Hacker News: pickleballcourt

New comment by pickleballcourt in "Work with the garage door up (2024)"

pickleballcourt — Sat, 25 Apr 2026 08:33:41 +0000

Love the title

New comment by pickleballcourt in "Launch HN: IonRouter (YC W26) – High-throughput, low-cost inference"

pickleballcourt — Fri, 13 Mar 2026 02:40:42 +0000

While your throughout is around 2x you still cost more then vercel ai model pricing for example for GLM-5: https://vercel.com/ai-gateway/models?q=glm

Is this a result of renting more expensive gpus?

The journey of how we built Manus [video]

pickleballcourt — Thu, 05 Feb 2026 12:01:14 +0000

Article URL: https://www.youtube.com/watch?v=Lm-WgmYRgmI

Comments URL: https://news.ycombinator.com/item?id=46898722

Points: 1

# Comments: 0

New comment by pickleballcourt in "Show HN: ShapedQL – A SQL engine for multi-stage ranking and RAG"

pickleballcourt — Fri, 30 Jan 2026 21:53:10 +0000

Thanks!

New comment by pickleballcourt in "Quack-Cluster: A Serverless Distributed SQL Query Engine with DuckDB and Ray"

pickleballcourt — Fri, 30 Jan 2026 21:52:40 +0000

Reminds me of smallpond from deepseek

New comment by pickleballcourt in "Show HN: ShapedQL – A SQL engine for multi-stage ranking and RAG"

pickleballcourt — Thu, 29 Jan 2026 14:45:07 +0000

Is there a major difference between pgvector and shapedql?

New comment by pickleballcourt in "Show HN: LemonSlice – Upgrade your voice agents to real-time video"

pickleballcourt — Tue, 27 Jan 2026 21:53:44 +0000

Thanks for sharing! Makes sense to go with latency first.

New comment by pickleballcourt in "Show HN: LemonSlice – Upgrade your voice agents to real-time video"

pickleballcourt — Tue, 27 Jan 2026 20:45:32 +0000

One thing I've learnt from movie production is actually what separates professional from amateur quality is in the audio itself. Have you thought about implementing personaplex from NVDIA or other voice models that can both talk and listen at the same time?

Currently the conversation still feels too STT-LLM-TTS that I think a lot of the voice agents suffer from (Seems like only Sesame and NVDIA so far have nailed the natural conversation flow). Still, crazy good work train your own diffusion models, I remember taking a look at the latest literature on diffusion and was mind blown by the advances in last years or so since u-net architecture days.

EDIT: I see that the primary focus is on video generation not audio.

Why 'Thank You' Might Be the Best Metric in AI Products

pickleballcourt — Thu, 18 Dec 2025 16:01:46 +0000

Article URL: https://flopsandfinance.substack.com/p/why-thank-you-might-be-the-best-metric

Comments URL: https://news.ycombinator.com/item?id=46314372

Points: 2

# Comments: 2

How much does it cost to generate an image?

pickleballcourt — Thu, 11 Dec 2025 12:22:48 +0000

Article URL: https://flopsandfinance.substack.com/p/how-much-does-it-cost-to-generate

Comments URL: https://news.ycombinator.com/item?id=46230506

Points: 1

# Comments: 0