News for hackers. Stuff that matters.

javantanna

Re:Bonsai: A Voxel Engine, from scratch on Tuesday December 16, 2025 @03:35AM (#46286159)
btw the license is nuts
Re:Ask HN: What is Safe Superintelligence Inc building? on Monday December 15, 2025 @06:16AM (#46272983)
Well, thanks
Re:What is the most effective way to learn programming? on Monday December 15, 2025 @03:47AM (#46271902)
Honestly, I struggled to learn programming through tutorials and step-by-step courses. What finally worked for me was building things. I learn the basics first—loops, conditionals, OOP—then look for unique project ideas with a real wow factor and build them end to end
Re:Countries with the Most Spoken Languages on Monday December 15, 2025 @02:19AM (#46271357)
ig india
thank you so much bro i am working on a project for around 9 months and stuck at this your post is exactly what i needed
Re:Ask HN: How do you give a local AI model long-term memory? on Saturday December 13, 2025 @03:22PM (#46257639)
thanks man i need to take a look to your code bcoz as you said hierarchical summaries i try to implement it didn't work for me like i am building a system which ocr pdf of legal contracts between parties so this way breaks when there is time to extract specific clauses as per contract
Re:Ask HN: How do you give a local AI model long-term memory? on Saturday December 13, 2025 @03:19PM (#46257600)
but in my company we work with legal documents so our data is soo confidential and we can't use apis i need to set it op offline!!
Re:My new killer SaaS (Script-as-a-Service) – safe-claude.com on Saturday December 13, 2025 @06:15AM (#46253754)
do i need claude api key ??
Re:Over 12,000 Startup Ideas Right Here on Saturday December 13, 2025 @05:57AM (#46253687)
I bought it and the documentation of startup ideas is really rich and properly explained its worth the money!!
Re:GitHub Is Down on Saturday December 13, 2025 @02:36AM (#46252827)
another service down in Ai Era!!
Well i really like the name BTW! guftaguu in hindi it means chat in private!

Ask HN: How do you give a local AI model long-term memory?

I’m running local LLMs using Ollama and hitting the usual wall:small context windows + no persistent memory = hard to solve multi-step or long-horizon tasks. For those who have built serious local setups: How do you give your model persistent memory? Vector DBs? RAG? Fine-tuned adapters? Some kind of external state management loop? Or a custom “memory module” you wrote yourself? I’m looking for...

( Read More... | 4 comments | on HN )
Yesterday, I was searching and couldn’t find it either. Now it all makes sense!
Re:SQLite JSON at Full Index Speed Using Generated Columns on Friday December 12, 2025 @01:08PM (#46246812)
Your website looks like supermemory.ai , BTW its pretty cool
Now theyrecently tightened visa rules related to pregnant travelers yk!!!
Re:Async DNS on Friday December 12, 2025 @01:04PM (#46246747)
Just curious how you approached performance bottlenecks — anything surprising you discovered while testing?