Ask HN: Imagine coding LLM's 1M times faster; what uses might there be?

Comments (1)

NitpickLawyer · 1h ago

Unlimited speed would probably help things like alphaevolve. Given a codebase / paper / idea and some definable "tests" or "goals", go nuts. Evolve, adversarial, n x m tries w/ documentation, n x m tries w/ papers in that field, etc.

For the moment, though, I'd take a "smarter" but slower model over x times faster. The current models are plenty fast already. They pump out 300-500 LoC files in seconds/minutes. That's plenty speed.

The Shaping of Narrative Through Internet Means (medium.com)

ASCII Animations (ascii.co.uk)

Conk (1995) – Action Game Construction Kit (semprini.me)

Show HN: Moyu Search – A better tab search (chromewebstore.google.com)

SubBuddy: Never Miss a Payment Again (subbuddy.io)

Windows 11 August Update Triggers NVMe Controller Failures; Phison Investigates (windowsforum.com)

Palantir stock slumps 9%, falling for a fifth straight day from record (cnbc.com)

Anyone using USDT for online payments? What do you buy and how?

The LattePanda Mu (taoofmac.com)

My Take on the iPadOS 26 Beta (taoofmac.com)

Wanted: Sane, tasteful Wi-Fi 6/7 Access Points (taoofmac.com)

Reading Books vs. Engaging with Them (cold-takes.com)

Webb spots a new, tiny moon orbiting Uranus (astronomy.com)

Why Is the US Punishing India – But Not China – For Buying Russian Oil? (thediplomat.com)

Show HN: College Student Program Across Various Country (github.com)

Doctor cracks the case of a 15-year-old who suddenly couldn't walk (economictimes.indiatimes.com)

Ask HN: Is there any plugin to use OpenAI or Claude API in Xcode like Cursor?

Stupidity as a Service (lawrenc.es)

Databricks Secures Series K Funding, Surpassing $100B Valuation (jphfeeds.top)

The Russia-Ukraine War Has Made North Korea More Dangerous (thediplomat.com)

Real-Time Mode in Apache Spark Structured Streaming (databricks.com)

Open source tools to lighten your summer (open-source-ward.com)

Mirrorshades, the Cyberpunk Anthology (rudyrucker.com)

Swinsian 3 Released – The Advanced Music Player for Mac (swinsian.com)

OpenAI makes GPT-5 'friendlier' after widespread user backlash (pcworld.com)

SPL Lightweight Multisource Mixed Computation Practices (github.com)

Show HN: I've launched my app on Hacker News and here is what happened (indiehackers.com)

Show HN: File-Fabric – Generate Files (docs, media, archives) for Testing

Show HN: BlipCut – AI Tool for Fast Video Translation and Localization (videotranslator.blipcut.com)

If Indian goods cannot go to the US, they can head to Russia (timesofindia.indiatimes.com)

From Sound to Meaning: A Deep Meaning Comprehension by DeepSeek (medium.com)

'The Lord Is Counting on Me to Stand on the Side of Israel' (zeteo.com)

Postgres in 2025: No managed service required? (docs.codefloe.com)

Show HN: Code – coding CLI with browser control and diffs (github.com)

Diploma of Nursing (menzies.vic.edu.au)

Virtuous Machines: Towards Artificial General Science (arxiv.org)

Understanding Go Error Types: Pointer vs. Value (blog.fillmore-labs.com)

Hkust Launches the Largest AI Education Sandbox Game "Aivilization" (aivilization.onrender.com)

How to build a world model? Introduction to Laplace Neuron (abibulic.github.io)

Why Detection Of Fake Content Isn't the Answer Anymore (inreality.io)

Show HN: Tempmail3 – Instant disposable email service, zero signup, open-source (tempmail3.com)

Show HN: VoxDemo – Create demo videos with AI voiceovers (voxdemo.com)

Building Ultra Cheap Energy Storage for Solar PV (austinvernon.substack.com)

Why is my device a touchpad and a mouse and a keyboard? (who-t.blogspot.com)

Accessibility Conformance Testing (Act) Rules Format 1.1 (w3.org)

Ask HN: Would you use an agent that migrates your stack (with benchmarks)? (charter-nlpt.vercel.app)

Show HN: Spectre, a coding agent for llama.cpp servers (github.com)

Monad Annoyance (macwright.com)

In AI push, China holds the first sports event for humanoid robots (nbcnews.com)

Qwen-Image-Edit: Image Editing with Higher Quality and Efficiency (qwenlm.github.io)

Ask HN: Imagine coding LLM's 1M times faster; what uses might there be?

Comments (1)