Hii, just made this project — would love your honest advice

Umm, I don't think this project is going to be very eye-catching or generate a lot of traction, and not because it isn't useful, but because nowadays there are hundreds of LLMs out there, each one with a better marketing team than the last... What I mean is that the masses are more focused on hype and FOMO than on actually figuring out which AI is best for them... Because at the end of the day, they have plenty to choose from at will and can test every button on that huge dashboard like a kid in a candy store... I think if there were only 8 models or fewer out there, then your project would be a boom... But under these conditions, no...

Building on that, I believe your perspective is extremely well-grounded and hits on a critical flaw in this kind of venture. The core issue goes beyond just market noise—it’s about internal redundancy. Most modern LLMs are already fully capable of doing an honest self-evaluation or comparing their architectures and capabilities against competitors. If a user wants to know where a specific AI falls short, they can literally just ask the model itself without needing an external third-party agent.

Furthermore, forcing users to visit an independent platform just to see strengths and weaknesses adds unnecessary friction in an ecosystem already suffering from severe tool and dashboard fatigue. Unless this platform evolves past a simple informational site into something deeply technical—like running live, automated empirical benchmarks, real-time cost-per-API tracking, or acting as an intelligent prompt router—it risks being completely redundant for the general public. Of course, that's just my opinion.

The interesting question is whether people need a filter for AI launches often enough to build a habit around it.

What would convince you that Reality Filter is solving a recurring decision problem, rather than being something people find interesting once and move on from?