Feed aggregator

Article URL: https://nautil.us/the-books-that-blew-these-scientists-minds-1267452/

Comments URL: https://news.ycombinator.com/item?id=47054843

Points: 1

# Comments: 1

Article URL: https://github.com/singlr-ai/scim-sql

Comments URL: https://news.ycombinator.com/item?id=47054821

Points: 1

# Comments: 0

Article URL: https://github.com/semos-labs/epist

Comments URL: https://news.ycombinator.com/item?id=47054799

Points: 1

# Comments: 0

While some Hollywood icons are feeling doom and gloom over the AI-generated clip, labor unions are fighting back with legal threats.

Article URL: https://familymemories.video

Comments URL: https://news.ycombinator.com/item?id=47054341

Points: 1

# Comments: 0

Article URL: https://y-cal.vercel.app/

Comments URL: https://news.ycombinator.com/item?id=47054304

Points: 1

# Comments: 0

Think about ordering pizza. Instead of opening DoorDash, browsing restaurants, customizing your order, and checking out, you could just say "Hey AI, order me a pepperoni pizza." Done. When you think about it that way, why would anyone need DoorDash anymore?

For simple stuff like this, yeah, AI replacing apps feels inevitable. But I've been thinking about the harder cases, and I'm not so sure.

## Most real-world apps solve really messy problems

Take something like Jira. It's not just a pretty interface sitting on top of simple logic. Jira is enforcing workflows across teams, managing who can see what, tracking state changes, hooking into CI/CD, keeping audit trails, all while dozens of people with different roles are working in it at the same time. Can you honestly replace all of that with "Hey AI, manage my project"?

## And then there's the cost

Here's the part that I think people aren't talking about enough.

Let's say you need to build a complex financial model in a spreadsheet. Hundreds of rows, nested formulas, pivot tables, cross-sheet references, the whole thing. Now imagine trying to do that entirely by talking to AI.

"Move that column." "No, the other one." "Now add a VLOOKUP that pulls from the other sheet." "Actually, the range is wrong." Every single back-and-forth is burning tokens. If the model is complex enough, you could easily spend more on AI inference in one sitting than you'd pay for Excel for an entire year. And the spreadsheet app just... does it. Deterministic logic, minimal compute, instant feedback.

I think this pattern holds more broadly than people realize: the more complex and repetitive the task, the more tokens you burn, and the harder it is for "just ask AI" to compete with a $10/month app on cost alone.

Here's where it gets ironic, though. AI coding agents like Cursor and Claude Code are making it way cheaper to build apps. So AI might not shrink the app market at all. It could actually grow it by making apps cheaper to create, while simultaneously making "just use AI directly" expensive for anything non-trivial.

*Curious what HN thinks:*

- Are we overestimating AI's ability to replace purpose-built software?

- Will inference costs drop and AI capabilities advance enough to make this argument irrelevant?

- Or will everyone just become a developer, building their own apps for their own needs?

Comments URL: https://news.ycombinator.com/item?id=47054299

Points: 1

# Comments: 0

Hey HN!

OpenClaw is notorious about its token usage, and for many the price of Opus 4.6 can be cost prohibitive for personal projects. The usual workaround is “just switch to a cheaper model” (Kimi k2.5, etc.), but then you are accepting a trade off: you either eat a noticeable drop in quality or you end up constantly swapping models back and forth based on usage patterns

I packaged Arch-Router (used by HF: https://x.com/ClementDelangue/status/1979256873669849195) into Plano and now calls from OpenClaw can get automatically routed to the right upstream LLM based on preferences you set. Preference could be anything that you can encapsulate as a task. For e.g. for daily calendar and email work you could redirect calls to k2.5 and for building apps with OpenClaw you could redirect that traffic to Opus 4.6

This hard choice of choosing one model over another goes away with this release.

Comments URL: https://news.ycombinator.com/item?id=47054250

Points: 1

# Comments: 0

Article URL: https://servo.org/slides/2026-02-fosdem-servo-web-platform/

Comments URL: https://news.ycombinator.com/item?id=47054247

Points: 1

# Comments: 0

Article URL: https://www.co-span.com/

Comments URL: https://news.ycombinator.com/item?id=47054220

Points: 2

# Comments: 0

Article URL: https://hornbeam.dev/

Comments URL: https://news.ycombinator.com/item?id=47054219

Points: 1

# Comments: 0

Article URL: https://www.bbc.com/news/articles/cx2d9yk2kpjo

Comments URL: https://news.ycombinator.com/item?id=47054211

Points: 2

# Comments: 0

Article URL: https://colliery-io.github.io/graphqlite/latest/

Comments URL: https://news.ycombinator.com/item?id=47054192

Points: 2

# Comments: 0

Article URL: https://accesslint.app

Comments URL: https://news.ycombinator.com/item?id=47054175

Points: 1

# Comments: 0

Article URL: https://jacobin.com/2026/01/left-nationalism-universalism-populism-melenchon/

Comments URL: https://news.ycombinator.com/item?id=47054170

Points: 1

# Comments: 1

Article URL: https://spectrum.ieee.org/data-center-sustainability-metrics

Comments URL: https://news.ycombinator.com/item?id=47054150

Points: 1

# Comments: 0

I know the trope:

"Ideas are cheap. Execution is everything - and a non-technical founder recruiting a senior engineer for equity is a red flag."

Totally fair.

Why I ask:

I’m 23 and non-technical. For ~3 years I’ve been studying distributed systems, developer tooling, AI codegen, and an infra concept around intent-based architecture.

I’ve attempted three startups in the past 18 months. All failed at execution because the system I’m trying to build is deeply technical, and I know enough to know I can’t “vibe code” my way into it.

That creates a catch-22:

- Building a shallow version invalidates the thesis. - Building it correctly requires engineers far stronger than me. - Engineers strong enough to build it are rightly skeptical of someone like me.

So here’s my question - the thing I feel most self-imprisoned in:

How do I, as a young non-technical founder, pitch this to deep-tech, systems-level engineers without sounding like a naive "idea guy"? What should I be doing right now to make myself undeniably useful to a technical co-founder of this caliber?

I truly appreciate any insight and am entrusted with any feedback you give.

Thank you much -Tim

Comments URL: https://news.ycombinator.com/item?id=47054141

Points: 1

# Comments: 4

Article URL: https://www.experimental-history.com/p/i-swear-the-ufo-is-coming-any-minute

Comments URL: https://news.ycombinator.com/item?id=47054134

Points: 3

# Comments: 0