News for hackers. Stuff that matters.

mnoronha

Re:The Perverse Incentives of Vibe Coding on Wednesday May 14, 2025 @10:29PM (#43991261)
Agree. My favorite workflow has been chatting with the LLM in the assistant panel of Zed, then making inline edits by prompting the AI with the context of that chat. That way, I can align with the AI on how the problem should be solved before letting it loose. What's great about this depending on how easy or hard the problem is for the LLM, I can shift between handholding / manual coding and vibe coding.
right -- much of the reasoning is also an argument against being an early employee. the main difference is the employee probably has more leverage against being diluted in later rounds.
Re:Gemma 3 QAT Models: Bringing AI to Consumer GPUs on Sunday April 20, 2025 @05:31PM (#43746663)
Any idea why MLX and ollama use such different amounts of ram?

Show HN: Infra.new, an AI Agent for Terraform

https://infra.new/

Hey HN, we built infra.new, an AI agent for Terraform that configures your infrastructure for cost and quality.General-purpose AI tools like ChatGPT and Cursor are most helpful for coding when the cost of running the code and verifying the output is low. For example, if you’re making a landing page, it’s cheap and (relatively) low-risk to let an LLM edit the files and check if the outputs look...

( Read More... | no comments | on HN )
Re:The Butter Thesis (2019) on Tuesday February 11, 2025 @04:46PM (#43018760)
collect this post as an nft? lol
Re:Operating System in 1,000 Lines – Intro on Wednesday January 08, 2025 @09:39AM (#42634703)
this comment reads... a lot like chatgpt

Ask HN: Does an application-level query optimizer like this exist?

Recently, I did a round of optimizations on an old side-project. There turned out to be two types of changes I made which made the bulk of the difference:1) Rewriting queries that were pulling data out of the DB that wasn't all used 2) Adding in-memory caching for repeated queriesA possible third, which I didn't need but could see being beneficial:3) Querying data in bulk when possible (e.g. n...

( Read More... | no comments | on HN )
Re:A decoder-only foundation model for time-series forecasting on Saturday February 03, 2024 @02:01AM (#39238156)
ah yes, the manifold of the universe (god) :)
I think this tradeoff was to keep it compatible with Collegeboard's (archaic) calculator policies, which forbid keyboards. [1] Being able to market the calculator as SAT-approved is probably super valuable since most students won't buy multiple calculators.

[1]: https://apstudent.collegeboard.org/takingtheexam/exam-polici...

Re:Note to Google Employees from CEO Sundar Pichai on Tuesday August 08, 2017 @07:04AM (#14956515)
"It’s true that women are socialized to be better at paying attention to people’s emotional needs and so on — this is something that makes them better engineers, not worse ones."

Doesn't sound like a claim that women naturally possess these skills to me.