How I'm Using AI Without Losing My Voice: A Real Creator's Take

Let's talk AI. Not the hype, not the fear — just what I'm actually using right now in July 2026 to run In Da Streets Radio, push my own music, and keep ninacapone.com moving without burning out. Because here's the thing: AI can be a tool in your hands or a crutch that makes your work sound like everybody else's. I'm choosing the first one.

This month I'm leaning hard on ChatGPT for lyric brainstorming — not writing my bars for me, but bouncing ideas. I'll drop a theme or a half-finished hook, ask it to give me five different angles or wordplay options, then I pick what resonates and rewrite it in my voice. It's like having a writing partner who never gets tired, but you still gotta be the editor. The free ChatGPT tier works fine for this, but I'm on the Plus plan at twenty bucks a month because the faster responses keep my flow state intact when I'm creating.

For podcast transcription and show notes, I'm using Descript. I record my interviews and commentary in Riverside or OBS, export the audio, drop it into Descript, and it spits out a transcript in minutes. Then I use that transcript to pull quotes for Instagram carousels in Canva, write blog posts, or generate YouTube descriptions. Descript's free tier gives you one hour of transcription per month; the Creator plan is twelve dollars monthly for ten hours. That's worth it when you're putting out multiple episodes a week and need to repurpose everything fast.

Image generation is where I'm still cautious. I've tested Midjourney and DALL-E for show artwork and social graphics, and yeah, they can create wild visuals. But I only use them for backgrounds or abstract elements — never for anything that's supposed to represent a real person or culture without my direct creative input on top. I don't want my brand looking like a thousand other AI-slop accounts. If I use an AI-generated background, I'm adding my own text, photos, and design choices in Canva so it still feels like me. Midjourney runs about ten dollars a month for the basic plan; DALL-E credits through OpenAI vary, but you can start small.

Here's my ethical line: AI assists, it doesn't replace. I will never let a bot write a full song, a full article, or a full episode script and slap my name on it. That's not creating, that's cosplaying as a creator. I use AI to speed up the grunt work — transcripts, first-draft social captions, brainstorming lists — so I have more energy for the parts that actually matter: the performance, the storytelling, the connection with my audience. If the tool makes me faster but I'm still the one making the final calls, that's a win.

The stuff that doesn't work? AI-generated voiceovers that sound robotic (ElevenLabs is getting better, but it's not there for long-form content yet), and any tool that tries to 'optimize' my content by making it sound like a corporate press release. I've also seen AI music generators like Suno and Udio — they're fun to mess with, but the output isn't ready for serious release. It's more like a sketch tool for now.

One hidden cost to watch: most AI platforms upsell you hard after the free trial. ChatGPT, Descript, Midjourney, Notion AI — they all have tiers that start small and balloon if you need team features or higher limits. Budget for twenty to fifty bucks a month total if you're serious, and track what you actually use versus what you're paying for. Cancel the subscriptions that aren't earning their keep.

Bottom line: AI is a power tool, not a personality transplant. Use it to clear the boring tasks so you can focus on what makes your work yours. If your audience can't tell the difference between you and a bot, you're doing it wrong. Stay human, stay intentional, and let the tech serve you — not the other way around.

Drop a comment or DM me at ninacapone.com with the AI tool or workflow you want me to break down next — I'm here for the real talk, always.