Hacker News AI Story Tags
Read a configurable Hacker News feed (top 30 by default) and tag each live story AI or Not AI from its title, URL and post text.
make a pipeline that reads hackernews and tags each story whether it's ai or not.
Typed to Jason, Jsonify's AI data engineer, who built the pipeline from it.
One message. 30 stories, 16 columns.
Results
30 stories, refreshed hourly
| Row# | Open row | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Not AI | PaulHoule | Hacker News title, URL and post text; linked article not read | 0 | top | false | false | 15 | 15 | 50060674 | — | Publishers Killed the Paperback, Then Blamed Readers | View row | ||||
| 2 | AI | ilreb | Hacker News title, URL and post text; linked article not read | 146 | top | true | false | 2 | 251 | 50058507 | — | GLM-Legal | View row | ||||
| 3 | AI | vishnukool | Hacker News title, URL and post text; linked article not read | 29 | top | true | false | 4 | 44 | 50059096 | Hey HN, I’m Vishnu from Agent37 ( https://agent37.com ). We build sandboxes for running AI agents at under 10% of the cost of most sandbox providers. Our customers spend an average of $0.37/month per sandbox. You can view a video walkthrough of the product here: https://youtu.be/DK5ISR2tbNo We started as one of those OpenClaw wrapper companies back in February (yup.. judge me). In time we added more and more harnesses. One of the goals I had at the time was to offer a free plan to users, so as to gain more market share. But it was too cost prohibitive to do it right away. Especially since we were bootstrapped and servers were running on my savings. Today we’re actually able to offer the free plan thanks to the optimizations we’ve made since. Over 90% of the new signups we have are just free plan users and they host their personal Hermes/Openclaw/Claude Code etc on our cloud. Our infrastructure cost for the free sandboxes averages around $0.05 to $0.10 per user per month. Here are some techniques that have worked for lowering cost: * When you look at agent harnesses, they have certain patterns in the way they consume CPU and RAM. They need CPU in bursts when they’re initializing or performing certain tasks, but beyond that, they’re often not peaking. They need more RAM for things like browser and computer-use tasks. Once the task is done, usage often comes back down. This is a key insight in lowering cost, so by efficiently managing the CPU/RAM utilization we could lower the cost. * Secondly, you have to skip the hyperscalers when renting the Hosts. I worked at Amazon before and they have 80-90% margins on VMs. So it was an obvious choice once I looked at the prices to go one layer down to vendors like Hetzner and OVH. This alone was probably one of the biggest cost levers in reducing the cost. * Thirdly, you also want to pick a technology stack that lets you efficiently cache and share resources across the host. We use gVisor for better utilization and fast boot times. It lets us use available host memory across sandboxes without reserving a full VM’s memory for each one. There are tradeoffs: our gVisor sandboxes don’t support workloads like Docker-in-Docker or Android emulators. * And finally the size and shape of the hosts you buy also makes a big difference. RAM is typically the bottleneck when it comes to cost and not the CPU. This is specifically true for agent harness workloads, so when buying servers, we optimize for packing more RAM into a given host. We also host full VM sandboxes for workloads that need them, including running Docker at a slightly higher price, but they are still pretty low cost given the rest of the techniques. Overall, my learning has been that just by focusing on a niche of sandbox usecases, and optimizing for it you can give a much better cost/performance tradeoff. I was previously a Senior Software Engineer at Amazon and an Engineering Manager at Quantcast. I built Amazon's internal AI chatbot, Diya, and their internal coding harness, Ziya. If you’re someone who’s looking to run agents on sandbox and are exploring a sandbox vendor, come check us out. We’re particularly looking for developers building messaging agents to try their workloads and challenge the limits and economics of our platform. Thanks! https://agent37.com | Launch HN: Agent37 (YC F26) – Low-cost agent sandboxes for running agents | — | View row | |||
| 4 | AI | DeepLogin | Hacker News title, URL and post text; linked article not read | 37 | top | true | false | 8 | 77 | 50060500 | — | Just say no to OpenAI and Anthropic | View row | ||||
| 5 | Not AI | djfergus | Hacker News title, URL and post text; linked article not read | 40 | top | false | false | 9 | 117 | 50055593 | — | Turn a WiFi 7 Router into a 320Msps Software Defined Radio Receiver | View row | ||||
| 6 | Not AI | joshcsimmons | Hacker News title, URL and post text; linked article not read | 19 | top | false | false | 29 | 78 | 50041573 | — | Running the unmodified Linux amdgpu driver on an iPad | View row | ||||
| 7 | Not AI | dkpk | Hacker News title, URL and post text; linked article not read | 126 | top | false | false | 28 | 137 | 50055375 | — | Incident with Actions, Pull Requests and Issues | View row | ||||
| 8 | Not AI | subset | Hacker News title, URL and post text; linked article not read | 42 | top | false | false | 25 | 233 | 50034206 | — | Calculus with Julia | View row | ||||
| 9 | AI | surprisetalk | Hacker News title, URL and post text; linked article not read | 10 | top | true | false | 13 | 57 | 50057668 | — | Kevin Kelly's 2016 AI Predictions, 10 Years Later | View row | ||||
| 10 | AI | whimblepop | Hacker News title, URL and post text; linked article not read | 19 | top | true | true | 1 | 60 | 50060694 | — | Codex CLI installs a persistent, self-updating background daemon by default | View row |
Produced by pipeline v1Operate a site this data was read from? Report a problem or request removal.
Pipeline
How Jsonify collects this data
Jason, Jsonify's AI data engineer, built this pipeline from a plain-language brief. It runs the same steps every time: read the sources, extract the fields, check every row, publish the dataset. When a source changes and a step fails, Jason diagnoses the failure, repairs the pipeline and verifies a sample before it resumes, so the dataset keeps its shape without anyone babysitting it.
- 01
Read Feed
Read a fresh ranked snapshot of the selected Hacker News feed.
- 02 · fan out
Read Story
Read each story's title, URL, post text, score, author and…
- 03 · fan out
Tag AI
Tag each story AI or not AI using its Hacker News title, URL…
- 04
Validate
Check story identities and tags, then preserve the source feed…
- 05 · output
Stories
Latest Hacker News stories with AI tags and source links.
30 stories · v1
Read Feed
Read a fresh ranked snapshot of the selected Hacker News feed.
Read Feed
Read a fresh ranked snapshot of the selected Hacker News feed.
Per input record: Read Story
Read each story's title, URL, post text, score, author and comment count from Hacker News.
Per input record: Tag AI
JudgeTag each story AI or not AI using its Hacker News title, URL and post text; flag uncertain decisions.
Validate
Check story identities and tags, then preserve the source feed ranking in the saved snapshot.
Stories
Latest Hacker News stories with AI tags and source links.
History
Versions and repairs
Every activated version of the pipeline, and the recorded outcome of each repair.
Version history
- v1
Added Hacker News story collection with AI/Not AI tags, uncertainty flags, story links and live engagement details.
Repairs
When a source changes and a step fails, Jason's repair and its outcome are recorded here.
Delivery options
Google SheetsExcel
Snowflake
S3
Postgres- Webhooks
Slack
Claude
ChatGPTCursor
- MCP
Build conversation
How was this pipeline built
The original chat with Jason, Jsonify's AI data engineer, up to the message that produced version 1. Shared by the pipeline's owner.
- User
make a pipeline that reads hackernews and tags each story whether it's ai or not.
JasonI’ll build a saved pipeline that reads Hacker News stories and tags each as AI or not AI. I’ll start with the top 30 stories; the saved data can be refreshed later or put on a schedule.
- Jason
Got it—hourly updates, and I’ll share the public page when it’s ready. At 30 stories per update, that’s about 21,600 collected rows per 30 days. Tags will use each story’s title, URL, and any text posted on Hacker News, rather than the full linked article.
- Jason
The pipeline is built and scheduled hourly. The first 30-story update is running now; I’ll verify the results and share the public page when it finishes.
Need a different one?
Describe it. Get verified rows, kept current.