I'm 17, built an AI hype-checker, and I genuinely don't know if it should be a product
https://bharatlearner18-del.github.io/reality-filter/
Built Reality Filter: paste an AI launch, get a verdict (worth trying / watch / too early / hype) backed by named sources. Some launches already have a verdict added — for the rest I research it myself. I'm 17, solo, still deciding project or product. Tell me straight.
The interesting question isn't whether AI hype-checking is useful — it's whether people need a verdict often enough to build a habit around it.
What would convince you that users need a recurring trust layer for AI launches rather than just using it occasionally when something feels suspicious?
I don't think people need a "trust layer" itself. I was thinking about a system where users pick the AI topics or companies they care about, and the system keeps tracking them. Whenever something new happens, it checks how much is real vs hype, what experts and users are saying, the risks, and then sends a simple summary by email.
It could help founders decide which tools are actually worth using, and they could share those updates with their team so everyone knows what's actually important in their field, what's just hype, and what deserves attention.
But honestly, I don't think many people would pay for that. Email updates feel a bit boring, and most people probably don't need this every day even founder's.
That's why I feel a simple search-first product might work better. No login, no setup—just open the website, search any AI company or topic, and instantly see the reality behind the hype. People may only come when they need it, but over time that could still become a habit. What do you think?
I think you've narrowed it down to a more important question than email vs. search. I'd be careful not to treat early usage as proof you've identified the underlying need—they can look very similar for quite a while.
I don't even have real users yet. Whenever I send it to someone, they try it out and give me a little bit of feedback—that's about it.
I appreciate the honesty. It's refreshing to see someone separate assumptions from evidence instead of pretending they're further along than they are.
I'd be interested in comparing a few thoughts outside the thread. What's the best email to reach you on?
Thanks - and honestly the separating is the only part I feel confident about right now.
Happy to compare notes. Easiest is X, I'm @bharatlearner there. Or if you'd rather keep it here, just say what you're chewing on and I'll answer properly in the thread - I'd slightly prefer that anyway, since anything useful we work out is then sitting where the next person can find it.
What's the angle you're thinking about? If it's the validation side, I've got a fairly unflattering set of numbers I'm happy to put on the table.
I appreciate that, and I actually agree that the best discussions are the ones other people can learn from.
The angle I'm thinking about is a bit too specific to do justice in a thread, and I'd rather not reduce it to a couple of comments. I don't really use X either.
If you're comfortable sharing it, what's the best email to reach you on instead?
what do you use "X" or "Instagram" ??
Neither, really.
The easiest way to reach me is hello@beryxa.com. Feel free to send me a message there, and we can continue the conversation.
This comment was deleted 2 months ago