HN Reader

Is it possible to write plain C iOS app in 2025?

102 points by iMario 3d ago 50 comments

Ask HN: Why is there no P2P streaming protocol like BitTorrent?

229 points by memet_rush 3d ago 220 comments

Ask HN: What are the best books on talking to users?

5 points by terrellbm 15h ago 6 comments

Ask HN: Have you been contracted for your repo?

6 points by gamescr 16h ago 1 comments

Ask HN: Why do people buy Nvidia RTX 5090 at x2 the MSRP?

10 points by behnamoh 1d ago 7 comments

Ask HN: Software Engineers to follow who have a healthy skepticism of AI

8 points by wronginternet4 18h ago 12 comments

Ask HN: The Future of SaaS

3 points by pramatosh5125 23h ago 1 comments

PolyThink: A Multi-Agent AI System to Eliminate Hallucinations

8 points by kuberwastaken 1d ago 9 comments

Ask HN: I feel pessimistic about the future – advice?

34 points by hnuu3 21h ago 26 comments

Ask HN: Git Rebase and Merge or Squash and Merge?

5 points by floriangosse 1d ago 3 comments

Ask HN: How do you monetize personal code if it's not an "app"?

171 points by splimeproject 5d ago 68 comments

Ask HN: How do people around you use AI?

3 points by nicbou 1d ago 10 comments

Ask HN: What is the future of tech roles?

6 points by farzd 1d ago 8 comments

A Scalable Standard for Clean ECommerce Data in LLMs (Fork of Llms.txt)

2 points by nicola_alessi 1d ago 0 comments

RocketSDR – Are they faking being backed with YC?

3 points by blendertom 1d ago 2 comments

Ask HN: What is the best font and styling combo for readability on the web?

16 points by duxup 2d ago 10 comments

Understanding Some Limits of DeepSeek

9 points by hazymemory 2d ago 0 comments

Ask HN: Why Don't More Product Teams Collaborate on Domain Models?

3 points by adangit 1d ago 4 comments

Tell HN: A realization I've had about working with AIs and building software

19 points by keepamovin 3d ago 15 comments

Ask HN: What's the Most Effective Way You've Improved Focus?

21 points by schappim 1d ago 22 comments

Restrictions on Our Freedom to Study Software: Train Manufacturer vs. "Hackers"

4 points by nathan_phoenix 1d ago 1 comments

Ask HN: Magic links are bad UX and make people's lives worse. Change my mind

51 points by figassis 4d ago 39 comments

Ask HN: Has anyone quit their startup (VC-backed) over cofounder disagreements?

77 points by stuck12345 6d ago 74 comments

Ask HN: How do you raise your kids in the age of AI?

4 points by MichaelMoser123 1d ago 23 comments

Ask HN: Why don't we have a functional DSL for data+embedding+API pipelines?

21 points by codingmoh 4d ago 21 comments

Ask HN: Dora Metrics: Are they here to stay?

3 points by riyanapatel 21h ago 1 comments

Ask HN: Slopsquat CVE?

3 points by ChuckMcM 3d ago 1 comments

Ask HN: What are the best books with problem sets that test your understanding?

12 points by jimsojim 4d ago 7 comments

Ask HN: How can sending emails cost Let's Encrypt five figures?

6 points by rrr_oh_man 2d ago 8 comments

PolyThink: A Multi-Agent AI System to Eliminate Hallucinations

8 kuberwastaken 9 4/17/2025, 8:55:04 AM
Excited to announce PolyThink Alpha's early access! Our multi-agent AI system fights hallucinations with consensus-driven, accurate answers from multiple models. I'd love for you to join the waitlist at https://www.polyth.ink/ as I'm planning to randomly roll out invites starting May. Feedback will shape our final launch! I'd love thoughts and suggestions too! What would you like to see here?

Comments (9)

stereo · 1d ago
Isn’t this basically the Swiss cheese model? If your two input AIs hallucinate, or your consensus AI misunderstands the input, you will still have confabulations in the output?
kuberwastaken · 1h ago
From all my testing, this never really happened even once honestly, plus the judge model (that I've kept strictly a reasoning model) also evaluates individually before "judging" the consensus.
TheKelsbee · 20h ago
I have this same thought, and have tried similar approaches.

OP: Have you trained or fine tuned a model that specifically reasons the worker model inputs against the user input? Or is this basically just taking a model and turning the temperature down to near 0?

kuberwastaken · 1h ago
Low temperature, heavy prompting to answer in a structured way. Sadly can't fine train models since this is API based but the approach does work!
sks38317 · 1d ago
I’m genuinely interested in how you arrived at the concept of using AI as a method to treat hallucinations. What inspired that approach?
kuberwastaken · 1h ago
Honestly, personal use cases. I am a STEM student and deal with a lot of "hard" questions that are about 60% of the time miscalculated by LLMs, I used to manually paste in approaches from say ChatGPT to DeepSeek and now grok and asked them what do you think is better. I created this out of necessity to automate this then realized how cool it can be if it scales further haha
tough · 19h ago
not op but LLM as Judge is a thing https://arxiv.org/abs/2411.15594
consumer451 · 1d ago
Very interesting. Will this be available as a meta model via API, allowing use in the coding tool of my choice?
kuberwastaken · 1h ago
Eventually yes, that's the plan! It's extremely good with code too, especially with more vague requests, tends to take about 2-3 rounds but almost always gets a great approach.