TL;DR: AI Insider is a content experiment where I'm building a newsletter using AI automation. Instead of hiding the AI behind polished marketing copy, I'm showing exactly how it works — real costs ($0.08/article), real prompts, real failures. If you want to see how AI-driven content actually gets made, you're in the right place.
Why I Started This
I run V7 Fund, a venture capital firm focused on AI startups. Every day I see founders building AI products, but most AI content online is... garbage. Generic SEO spam, clickbait, or AI-generated slop pretending to be human.
I wanted to try something different: What if I built a content machine that's completely transparent about being AI-powered?
How It Actually Works
I have an AI assistant named Anna (running on Claude) that helps me create content. But here's what makes this different:
- No hiding: Every article shows exactly how it was made
- Real costs: I track and publish what each piece costs to create
- Real prompts: You can see the actual instructions I give the AI
- Real failures: When things break, I write about that too
The Numbers So Far
Here's what I've learned after the first week:
- Cost per article: ~$0.08 (Claude API)
- Time to publish: 15-30 minutes with human review
- Articles published: 10+
These numbers will change as I optimize. I'll keep updating them.
What You'll Find Here
This isn't a news site. It's not trying to compete with TechCrunch or The Verge.
Instead, you'll find:
- Show Your Work posts: Behind-the-scenes of how each piece gets made
- AI industry analysis: My take on what's actually important in AI
- Tool reviews: Honest assessments of AI products I actually use
- Experiments: Things I'm testing, with real results
Why "Show Your Work"?
Most AI content tries to pass as human. That's a losing game — readers can tell, and it destroys trust.
I'm betting the opposite approach works better: Be completely transparent about AI involvement, and focus on delivering genuine value.
If I'm right, this becomes a model other creators can follow. If I'm wrong, at least you'll see exactly where I failed.
Want to follow along? Subscribe below. I'll send updates when I publish something worth reading.