Tech should help us be creative. AI is taking the wheel
Making something worth listening to used to take time, taste, and a stomach for failure. Now, anyone can type a sentence into a model and get a song that sounds "good enough" for a scroll.
Three AI-made tracks recently hit Spotify's Viral 50. That's not "democratization" - that's volume. If a machine does the heavy lifting, can you claim the craft? You can press upload, but did you actually say anything?
The real threat isn't creativity - it's scale
At the core of the mess is the big S-word: scale. Platforms optimize for more files, more minutes, more clicks. Quality is secondary to quantity because quantity feeds the machine.
Streaming services favor time spent, not work finished or joy felt. Thousands of AI songs go up daily from non-musicians, burying the stuff with a pulse. Spotify hosts well over 100 million tracks - do we need more noise layered on top of that buffet? Here's their own count.
The result: a goopy stew of same-sounding content where "good" and "bad" flatten into "present." The few bright moments surface by accident, not by care.
Why this hits creatives hardest
When a model drafts your song, script, or thank-you note, you hand your voice to a mediator. The machine speaks in your place, trained on averages, optimized for safe probabilities.
Old tools extended your intent - a pencil, a DAW, a camera. AI can replace intent if you let it. A song about grit hits different when it's lived. Generated perseverance is cosplay.
And the metrics don't care if you enjoyed the two minutes of an AI track named for a bodily function. They only record that you stayed for two minutes.
A practical playbook to keep your voice in an AI-saturated feed
- No-ghostwriting rule: use AI for research, variations, and cleanup - never for the core message or melody. Draft the heart by hand, then polish with tools.
- Set constraints: one take before plugins, 200-word limit per idea, three references max. Constraints protect taste from bloat.
- Fewer, better: pick a sustainable cadence (weekly > daily). Quality compounds; spam dilutes trust.
- Own your channel: build an email list or community you control. Don't let an algorithm gate your audience.
- Curate like a pro: keep a personal canon of influences. Share playlists, books, and films that sharpen your taste.
- Track real signals: saves, replies, completions, repeat listens. Views and watch time are vanity if no one returns.
- Credit and consent: if you use models, disclose. Don't clone voices without permission. Keep raw project files for provenance.
- Human collabs > model blends: work with photographers, musicians, editors. Real chemistry beats synthetic polish.
- Skill reps: daily writing, scales, sketching. Craft still beats tooling in the long run.
- Timebox the feed: set strict windows for scrolling and "idea fishing." Protect deep work with calendar blocks.
What to do as the flood grows
Ignore most of it. Be a filter, not a funnel. Share the best work you find and explain why it's good - that builds taste, authority, and community.
Small is premium now. Intention is scarce. People will pay for work with edges - the kind that couldn't exist without you.
If you still want AI in your stack, use it without losing your voice
Treat AI like a studio assistant, not a stand-in. Learn prompts, auditing, and ethics so the tool bends to your taste - not the other way around. A curated starting point helps:
- AI courses by job - find practical workflows that keep you in the driver's seat.
The internet will keep rewarding scale. Your edge is refusing to become average. Keep the human in the loop - on purpose.
Your membership also unlocks:
