Researchers cobbled together funding and time to show how quantum computing could aid in the development of drugs to help underserved populations and combat rare diseases.
Lip Sync AI - Create Your Talking Videos Instantly Comments URL: https://news.ycombinator.com/item?id=48879798 Points: 2 # Comments: 0
I'm a experienced/senior developer which is frequently using ai, guiding coding agents, etc. I wonder, how does my job look like in a few years? Which skills might be the best ones to have? Currently, having business knowledge, development experience helps greatly with guiding coding agents, creating MVPs/PoCs in "no time", improving code, etc. But what if coding agents/ai would overtake this job? Comments URL: https://news.ycombinator.com/item?id=48879727 Points: 4 # Comments: 6
Anthropic analyzed 1.2 million Claude Cowork sessions from more than 600,000 organizations. About half of all usage goes toward business processes and text creation, what Anthropic calls "the work around the work." That means tasks like compiling status reports, building onboarding checklists, or putting together slide decks. Software development barely shows up in Cowork because developers stick with Claude Code for that. The article Claude Cowork's biggest use case is the mundane office work n
Article URL: https://runewardd.github.io/runeward/ Comments URL: https://news.ycombinator.com/item?id=48879713 Points: 1 # Comments: 0
Article URL: https://thehill.com/policy/technology/5964282-privacy-concerns-instagram-ai/ Comments URL: https://news.ycombinator.com/item?id=48879675 Points: 1 # Comments: 0
OpenAI CEO Sam Altman now says he's "pretty sure" AI has created more jobs than it's eliminated. That's a sharp turn from his earlier warnings about entire professions disappearing. Anthropic CEO Dario Amodei is walking back similar claims, too. But studies so far back neither the old doomsday predictions nor optimism. The article OpenAI CEO Altman is now "pretty sure" AI is net job-creating, which is quite the pivot from predicting mass layoffs appeared first on The Decoder .
Article URL: https://www.wsj.com/tech/ai/anti-ai-activists-disappearance-sam-kirchner-6872879f Comments URL: https://news.ycombinator.com/item?id=48879571 Points: 1 # Comments: 0
The hormone appears to be involved in impulse control. ScienceAlert stories are written, fact-checked, and edited by humans, never generated by AI. Don't miss a story, subscribe here.
Hi HN! This is my first ever open-source project. It's nothing fancy, but it scratches an itch: claude --resume only works if you remember which directory the session was in. dejaview is a TUI that reads the transcripts in ~/.claude/projects/, groups sessions by project, and lets you jump back into any of them with Enter. Read-only, local-only, no config. Built with Textual. Try it: uvx dejaview Feedback is very welcome! Comments URL: https://news.ycombinator.com/item?id=48879501 Points: 1 # Com
Experimental Godot 4.7 project running Gemma 4 inference entirely in GDScript and Vulkan compute shaders. Comments URL: https://news.ycombinator.com/item?id=48879481 Points: 1 # Comments: 0
Article URL: https://discuss.huggingface.co/t/finetuning-a-reasoning-llm-with-supervised-or-reinforcement-learning/176449 Comments URL: https://news.ycombinator.com/item?id=48879403 Points: 2 # Comments: 0
An economics professor at Brown University suspects most of his 86 students used AI to cheat on a take-home exam that averaged 96 percent. When he made the final an in-person test, 18 students dropped the course, nine didn't show up, and the average fell to 48.6 percent. Two large studies from China and UC Berkeley back up his case: where students lean on AI for homework, their proctored exam scores tank. The article Grades dropped from 96 to 48 percent when a Brown professor made students take
<p>Climate change is one of the biggest challenges of our generation, and carbon credits have become an important tool in reducing global emissions. But one question kept coming to my mind:</p> <p><strong>How do we know whether a carbon project is actually making the impact it claims?</strong></p> <p>That question inspired me to build <strong>CarbonSense</strong>—an AI-powered platform that explores how satellite imagery and Google's Gemini can help make carbon project monitoring more transparen
<pre style='white-space:pre-wrap;width:81ex'>improve(ui): unify the New session draft with the start-screen layout (#105069) The /new draft page now renders the same welcome block as the empty-chat start screen, with the draft controls folded in launcher-style: hero on top, a quiet borderless target row (agent, node, folder, worktree) directly above the mid-screen composer, and the recent chats below. The page reuses renderWelcomeState via hint/composer slots so the surfaces cannot drift apart;
Article URL: https://neutralityproject.org/results.html Comments URL: https://news.ycombinator.com/item?id=48879363 Points: 4 # Comments: 0
<p>Most APIs are "agent-compatible": an agent <em>can</em> call them. Far fewer are <em>agent-native</em>: an agent can accomplish the whole job on its own, with no human un-sticking it halfway. We spent a few weeks turning <a href="https://send16.com" rel="noopener noreferrer">Send16</a> (an email platform) from the first into the second, and the gap turned out to be smaller and weirder than expected. Here's the whole thing.</p> <h2> The wall: onboarding, not the API </h2> <p>Our send endpoint
<pre style='white-space:pre-wrap;width:81ex'>test(agents): skip region-blocked xAI models</pre>
<pre style='white-space:pre-wrap;width:81ex'>feat(agents): sessions_search full-text recall over past session transcripts (#105057) Adds a core sessions_search agent tool backed by a SQLite FTS5 index that lives next to the transcript rows in the per-agent database. Indexing is transactional: user/assistant text is indexed inside the same write transaction that persists the event, deletes drop index rows in the same transaction, and branch rewinds mark the session dirty so the next search rebuil
<p>I build integrations for a living: ecommerce to ERP, CRM to invoicing, and lately ERPs to AI agents. Every one of them has the same quiet security problem, and most "ISO 27001 for developers" content misses it because it stays at the policy level.</p> <p>Here is the thing nobody says out loud: <strong>integrating two systems means taking data out of the place where it lived protected and moving it across a network you do not control.</strong> Names, tax IDs, amounts, addresses, emails. That j
<p><em>Generating content is interesting. Generating content, creating artwork, and publishing it without human intervention is where things start to get fun.</em></p> <h2> The Final Challenge </h2> <p>By this point, WizardThoughts had come a long way.</p> <p>The platform could:</p> <p>✅ Generate content using Amazon Bedrock<br><br> ✅ Generate matching image prompts<br><br> ✅ Create original artwork using Stability AI<br><br> ✅ Store assets in Amazon S3<br><br> ✅ Track everything in DynamoDB </p
<p>You're staring at a tender. £2M/year, five years, NHS trust. Looks good on paper. The deadline is six weeks out and your team can probably handle it.</p> <p>So you start writing. Three weeks later, you're 40 pages deep and someone asks: "Did we check how many incumbents are bidding?"</p> <p>Three. Plus two others you hadn't heard of.</p> <p>Now you're in a bidding war with a 15% markup, defending against suppliers who've held the contract for a decade. The win probability was never better tha
<p>A <a href="https://dev.to/annavi11arrea1/simple-benchmark-review-ollama-on-jetson-nano-5gee">Jetson Nano and Ollama benchmark review</a> appeared among DEV's top articles on <strong>2026-07-12</strong>. Edge inference experiments are valuable because they make constraints physical: memory is finite, heat accumulates, and a product cannot hide behind average cloud latency.</p> <p>They are also easy to overgeneralize. Before comparing numbers, capture the test envelope.</p> <h2> Device and soft
<p><a href="https://www.iroh.computer/blog/mesh-llm" rel="noopener noreferrer">Mesh LLM</a> reached 215 points in the Hacker News snapshot I reviewed at <strong>2026-07-12 08:00 UTC</strong>. Its author describes an OpenAI-compatible API that can run locally, route to a peer, or split model layers across multiple machines.</p> <p>That is an appealing use of existing hardware. Operationally, however, a collection of GPUs becomes capacity only after the platform can explain what happens when one n
"So many important endocrine functions beyond just reproduction." ScienceAlert stories are written, fact-checked, and edited by humans, never generated by AI. Don't miss a story, subscribe here.
Hi HN, I’ve been working on Attestor, an open-source execution boundary for autonomous AI agents. Comments URL: https://news.ycombinator.com/item?id=48879241 Points: 3 # Comments: 0
The AgenticSTS project replaces the ever-growing chat log of AI agents with five separate memory layers. Tested on the card game Slay the Spire 2, the prompt stays at around 5,000 tokens instead of ballooning past 500,000. The agent wins 6 out of 10 games, while competing agents don't win any. The article AI agents win at Slay the Spire 2 after researchers replace growing chat logs with structured memory appeared first on The Decoder .
Article URL: https://www.hackster.io/c_m_cooper/announce-for-unify-e9c652 Comments URL: https://news.ycombinator.com/item?id=48879183 Points: 1 # Comments: 0
Article URL: https://www.inkfold.app/ Comments URL: https://news.ycombinator.com/item?id=48879174 Points: 1 # Comments: 0
Bots, not people, are now the ones who use the internet the most
Article URL: https://buy.stripe.com/5kQ5kEaUKd8w0qr0FKe3e00 Comments URL: https://news.ycombinator.com/item?id=48879076 Points: 1 # Comments: 0
Article URL: https://www.evolvinglab.ai/blog/research-world-model Comments URL: https://news.ycombinator.com/item?id=48879047 Points: 1 # Comments: 0
I think from this blog you may misunderestimate how absolutely giddy I am about AI. I did hacking from 2007-2014, after that my whole career has been devoted to AI. I love the progress. I’m so excited for the new LLMs, self driving cars, video generation models, and coding agents. I set up a Linux box with opencode on my local GLM-5.2 last week and wow like just saying install tmux with the geohot configuration works; the Year of the Linux Desktop is finally here!
Article URL: https://defragzone.substack.com/p/the-ten-commandments-of-ai-usage Comments URL: https://news.ycombinator.com/item?id=48878905 Points: 1 # Comments: 0
Article URL: https://open.spotify.com/track/50mUiidUauqseqtm2Kr7U6 Comments URL: https://news.ycombinator.com/item?id=48878861 Points: 2 # Comments: 0
Article URL: https://github.com/ktock/llmlet Comments URL: https://news.ycombinator.com/item?id=48878859 Points: 1 # Comments: 0
Article URL: https://apps.apple.com/us/app/msk-ai-cto-agent/id6778822216 Comments URL: https://news.ycombinator.com/item?id=48878839 Points: 1 # Comments: 0
Article URL: https://www.washingtonpost.com/national-security/2026/07/06/why-anthropic-alleges-chinese-firms-are-distilling-knowledge-claude/ Comments URL: https://news.ycombinator.com/item?id=48878812 Points: 3 # Comments: 1
<p> A tiny cat that watches your AI coding agents for you </p> <p> <a href="https://www.producthunt.com/products/tablo-a-desktop-cat-widget?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/1194103?app_id=339">Link</a> </p>
Article URL: https://www.reuters.com/world/china/chinas-deepseek-developing-its-own-ai-chip-sources-say-2026-07-07/ Comments URL: https://news.ycombinator.com/item?id=48878697 Points: 3 # Comments: 0
