HackerNews

The End of an Era

Article URL: https://hughhowey.com/the-end-of-an-era/ Comments URL: https://news.ycombinator.com/item?id=49121980 Points: 254 # Comments: 278
HackerNews

Moonshot built on 20k Nvidia chip cluster from Alibaba

Article URL: https://www.bloomberg.com/news/articles/2026-07-31/moonshot-s-kimi-built-on-20-000-nvidia-chip-cluster-from-alibaba Comments URL: https://news.ycombinator.com/item?id=49122838 Points: 50 # Comments: 31
HackerNews

Situational Awareness Down 67% in July in AI Stock Rout

Article URL: https://www.wsj.com/finance/investing/situational-awareness-down-67-in-july-in-ai-stock-rout-cd19901f Comments URL: https://news.ycombinator.com/item?id=49122994 Points: 93 # Comments: 92
HackerNews

U.S. debt-to-GDP ratio reaches 123%

Article URL: https://fred.stlouisfed.org/series/GFDEGDQ188S Comments URL: https://news.ycombinator.com/item?id=49123258 Points: 10 # Comments: 0
HackerNews

Better to Beg Forgiveness

Article URL: https://pluralistic.net/2026/07/31/just-do-it/ Comments URL: https://news.ycombinator.com/item?id=49123300 Points: 27 # Comments: 10
HackerNews

Elevators

Article URL: https://john.fun/elevators Comments URL: https://news.ycombinator.com/item?id=49124218 Points: 30 # Comments: 3
HackerNews

Is AI Reasoning Right for the Wrong Reasons?

Article URL: https://www.quantamagazine.org/is-ai-reasoning-right-for-the-wrong-reasons-20260731/ Comments URL: https://news.ycombinator.com/item?id=49124358 Points: 27 # Comments: 3
🌍 TheVerge

Sony pushes forward with ditching discs, despite backlash

Sony has received a lot of backlash from PlayStation fans since announcing that it's killing physical game disc production, but that hasn't swayed its decision. During Sony's latest earnings call, chief financial officer Lin Tao said that while the company "put in a lot of thought and time" when considering the opposition put forward by […]
🌍 Dev.to

Your RAG copilot can't count — stop letting it try

Your RAG copilot can't count — stop letting it try A user asked our document-search copilot a completely reasonable question: how many documents a specific person authored. The copilot answered with a number. Confidently. With a nice sentence around it. The real answer, computed by the database, was 84 . The copilot said — because it did what every RAG system does when you ask it an aggregation question: it counted what it could see. Here's the claim I want to defend, because I think ...
🌍 Dev.to

A Consent-First Phone Privacy Triage Flow for Support Teams

Privacy incidents on personal phones often begin with a vague report: battery drain, unfamiliar login alerts, a device listed in a messaging account, or an app with permissions that do not match its purpose. Support teams need a repeatable way to respond without escalating fear or encouraging unauthorized access. This article presents a consent-first triage flow that developers, help-desk teams, and security educators can adapt for user-facing support. Start with an authorization ...
How BrowserAct Fixed the Stale-Selector Failures Breaking My Browser Tasks
🌍 Dev.to

How BrowserAct Fixed the Stale-Selector Failures Breaking My Browser Tasks

Disclosure: BrowserAct sponsored this piece. The BrowserAct links below are affiliate-tracked — I get credit if you sign up through them. I keep hitting the same failure building agent workflows in Claude Code: the agent captures a selector once, the page re-renders with a new build hash, and the next run breaks on an element that's still right there on the screen — just under a different id. BrowserAct is a browser automation platform built for AI agents — real browser control, persistent b...
🌍 Dev.to

A user spent four days designing a feature for my project. The right answer was zero lines of code.

Last week I got the best feature request my open-source project has ever received. It was detailed, technically literate, fail-closed by design, and came from someone who clearly runs the tool harder than I do. I'm not going to build it. And the person who asked for it agrees — because the conversation ended with him shipping the fix, in his own layer, as a package other people can now use. I want to walk through how that happened, because the whole arc is a case study in a question I th...
Your Data Warehouse Is Not Your Data Pipeline
🌍 Dev.to

Your Data Warehouse Is Not Your Data Pipeline

Teams keep forcing their warehouse to do integration work it was never designed for. The result is ballooning costs, opaque failures, and architectures that become harder to maintain the more they ‘succeed.’ Here’s the case for separating data movement from analytics storage. The expensive truth about modern data stacks Spend enough time around data platform teams and you hear the same story. A company builds out its "modern data stack" — warehouse, processing layer, orchestrator...
🌍 Dev.to

Gemini Robotics 2 Has Not Been Announced: What Google DeepMind Actually Offers

Google DeepMind has not announced a product called Gemini Robotics 2 . The name does not appear in the supplied first-party material describing the company’s robotics work. What is documented is the broader Gemini Robotics family, including Gemini Robotics 1.5, Gemini Robotics-ER 1.6, and On-Device variants. That distinction is important for technology leaders assessing robotics AI. Product names are not cosmetic details in enterprise planning. They shape vendor evaluations, architecture dis...
🌍 Dev.to

Building an Elden Ring Build Calculator: What Game Developers Can Learn From RPG Stat Systems

When players talk about RPG builds, they usually talk about weapons, armor, skills, and damage numbers. But behind every build decision is a much more interesting problem: A stat optimization problem. Games like Elden Ring look simple on the surface. A player chooses: Strength Dexterity Intelligence Faith Arcane Then equips a weapon. However, the relationship between these choices is not linear. A weapon may scale with multiple attributes. A stat may provide strong...
🌍 Dev.to

Hardening an AI coding agent: the failures, and the code that fixed them

At Univoco we build retrieval-augmented assistants over a customer's own documentation. One of them is a coding agent that writes code for a proprietary document layout engine. The useful thing about that engine is that there is effectively nothing about it on the public internet. The model cannot have learned it. Everything the agent knows comes from a private documentation index and the tools we give it to search that index. That makes it an unusually clean laboratory. When memorisatio...
🌍 Dev.to

Your AI SOC Agent Has a Task Role. Does It Also Have Your S3 Credentials?

✓ Human-authored analysis; AI used for formatting and proofreading. Agentic security platforms are arriving. ECS-hosted investigation agents that invoke Bedrock models, write case artifacts to S3, mount shared session state via EFS, and expose interactive shells through ALBs. The architecture is sound. The security configuration surface is enormous. Most of it isn't covered by existing benchmarks because the benchmarks were written before agents existed. I audited a published deployme...
🌍 Dev.to

How to Connect an OpenAI SDK App to an OpenAI-Compatible API

Many applications already use the OpenAI SDK. Moving to another compatible gateway should not require rewriting the whole application. In most cases, the main changes are the API key, base URL, and model ID. This guide shows the basic migration flow with Route Key, an OpenAI-compatible AI API gateway and multi-model router. What you need Before starting, prepare: An OpenAI-compatible SDK A Route Key API key A supported model ID A test environment where you can safel...
🌍 Dev.to

Functional Error Handling in TypeScript: Beyond try-catch

Every TypeScript developer has written something like this: function getUser ( id : string ): User { const user = db . findUser ( id ); if ( ! user ) throw new Error ( " User not found " ); return user ; } The function signature says it returns a User . But that's a lie. It might throw, and nothing in the type system warns you about it. The caller has to remember to wrap it in try-catch , hope they handle the right err...