No reviews yetBe the first to leave a review for Ithy
Layout Agent 2.0 — Thinks before it builds
Thinks before it builds
Promoted
Multi-model consensus is genuinely interesting for research-focused tasks. One thing I keep wondering about products in this space: when models disagree, how does Ithy surface that divergence rather than just averaging it away?
Report
The multi-model aggregation angle is interesting, but when models disagree, which one wins? That tie-breaker logic is probably where Ithy lives or dies.
Report
Great job. I have used and recommended Ithy since it first started. I think Ithy deserves a big pat on the back.
@antonio_tena Thank you Antonio for the support! If you remember the first version of Ithy, it was very slow and had basically no features - thanks for sticking with us lol
Report
The hard part isn't combining outputs, it's knowing when to trust which model. Most ensemble approaches paper over that with a final summarization pass and call it done.
Report
Bro, this is insanely cool! I just tried it, and honestly, if you take this to market, it’s going to boom! Preplexity is nothing compared to this. I’m a developer, and the way it generated the article blew my mind.
Also, if you have any web or app-related work, I’d love to collaborate. I’m looking to connect and work with like-minded people
@eileen_zhu1 Thank you! Yes, even if some of the answer might be "too much", we hope Ithy brings at least 1 unique perspective that other AIs don't. We really try to make sure you learn something new every time you ask a question!
Report
I teach AI Tools, and I find the combination of four LLMs to be an excellent way to demonstrate how to eliminate bias. This format is fantastic, and I highly recommend it.
@waltraud_krickl Great point, bias elimination is definitely something that we're always trying to be the best at. Since every LLM has some sort of bias, using Ithy will at least let you consider multiple perspectives together.
Multi-model consensus is genuinely interesting for research-focused tasks. One thing I keep wondering about products in this space: when models disagree, how does Ithy surface that divergence rather than just averaging it away?
The multi-model aggregation angle is interesting, but when models disagree, which one wins? That tie-breaker logic is probably where Ithy lives or dies.
Great job. I have used and recommended Ithy since it first started. I think Ithy deserves a big pat on the back.
Ithy
@antonio_tena Thank you Antonio for the support! If you remember the first version of Ithy, it was very slow and had basically no features - thanks for sticking with us lol
The hard part isn't combining outputs, it's knowing when to trust which model. Most ensemble approaches paper over that with a final summarization pass and call it done.
Bro, this is insanely cool! I just tried it, and honestly, if you take this to market, it’s going to boom! Preplexity is nothing compared to this. I’m a developer, and the way it generated the article blew my mind.
Also, if you have any web or app-related work, I’d love to collaborate. I’m looking to connect and work with like-minded people
Portfolio https://sujal-sb.vercel.app/
Ithy
Thank you @sujal_bagavan! Perplexity is our biggest competitor ;)
And we love your portfolio - let's get in touch!
The speed and polish are seriously impressive. Loved how it doesn’t just answer, but turns every query into a media-rich deep dive.
Ithy
@eileen_zhu1 Thank you! Yes, even if some of the answer might be "too much", we hope Ithy brings at least 1 unique perspective that other AIs don't. We really try to make sure you learn something new every time you ask a question!
I teach AI Tools, and I find the combination of four LLMs to be an excellent way to demonstrate how to eliminate bias. This format is fantastic, and I highly recommend it.
Ithy
@waltraud_krickl Great point, bias elimination is definitely something that we're always trying to be the best at. Since every LLM has some sort of bias, using Ithy will at least let you consider multiple perspectives together.