Ask HN: What Are You Working On? (December 2025)

I am also writing an "indie book" on getting to $100k revenue (and I write it inside LakyAI, no less!).

It uses a pipeline involving several AI models, including Nano Banana and OpenAI’s gpt-image-1 for the raw image generation.

https://universymbols.com/

I'm building out a new concept around training AI computer use agents on real sensitive tasks without PII exposure. My first demo releasing soon is a dataset of AI agent with human assisted tasks on things like paying my personal credit card or doing bank transfers.

Main things: 1. I don't modify the website I operate on 2. I take full videos and record all AI agent logs and all human actions 3. I don't modify any of those logs and will release them to the public.

I am working towards a future where AI companies are paid to generate the data they need for AI agent operations instead of paying massive sums to generate synthetic data. Imagine a future where labeling companies are completely sidestepped by simply training on production tasks directly.

Redactsure.com

Infrastructure architects think in terms of building blocks in "high-level designs" and those building blocks are often socialised/expressed in Visio/Spreadsheets. Thinking in building blocks is now more necessary than ever because of the sheer size of the infra being designed/deployed.

This approach is problematic after the design phase because there's a lossy translation to where the low-level design lives, often referred to as the Source of Truth, like NetBox.

NetBox Designs allows users to express composable, versioned, and templatizable building blocks that can be rendered to low level designs. No lossy translations, and you can always check in the future "does my LLD still match my HLD and if not, where?"