through the third position, which almost nobody is articulating, even though it's the only one that actually works. And then because this show is about getting useful work done in the real world, we're going to land on a specific example of what disciplined amplification looks like in practice. The Spotify page stack, marquee, showcase, discovery mode. Three tools sitting side by side on the same dashboard. Most artists pick the wrong one and burn the budget. This is the AI music revolution. Let's get into it. The Washington Post ran a piece this month on what it called the AI content economy. The reporting was unsentimental, which is what made it useful. The volume of AI-generated content flooding every distribution channel has reached a point where the publishing ecosystem can no longer absorb it. Books on Amazon, articles on the open web, tracks on streaming platforms, academic papers on preprint servers. The supply curve broke. This is not the first piece to describe this. It's one of the first that documents it. Joel Waldfogel at the University of Minnesota has been studying the AI book problem on Amazon for the better part of a year. The bottom of the Amazon book market has been overrun with AI-generated titles, indistinguishable in their metadata from human-authored books, sold at the same prices, competing for the same algorithmic self-space. The reader cannot tell which is which after the purchase, and often not even then. Deezer publishes its own daily intake numbers, 75,000 fully AI-generated tracks uploaded per day. 40% of all uploads. That is not a market disruption, that is a market saturation event. Archiv tightened its submission policy in January 2026. The preprint server that has powered scientific dissemination 30 years now refuses certain categories of fully AI-generated submissions because the moderation cost exceeded the value of the contributions. The infrastructure of science itself bent under the weight. This is the noise economy. The total amount of content reaching every channel exceeds the total attention available by orders of magnitude. And AI generation has dropped the marginal cost of content production to nearly zero. The cap on production no longer exists. The cap on attention does. This is the landscape every working creator is operating inside of right now. There's no opting out of it. There's only how you respond. The two responses are publicly visible right now, and both are dead ends for structural reasons. Position A is the AI content farm, the creator who uses AI to manufacture volume. 40 books a year, all AI generated, 100 tracks a month, all generated and uploaded, a thousand articles produced and indexed. The bet is that volume wins inside the noise economy. The strategy is to be the noise. This is failing in real time. Spotify deplatformed 75 million spam pattern tracks in 12 months. Amazon has tightened metadata enforcement against undisclosed AI uploads. These are excludes flagged AI tracks from algorithmic playlists. The platforms hunt for these patterns specifically because the platforms are also drowning in noise. And the only economic move available to them is suppression of obviously machine-generated catalogs. Position A is also being competed against by AI content farms with deeper pockets and faster generation pipelines, all chasing the same suppressed shelf space. The race ends in zero royalties and account flags. Position B is the AI rejection, the creator who refuses the technology entirely. Real artists do not use AI. The position has moral clarity. It also has structural disadvantages that compound over time. Position B fails because rejecting AI does not exempt the creator from the noise economy. The same distributor channels are still saturated. The same algorithms still serve listeners. The same reader still cannot tell at the metadata layer which book on Amazon is human and which one is machine. Position B does not reduce the noise, it just reduces the rejector's tools for navigating it. The creator who refuses to use AI for distribution amplification is competing in a noise economy with one hand tied behind their back. The moral position is intact. The career math is brutal. So position A loses to platform suppression. Position B loses to structural disadvantages. The third position exists, and almost nobody is articulating it. The third position starts from a distinction position A and B have collapsed. Signal and amplification are different problems. They use different tools. And the failure of both positions is in treating them like the same problem. Signal is what gets created. It is the song. It's the book, the article, the catalog, the body of work. It's human-originated. It carries the artist's taste, judgment, choices in perspective. It is the thing the audience eventually decides to pay attention to or not when they encounter it. Signal is not a quantity problem. It's a quality problem. The signal is good when the human who created it has done the work to make it good. AI cannot produce signal because AI does not have taste, judgment, or stakes. AI can produce material that looks like signal. The audience knows the difference eventually, and the platforms are getting faster at knowing the difference up front. Amplification is what happens to the signal after it is created. It's the distribution, the production scaling, the variant testing, the platform navigation, the metadata optimization, the social posting, the email writing, the workflow automation. Amplification is a quantity problem disguised as a quality problem. Most of the work of getting signal in front of an audience is mechanical, repetitive, and well suited for AI execution. The third position uses human capacity for signal and AI capacity for amplification. The creator writes the song, the book, the article. The creator directs the production, makes the decisions, owns the outcome. AI handles the part of distribution that are mechanical, the social variants, the email drafts for testing, the metadata cleanup, the workflow automation, the volume of small executional tasks that used to consume the creator's time and produce nothing of artistic value. The creator using AI this way is not feeding the noise economy. They are using AI to cut through it. The signal is human, the amplification is machine, the ratio is intact. This is lane two publishing, the direct parallel to lane two production. In lane two music production, the human directs the AI generation, makes the creative decisions, owns the masters. The AI does the part the human cannot do alone at speed. The output is human-authored, AI-assisted. The work is the human's. In lane two publishing, the same logic holds to everything that surrounds the work. The blog post is written by the human, the AI handling the formatting, the social variants, the metadata. The book is authored by the human, with AI handling production tasks that used to require a publishing team. The catalog strategy is the humans, with AI handling the executional volume that distribution at scale now requires. Lane 2 Creator is not less human because they use AI. They are more present in the parts that actually matter, because AI is handling the parts that did not need their presence in the first place. I use AI heavily is not in conflict with this position. It is part of the position. The amount of AI in the workflow is not the question. The location of the AI and the workflow is the entire question. AI in the signal layer produces slot. AI in the amplification layer produces leverage. The creator who refuses to use AI on the amplification side is performing virtue at the cost of reach. The creator who is using AI on the signal side is producing noise at the cost of meaning. The third position is what is left for working creators who want to do real work without being either drowned or marginalized. I use AI to amplify the signal I create, not to create the signal itself. That is the position. That is what lane two publishing means. Now here's where the episode gets practical, because saying AI belongs in amplification is the philosophical answer. The operational answer is harder because even when you put AI in the right place, you still can do amplification badly. You can still spend money on the wrong tools. You can still throw budget at a problem the budget can't solve. I want to walk through one specific example of what disciplined amplification looks like because I see creators get this wrong constantly. The Spotify page stack. Three tools that look interchangeable on the dashboard. They are not. Marquee is a full-screen takeover. A targeted listener opens Spotify and your track appears at a full screen promotion before they reach any other feature. It is the most intrusive format Spotify offers, which is also why it has the highest intent rate. Brian Hazard's published campaign data shows 29.6% intent rate for marquee compared to 9.6% for Showcase in equivalent context. Three times higher engagement. Minimum spend is $100. Eligibility window is 18 days after release. Shocase is a home feed banner. The listener opens Spotify and your track appears in their home feed alongside the personalized playlist and recommended artists they normally see. Same $100 minimum, lower cost per click, lower intent rate, fixed 14-day duration. The eligibility window is open to any track at any age, which is the one thing Showcase has that Marquee does not. Discovery Mode is not an ad at all. It's a royalty trade. You opt a track in and Spotify increases its algorithmic distribution in radio, autoplay, and mixed playlist context. In exchange for a 30% reduction in royalty rate on the streams that come from that amplification. If a stream would normally pay 4 tenths of a cent, it pays a little under 3-tenths of a cent instead. The math only works if the volume uplift exceeds 30%. Practitioner data suggests genuine uplift falls between 15 and 40%, depending on the track, genre, and baseline engagement. If your uplift is 15%, you lose money. If it's 40%, you gain. Three tools, same dashboard, completely different jobs. Marquee is the right tool when the track is brand new. The baseline save rate is already at least 4%. The catalog clears Spotify's monthly active listener threshold in at least one target market, and the goal is to seed algorithmic signals during the pulse release window. Hazard's data shows that a 100 to 150 marquee campaign in week one produces measurably better algorithmic outcomes in weeks three through five than the same dollars split across four smaller campaigns over a month. Concentrate the spend, run at once, front load it into days one through seven. The mechanism is not the impression itself. It is the save and completion rate signals the impression generates. When 200 listeners click your marquee and 20 of them save the track, Spotify's algorithm logs that 10% save rate and escalates your visibility through release radar, radio, and discover weekly. 60% of marquee campaigns see measurable algorithmic uplifts in the two to four weeks following the campaign. The campaign is a catalyst. The algorithm does the actual work. Showcase is the right tool when the track is older than 18 months and marquee will not accept it. Or when the goal is conversion to owned channels rather than streams. Use Showcase as a secondary lever, not a primary one. Never use it for a new release when Marquee is eligible. Now, Discovery Mode is the right tool when the track is past release week. The save rate is already above 6%, and the goal is to extend the life cycle of an existing winner. Discovery Mode does not save weak tracks, it amplifies tracks that already have algorithmic momentum and it costs royalty share to do it. Pick one or two tracks per year that are genuinely working and give them the boost. Do not enable it across the catalog. The pre-spin diagnostic before any Spotify paid spend. Four questions. Does the track have a baseline save rate above 4%? If not, no campaign will fix it. Does the catalog clear Spotify's eligibility threshold? Marquee requires at least 1,000 streams in your target market over the last 28 days, plus at least 5,000 monthly active listeners in that target market. If the catalog does not clear those thresholds, the tool's not available regardless of budget. Is there a three-day buffer before launch? Marquee needs the track live for at least three days. Is the budget at least $100? If less, paid Spotify tools are not eligible. The dashboard makes the three tools look equivalent. The math says they are not. Match the tool to the job, run it in the right window, measure the algorithmic payoff, and the spin works. Default to the wrong tool and the $100 minimum becomes a $100 spent on signal that the algorithm never amplifies. I want to land on the connection between these two arguments because the connection is the whole point of the episode. The manifesto argument was philosophical. AI in the signal layer produces slop. AI in the amplification layer produces leverage. The location matters more than the volume. The Spotified Page Stack argument is tactical. Within amplification, the specific tool you pick matters too. Marquee, showcase, and discovery mode all live in the amplification layer. They are not interchangeable. The discipline doesn't end with use AI in amplification. The discipline runs all the way down. These are the same arguments at two scales. The third position is not a one-time decision. It is an operating posture. Every choice in the workflow is the same choice. Where am I putting the human work? Where am I putting the AI work? Where am I putting the money? Is each piece of leverage matched to the job it actually does? Or am I defaulting because the dashboard makes it look easy to default? Lane 2 creator wins the noise economy by getting that question right consistently. Not just once. Every release, every campaign, every workflow decision. If you want the full breakdown of how discipline amplification actually works across the platforms that matter, Unlock Music Promotion is the book. The Spotified Page Stack lives in chapter 5, including the campaign math, the third-party playlist comparison, and the plan run review framework for measuring algorithmic lift. 999 at jgbslab.com or on Amazon. If you want the full methodology infrastructure that turns the lane two position into a working catalog, the premium starter kit is the move. The three books Unlock Suno, Unlock Music Rights and Registration, Unlock Music Promotion, plus the blueprints, plus the reference cards, plus the three-song sprint course, plus Fader, the Custom AI Studio Manager. $67 at jgbeatlab.com backslash store. The full design of the workflow package and one entry point. And if you want everything I publish, every book, every Red Lab protocol research report, every blueprint, the sprint course, fader, the community of people running this methodology with intention. Red Lab Access is the move. One price, lifetime access, $117 at jgbeatslab.com. The noise economy is real. The two failed positions are visible. The third one is the design. Signal is yours. Amplification is delegated. The location of AI in the workflow is the entire question. Get that question right, run the methodology, and the work cuts through. Stop gambling. Start directing. Thanks for listening. New episode every Friday. Subscribe so you don't miss next week. I'll see you then.