Ask HN: Why hasn't x86 caught up with Apple M series?
418 points by stephenheron 2d ago 599 comments
iOS 26 Launches Sept 15 – Even GPT-5 Doesn't Know It Exists
2 points by rileygersh 6h ago 5 comments
Ask HN: Best codebases to study to learn software design?
100 points by pixelworm 3d ago 89 comments
Stop squashing your commits. You're squashing your AI too
4 points by jannesblobel 1d ago 9 comments
Show HN: A math tutor that won't hallucinate answers (ChatGPT and Photomath)
1 noygaryan 0 8/27/2025, 2:46:23 PM thinkercan.com ↗
Hi HN,
I'm the creator of thinkercan.
The best way to describe it is ChatGPT for reasoning + Photomath for the step-by-step solution.
I built this because I wanted a tool that could understand a student's question in natural language (even word problems) like ChatGPT, but then generate a perfect, step-by-step mathematical solution like Photomath.
The core problem with using LLMs alone for math is that you can't trust their process. They hallucinate steps and get the final answer wrong. My goal was to fix that.
Under the hood, thinkercan uses an LLM as an intelligent interface to understand the query. But the crucial part is that it then calls a dedicated symbolic math engine to generate the entire solution path—every single step, from start to finish.
This means you get a solution that is both easy to understand and mathematically sound from top to bottom.
Here’s a 50-second video that shows the process: https://www.youtube.com/watch?v=nI05v-Yltyc
We're still in the early stages (just added calculus support), and I'm here to answer any questions. I would be grateful for any feedback on the approach and the product.
No comments yet