The best Hacker News stories from Show from the past day

from

All Show

from the past*

day week month

with these search terms:

Create follow Go back

Latest posts:

Show HN: I built an app to stop me doomscrolling by touching grass

i wanted to change the habit of reaching for my phone in the morning and doomscrolling away an hour so i built an app to help me. now i have to literally touch grass before accessing my most distracting apps<p>the app is built in swiftui, uses the screen time apis provided by apple and google vision to recognise grass or not<p>i'd love to get your thoughts on the concept.

Show HN: I built an app to stop me doomscrolling by touching grass

i wanted to change the habit of reaching for my phone in the morning and doomscrolling away an hour so i built an app to help me. now i have to literally touch grass before accessing my most distracting apps<p>the app is built in swiftui, uses the screen time apis provided by apple and google vision to recognise grass or not<p>i'd love to get your thoughts on the concept.

Show HN: I built an app to stop me doomscrolling by touching grass

i wanted to change the habit of reaching for my phone in the morning and doomscrolling away an hour so i built an app to help me. now i have to literally touch grass before accessing my most distracting apps<p>the app is built in swiftui, uses the screen time apis provided by apple and google vision to recognise grass or not<p>i'd love to get your thoughts on the concept.

Show HN: We made a Meta Quest3 see through walls

Show HN: AI-native browser game that users can craft unlimited 3D items

Most games have limits. You can only use preset features, need coding for customization & adding mods, require expensive extra devices. We wanted to remove those barriers.<p>That’s why we are building space zero—a browser based 3D world powered by AI. We plan to build players can freely mix items to generate unexpected creations with unique properties and sounds. Also the world itself is dynamically generated, evolving endlessly.<p>I uploaded a demo version I’ve been working on for the past month! I hope to get any feedbacks or comments :)

Show HN: Benchmarking VLMs vs. Traditional OCR

Vision models have been gaining popularity as a replacement for traditional OCR. Especially with Gemini 2.0 becoming cost competitive with the cloud platforms.<p>We've been continuously evaluating different models since we released the Zerox package last year (<a href="https://github.com/getomni-ai/zerox">https://github.com/getomni-ai/zerox</a>). And we wanted to put some numbers behind it. So we’re open sourcing our internal OCR benchmark + evaluation datasets.<p>Full writeup + data explorer here: <a href="https://getomni.ai/ocr-benchmark">https://getomni.ai/ocr-benchmark</a><p>Github: <a href="https://github.com/getomni-ai/benchmark">https://github.com/getomni-ai/benchmark</a><p>Huggingface: <a href="https://huggingface.co/datasets/getomni-ai/ocr-benchmark" rel="nofollow">https://huggingface.co/datasets/getomni-ai/ocr-benchmark</a><p>Couple notes on the methodology:<p>1. We are using JSON accuracy as our primary metric. The end goal is to evaluate how well each OCR provider can prepare the data for LLM ingestion.<p>2. This methodology differs from a lot of OCR benchmarks, because it doesn't rely on text similarity. We believe text similarity measurements are heavily biased towards the exact layout of the ground truth text, and penalize correct OCR that has slight layout differences.<p>3. Every document goes Image => OCR => Predicted JSON. And we compare the predicted JSON against the annotated ground truth JSON. The VLMs are capable of Image => JSON directly, we are primarily trying to measure OCR accuracy here. Planning to release a separate report on direct JSON accuracy next week.<p>This is a continuous work in progress! There are at least 10 additional providers we plan to add to the list.<p>The next big roadmap items are: - Comparing OCR vs. direct extraction. Early results here show a slight accuracy improvement, but it’s highly variable on page length.<p>- A multilingual comparison. Right now the evaluation data is english only.<p>- A breakdown of the data by type (best model for handwriting, tables, charts, photos, etc.)

Show HN: Benchmarking VLMs vs. Traditional OCR

Vision models have been gaining popularity as a replacement for traditional OCR. Especially with Gemini 2.0 becoming cost competitive with the cloud platforms.<p>We've been continuously evaluating different models since we released the Zerox package last year (<a href="https://github.com/getomni-ai/zerox">https://github.com/getomni-ai/zerox</a>). And we wanted to put some numbers behind it. So we’re open sourcing our internal OCR benchmark + evaluation datasets.<p>Full writeup + data explorer here: <a href="https://getomni.ai/ocr-benchmark">https://getomni.ai/ocr-benchmark</a><p>Github: <a href="https://github.com/getomni-ai/benchmark">https://github.com/getomni-ai/benchmark</a><p>Huggingface: <a href="https://huggingface.co/datasets/getomni-ai/ocr-benchmark" rel="nofollow">https://huggingface.co/datasets/getomni-ai/ocr-benchmark</a><p>Couple notes on the methodology:<p>1. We are using JSON accuracy as our primary metric. The end goal is to evaluate how well each OCR provider can prepare the data for LLM ingestion.<p>2. This methodology differs from a lot of OCR benchmarks, because it doesn't rely on text similarity. We believe text similarity measurements are heavily biased towards the exact layout of the ground truth text, and penalize correct OCR that has slight layout differences.<p>3. Every document goes Image => OCR => Predicted JSON. And we compare the predicted JSON against the annotated ground truth JSON. The VLMs are capable of Image => JSON directly, we are primarily trying to measure OCR accuracy here. Planning to release a separate report on direct JSON accuracy next week.<p>This is a continuous work in progress! There are at least 10 additional providers we plan to add to the list.<p>The next big roadmap items are: - Comparing OCR vs. direct extraction. Early results here show a slight accuracy improvement, but it’s highly variable on page length.<p>- A multilingual comparison. Right now the evaluation data is english only.<p>- A breakdown of the data by type (best model for handwriting, tables, charts, photos, etc.)

Show HN: Benchmarking VLMs vs. Traditional OCR

Vision models have been gaining popularity as a replacement for traditional OCR. Especially with Gemini 2.0 becoming cost competitive with the cloud platforms.<p>We've been continuously evaluating different models since we released the Zerox package last year (<a href="https://github.com/getomni-ai/zerox">https://github.com/getomni-ai/zerox</a>). And we wanted to put some numbers behind it. So we’re open sourcing our internal OCR benchmark + evaluation datasets.<p>Full writeup + data explorer here: <a href="https://getomni.ai/ocr-benchmark">https://getomni.ai/ocr-benchmark</a><p>Github: <a href="https://github.com/getomni-ai/benchmark">https://github.com/getomni-ai/benchmark</a><p>Huggingface: <a href="https://huggingface.co/datasets/getomni-ai/ocr-benchmark" rel="nofollow">https://huggingface.co/datasets/getomni-ai/ocr-benchmark</a><p>Couple notes on the methodology:<p>1. We are using JSON accuracy as our primary metric. The end goal is to evaluate how well each OCR provider can prepare the data for LLM ingestion.<p>2. This methodology differs from a lot of OCR benchmarks, because it doesn't rely on text similarity. We believe text similarity measurements are heavily biased towards the exact layout of the ground truth text, and penalize correct OCR that has slight layout differences.<p>3. Every document goes Image => OCR => Predicted JSON. And we compare the predicted JSON against the annotated ground truth JSON. The VLMs are capable of Image => JSON directly, we are primarily trying to measure OCR accuracy here. Planning to release a separate report on direct JSON accuracy next week.<p>This is a continuous work in progress! There are at least 10 additional providers we plan to add to the list.<p>The next big roadmap items are: - Comparing OCR vs. direct extraction. Early results here show a slight accuracy improvement, but it’s highly variable on page length.<p>- A multilingual comparison. Right now the evaluation data is english only.<p>- A breakdown of the data by type (best model for handwriting, tables, charts, photos, etc.)

Show HN: Jq-Like Tool for Markdown

There have been a few times I wanted the ability to select some text out of a Markdown doc. For example, a GitHub CI check to ensure that PRs / issues / etc are properly formatted.<p>This can be done to some extent with regex, but those expressions are brittle and hard to read or edit later. mdq uses a familiar pipe syntax to navigate the Markdown in a structured way.<p>It's in 0.x because I don't want to fully commit to the syntax being stable, in case real-world testing shows that the syntax needs tweaking. But I think the project is in a pretty good spot overall, and would be interested in feedback!

Show HN: Jq-Like Tool for Markdown

There have been a few times I wanted the ability to select some text out of a Markdown doc. For example, a GitHub CI check to ensure that PRs / issues / etc are properly formatted.<p>This can be done to some extent with regex, but those expressions are brittle and hard to read or edit later. mdq uses a familiar pipe syntax to navigate the Markdown in a structured way.<p>It's in 0.x because I don't want to fully commit to the syntax being stable, in case real-world testing shows that the syntax needs tweaking. But I think the project is in a pretty good spot overall, and would be interested in feedback!

Show HN: Jq-Like Tool for Markdown

There have been a few times I wanted the ability to select some text out of a Markdown doc. For example, a GitHub CI check to ensure that PRs / issues / etc are properly formatted.<p>This can be done to some extent with regex, but those expressions are brittle and hard to read or edit later. mdq uses a familiar pipe syntax to navigate the Markdown in a structured way.<p>It's in 0.x because I don't want to fully commit to the syntax being stable, in case real-world testing shows that the syntax needs tweaking. But I think the project is in a pretty good spot overall, and would be interested in feedback!

Show HN: Japanese City Name Generator – Using a Simple 3-Layer MLP

I trained and deployed my first model: a Japanese city name generator using just a 3-layer MLP under the hood. It runs in the browser fully locally on the onnx runtime.<p>Trained on <2,000 real Japanese city names, what's interesting is that on this simple task the simple MLP performed better than more complex models which tended to overfit and generate existing names.

Show HN: BookWatch – Animated book summaries for visual learners

Hey HN! I’m Miran Antamian, founder and CEO of BookWatch, and I’m thrilled to introduce BookWatch.com. The first AI-powered Visual Library for learners who hate reading but love growing.<p>After 1.5 years of building, 65,000+ mobile app users, and countless animated summaries, we’re launching our New Web App to make learning even more effortless. The Problem: Visual Learners Can’t Learn from Books in Text or Audio Formats Did you know that 65% of people identify as visual learners? Yet most book knowledge remains trapped in text format. Many of us want the wisdom of great books but struggle with: - Finding time to read entire books - Maintaining focus through hundreds of pages - Remembering key concepts without visual reinforcement - Traditional reading is slow, boring, and leaves millions of learners behind.<p>Our Solution: Learn by Watching, Not Reading<p>BookWatch uses proprietary AI technology to transform non-fiction bestsellers into animated video summaries that capture the core ideas in a visually engaging format.<p>Some animated book summaries you might like:<p>- Hackers and Painters by Paul Graham: <a href="https://www.bookwatch.com/videos/aGtpl1QLtyzN6JlqRze7" rel="nofollow">https://www.bookwatch.com/videos/aGtpl1QLtyzN6JlqRze7</a><p>- Zero to one by Peter Thiel: <a href="https://www.bookwatch.com/videos/rSvhERHHBoF8bpTlXnyi" rel="nofollow">https://www.bookwatch.com/videos/rSvhERHHBoF8bpTlXnyi</a><p>- The lean startup by Eric Reese: <a href="https://www.bookwatch.com/videos/dTwT3JdLWvpAsOVGbePa" rel="nofollow">https://www.bookwatch.com/videos/dTwT3JdLWvpAsOVGbePa</a><p>- Blitz Scaling by Reed Hoffman and Chris Y: <a href="https://www.bookwatch.com/videos/ogoFpks8aRGn8OgmKP2F" rel="nofollow">https://www.bookwatch.com/videos/ogoFpks8aRGn8OgmKP2F</a><p>What Makes BookWatch Special?<p>(1) AI-Powered Personal Book Expert Every video comes with an AI expert that knows everything about that specific book. Ask how to apply concepts to your unique situation: "How can I use these ideas to improve my startup?" or "How would this apply to my leadership role?". Our AI expert will give you specific advice based on the book on how to apply it in your life or business situation.<p>(2) Smart Note-Taking System Our one-click note system captures what's being said at that exact moment in the video. No need to pause or get distracted - just click and continue watching! Review and organize your notes later.<p>(3) AI Book Recommendation Engine Don't know where to start? Tell our AI what you want to learn ("I want to be a better public speaker" or "I want to be more productive" or “I want to learn b2b sales”), and it'll recommend the perfect books from our library.<p>(4) Progress Tracking & Gamification Stay motivated with achievement badges, learning streaks, and a visual representation of your growing knowledge base. See how many days have you been learning and which books helped you more.<p>Business Model<p>We've adopted a freemium approach: Free tier: 60 minutes of learning per day Premium: $89/year or $13/month for unlimited access<p>Where to Find Us<p>Web app: BookWatch.com iOS & Android apps: Search "BookWatch" in app stores Check out our growing YouTube community<p>Over 65,000 users have learned with BookWatch and we have grown on youtube organically to 100,000+ subscribers and 4.5M views!<p>I'd love to hear your thoughts and feedback! What books would you like to see in visual format first? What features would make this even more valuable for your learning journey?<p>Thank you HN community!

Show HN: BookWatch – Animated book summaries for visual learners

Hey HN! I’m Miran Antamian, founder and CEO of BookWatch, and I’m thrilled to introduce BookWatch.com. The first AI-powered Visual Library for learners who hate reading but love growing.<p>After 1.5 years of building, 65,000+ mobile app users, and countless animated summaries, we’re launching our New Web App to make learning even more effortless. The Problem: Visual Learners Can’t Learn from Books in Text or Audio Formats Did you know that 65% of people identify as visual learners? Yet most book knowledge remains trapped in text format. Many of us want the wisdom of great books but struggle with: - Finding time to read entire books - Maintaining focus through hundreds of pages - Remembering key concepts without visual reinforcement - Traditional reading is slow, boring, and leaves millions of learners behind.<p>Our Solution: Learn by Watching, Not Reading<p>BookWatch uses proprietary AI technology to transform non-fiction bestsellers into animated video summaries that capture the core ideas in a visually engaging format.<p>Some animated book summaries you might like:<p>- Hackers and Painters by Paul Graham: <a href="https://www.bookwatch.com/videos/aGtpl1QLtyzN6JlqRze7" rel="nofollow">https://www.bookwatch.com/videos/aGtpl1QLtyzN6JlqRze7</a><p>- Zero to one by Peter Thiel: <a href="https://www.bookwatch.com/videos/rSvhERHHBoF8bpTlXnyi" rel="nofollow">https://www.bookwatch.com/videos/rSvhERHHBoF8bpTlXnyi</a><p>- The lean startup by Eric Reese: <a href="https://www.bookwatch.com/videos/dTwT3JdLWvpAsOVGbePa" rel="nofollow">https://www.bookwatch.com/videos/dTwT3JdLWvpAsOVGbePa</a><p>- Blitz Scaling by Reed Hoffman and Chris Y: <a href="https://www.bookwatch.com/videos/ogoFpks8aRGn8OgmKP2F" rel="nofollow">https://www.bookwatch.com/videos/ogoFpks8aRGn8OgmKP2F</a><p>What Makes BookWatch Special?<p>(1) AI-Powered Personal Book Expert Every video comes with an AI expert that knows everything about that specific book. Ask how to apply concepts to your unique situation: "How can I use these ideas to improve my startup?" or "How would this apply to my leadership role?". Our AI expert will give you specific advice based on the book on how to apply it in your life or business situation.<p>(2) Smart Note-Taking System Our one-click note system captures what's being said at that exact moment in the video. No need to pause or get distracted - just click and continue watching! Review and organize your notes later.<p>(3) AI Book Recommendation Engine Don't know where to start? Tell our AI what you want to learn ("I want to be a better public speaker" or "I want to be more productive" or “I want to learn b2b sales”), and it'll recommend the perfect books from our library.<p>(4) Progress Tracking & Gamification Stay motivated with achievement badges, learning streaks, and a visual representation of your growing knowledge base. See how many days have you been learning and which books helped you more.<p>Business Model<p>We've adopted a freemium approach: Free tier: 60 minutes of learning per day Premium: $89/year or $13/month for unlimited access<p>Where to Find Us<p>Web app: BookWatch.com iOS & Android apps: Search "BookWatch" in app stores Check out our growing YouTube community<p>Over 65,000 users have learned with BookWatch and we have grown on youtube organically to 100,000+ subscribers and 4.5M views!<p>I'd love to hear your thoughts and feedback! What books would you like to see in visual format first? What features would make this even more valuable for your learning journey?<p>Thank you HN community!

Show HN: Willpayforthis.com – Ideas people will pay for

Ah, there's a dumb easy hack to figure out what ideas people will pay for. Search "I'd pay for" on Twitter and you'll find hundreds of posts from people talking about pain points and products they'd pay for to solve them.<p>Do this enough and you realize you have to filter through a lot of slop. slop. slop.<p>I created willpayforthis.com to accumulate high signal, high quality posts and save you some time.<p>I love thoughts from the community on how I can make it better, save you time, and help you work on the best ideas.

Show HN: I Built a Visual Workflow Automation Platform – FlowRipple

FlowRipple is designed to streamline and automate business processes with ease. Whether you're a developer, business owner, or marketer, our platform lets you build custom workflows that can be triggered by events from your applications, webhooks, or on a schedule. We’ve just gone live and are offering an exclusive Early Access Program with some incredible perks to get you started.

Show HN: Txeo – A Modern C++ Wrapper for TensorFlow

Txeo is a lightweight and intuitive C++ wrapper for TensorFlow, designed to simplify TensorFlow C++ development while preserving high performance and flexibility. Built entirely with Modern C++, Txeo allows developers to use TensorFlow with the ease of a high-level API, eliminating the complexity of its low-level C++ interface.

Show HN: Txeo – A Modern C++ Wrapper for TensorFlow

Txeo is a lightweight and intuitive C++ wrapper for TensorFlow, designed to simplify TensorFlow C++ development while preserving high performance and flexibility. Built entirely with Modern C++, Txeo allows developers to use TensorFlow with the ease of a high-level API, eliminating the complexity of its low-level C++ interface.

Show HN: Txeo – A Modern C++ Wrapper for TensorFlow

Txeo is a lightweight and intuitive C++ wrapper for TensorFlow, designed to simplify TensorFlow C++ development while preserving high performance and flexibility. Built entirely with Modern C++, Txeo allows developers to use TensorFlow with the ease of a high-level API, eliminating the complexity of its low-level C++ interface.

Show HN: I built an AI voice agent for Gmail

Hello again, HN! I’ve been using my DSL to create new voice experiences.<p>I’ve made an AI-powered email client for Gmail that you talk to, using your microphone. (I <i>highly recommend</i> using earbuds or headphones! Or the best is with Ray-Ban Meta glasses.)<p>Some fun things: Every user’s agent has a slightly different personality. You can train it by asking it to remember things for next time. And it presents some generative UI while you use it.<p>This is the first time I’m showing this publicly. I’d love your feedback! What works well, and what doesn’t?<p>I previously did a Show HN for ‘D&D meets Siri’: <a href="https://news.ycombinator.com/item?id=41328794">https://news.ycombinator.com/item?id=41328794</a>. I’m thinking of releasing the framework/DSL that I’m using to craft these experiences. Would that be interesting? Would you want to build voice apps?