AI News Today

Wanted to try something new, decided to create a mutable crm, which re renders the UI after you apply a change. The llm only modifies the db via tool use, so he doesnt directly write sql, only choose which function to use and the params. Can only revert the last change, and dropping tables is not available to the llm. Looking for feedback, opinions, something you would do different. Taking my first steps in AI implementations. Heavily assisted by claude code Comments URL: https://news.ycombinato

if you have solid insights on AI usage in industrial process mining this is the right place to start. Comments URL: https://news.ycombinator.com/item?id=48971319 Points: 1 # Comments: 0

The science is tricky to untangle. ScienceAlert stories are written, fact-checked, and edited by humans, never generated by AI. Don't miss a story, subscribe here.

Article URL: https://github.com/chalie-ai/chalie Comments URL: https://news.ycombinator.com/item?id=48971042 Points: 4 # Comments: 1

On the latest episode of Equity, we debate whether Apple's lawsuit will cast over OpenAi's much-discussed plans to get into hardware and go public.

Article URL: https://github.com/AdamACE9/bothread Comments URL: https://news.ycombinator.com/item?id=48970865 Points: 1 # Comments: 0

<pre style='white-space:pre-wrap;width:81ex'>fix(system-agent): preserve setup success on audit failure (#111504)</pre>

Break reminders are your friend.

Article URL: https://www.thecareertoolkitbook.com/blog/ai-requires-a-labor-market-bailout Comments URL: https://news.ycombinator.com/item?id=48970332 Points: 2 # Comments: 0

Article URL: https://www.wsj.com/tech/ai/ai-chip-startup-etched-is-in-talks-for-20-billion-valuation-caf1787d Comments URL: https://news.ycombinator.com/item?id=48970146 Points: 2 # Comments: 1

I would never go so far as to say there's no place for AI in music (I'm a fan of Holly Herndon, after all). But I generally find music made with generative AI to be offensively boring, especially the outputs of Suno. So I'm having a bit of a tough time processing the fact that […]

Article URL: https://vivekhaldar.com/articles/compiling-an-ai-agent-skill/ Comments URL: https://news.ycombinator.com/item?id=48969971 Points: 1 # Comments: 0

Hello HN! I found my self always doing some kind of polling loop for my applications when i want to watch data as it comes in, however i find striking the balance between aggressive polling and the responsiveness tricky. So i wrote (with help from claude) pgnudg. pgnudge is a small async Python library that wakes your app when specific Postgres tables change. It only tells you which tables moved. You refetch, and the database stays the source of truth.It reads the WAL over the replication protoc

Article URL: https://github.com/Helldez/BigMoeOnEdge Comments URL: https://news.ycombinator.com/item?id=48969917 Points: 1 # Comments: 0

Article URL: https://github.com/bjoern-janson/resolution-horizon Comments URL: https://news.ycombinator.com/item?id=48969902 Points: 1 # Comments: 0

Article URL: https://arxiv.org/abs/2607.04003 Comments URL: https://news.ycombinator.com/item?id=48969887 Points: 3 # Comments: 0

Article URL: https://osf.io/preprints/psyarxiv/5y6m4_v1 Comments URL: https://news.ycombinator.com/item?id=48969839 Points: 3 # Comments: 1

<p> Your personal context file for every agent. </p> <p> <a href="https://www.producthunt.com/products/creed-2?utm_campaign=producthunt-atom-posts-feed&utm_medium=rss-feed&utm_source=producthunt-atom-posts-feed">Discussion</a> | <a href="https://www.producthunt.com/r/p/1200738?app_id=339">Link</a> </p>

Let's discover how neural networks learn, step by step The post Backpropagation Explained for Beginners (Part 1): Building the Intuition appeared first on Towards Data Science .

According to the New York Times, federal agents have been told that the FBI will no longer be investigating confrontations involving ICE agents. The DHS and DOJ denied the change in policy to The Times. The reported change in guidance follows renewed scrutiny of violence by ICE agents who have killed two people in the […]

<pre style='white-space:pre-wrap;width:81ex'>fix(agents): keep web search error serialization surrogate-safe (#111327) * fix(agents): keep web search error serialization surrogate-safe * test(agents): pin web search truncation prefix Preserve the pre-existing provider_error fallback while proving the exact UTF-16-safe retained prefix.\n\nCo-authored-by: zenglingbiao <zeng.lingbiao@xydigit.com> --------- Co-authored-by: Peter Steinberger <steipete@gmail.com></pre>

<blockquote> <p><strong>Abstract.</strong> Notes from testing whether to move KeibiDrop, a post-quantum encrypted peer-to-peer filesystem, from gRPC over TCP to gRPC over QUIC. The product reads remote files on demand, so the goal is to survive an IP change, keep seek latency low, stay correct, then go fast, in that order. Measured on a 12-core macOS laptop, a 4-core Linux VPS, and a real link between a laptop in Barcelona and the VPS at 72 ms round trip. Findings: on macOS the QUIC transport is

<pre style='white-space:pre-wrap;width:81ex'>fix(ai): honor Retry-After when retry-after-ms is unparseable (#111353) * fix(ai): honor Retry-After when retry-after-ms is unparseable The ChatGPT Responses retry path returned as soon as `retry-after-ms` was present, so a malformed or empty value discarded a valid `Retry-After` header sent alongside it and the client fell back to blind exponential backoff. The sibling parser in `provider-transport-fetch.ts` treats the two headers as ordered preferen

<h1> Building Production-Grade LLM Evaluation Pipelines: From Vibes to Metrics </h1> <p><em>How we replaced "looks good to me" with automated evaluation catching 92% of hallucinations before deployment</em></p> <h2> The Problem: Why "Vibe Checks" Fail in Production </h2> <p>Three months ago, our team shipped a RAG-based customer support assistant. It worked great in testing — we'd ask it questions, read the answers, and say "yeah, that looks right."</p> <p>Then it hit production.</p> <p>A custom

<pre style='white-space:pre-wrap;width:81ex'>fix(tavily): reject blank extract URLs (#111333) * fix(tavily): reject blank extract URLs Co-authored-by: chatgpt-codex-connector[bot] <199175422+chatgpt-codex-connector[bot]@users.noreply.github.com> * fix(tavily): normalize tool string arrays Co-authored-by: VectorPeak <73048950+VectorPeak@users.noreply.github.com> --------- Co-authored-by: chatgpt-codex-connector[bot] <199175422+chatgpt-codex-connector[bot]@users.noreply.github.com> Co-authored-by:

<p> Manage your team of AI agents by voice, from anywhere </p> <p> <a href="https://www.producthunt.com/products/openbase-2?utm_campaign=producthunt-atom-posts-feed&utm_medium=rss-feed&utm_source=producthunt-atom-posts-feed">Discussion</a> | <a href="https://www.producthunt.com/r/p/1200703?app_id=339">Link</a> </p>

Article URL: https://aimaketattoo.com/ai-tattoo-generator Comments URL: https://news.ycombinator.com/item?id=48969401 Points: 1 # Comments: 0

<p> Free browser tools for everyday tasks, no login </p> <p> <a href="https://www.producthunt.com/products/free-ai-tools-4?utm_campaign=producthunt-atom-posts-feed&utm_medium=rss-feed&utm_source=producthunt-atom-posts-feed">Discussion</a> | <a href="https://www.producthunt.com/r/p/1200702?app_id=339">Link</a> </p>

I'm curious. Do you use todo list app? pen & paper? or any digital apps you're using daily? Personally, i still use Obsidian for notes tasks and sometimes Jira for following tasks with my teammates. In the Agentic AI era, does the way you "manage" tasks change? What are the moments in the normal-day workflow you have that leads to you changing your method? Would love to hear your stories and complaints. Comments URL: https://news.ycombinator.com/item?id=48969329 Points: 1 # Comments: 1

Article URL: https://charity.wtf/p/ai-demands-more-engineering-discipline Comments URL: https://news.ycombinator.com/item?id=48969326 Points: 1 # Comments: 0

Connect all the things and watch what happens

Welcome back to TechCrunch Mobility, your hub for the future of transportation and now, more than ever, how AI is playing a part.

Article URL: https://twitter.com/kimi_moonshot/status/2078855608565207130 Comments URL: https://news.ycombinator.com/item?id=48969291 Points: 35 # Comments: 4

Article URL: https://justrepl.com Comments URL: https://news.ycombinator.com/item?id=48969224 Points: 1 # Comments: 1

<pre style='white-space:pre-wrap;width:81ex'>feat: avatar upload UI, self-service profiles, server-side avatar route (#111421) * feat: avatar upload UI, self-service profiles, server-side avatar proxy * refactor(ui): make resolveAvatar synchronous so avatar updates render immediately resolveAvatar no longer does async work (gravatar moved server-side), so the until()-deferred render left the identity chip img stale after updateSelfUser. Rendering synchronously reuses the img element and reflects

<pre style='white-space:pre-wrap;width:81ex'>feat(agents): authoritative write diffs end to end (#111456) * feat(agents): authoritative write diffs end to end The write tool now reports what actually happened: no-op (changed:false), created files (created:true with a bounded numbered diff), and overwrites (created:false with a real old-to-new diff when the old content is readable text within byte, line, and edit-distance budgets — unknown removals are never fabricated). The gateway history proje

Article URL: https://hackenewhome.blogspot.com/p/my-take-on-ai-vc-benchmark-bloggers.html Comments URL: https://news.ycombinator.com/item?id=48969138 Points: 1 # Comments: 0

Article URL: https://ahaprompt.app Comments URL: https://news.ycombinator.com/item?id=48969127 Points: 1 # Comments: 0

Article URL: https://neutralityproject.org/ Comments URL: https://news.ycombinator.com/item?id=48969113 Points: 1 # Comments: 0

Three rigorous prompts to parse and recognize intent -> resulting in clear instructions for a controller which locks onto your requirements. Lets you get work done without the usual role-play. Comments URL: https://news.ycombinator.com/item?id=48969107 Points: 2 # Comments: 0