Hacker News: Escapade5160

New comment by Escapade5160 in "Our position on open-weights models"

Escapade5160 — Mon, 27 Jul 2026 22:27:23 +0000

Perhaps they saw it crystal clear.

New comment by Escapade5160 in "Show HN: Abralo – Free, easy way to run several Claude Code agents in one window"

Escapade5160 — Sat, 11 Jul 2026 04:15:08 +0000

Cmux.... Or Orca.

New comment by Escapade5160 in "Every new car sold in the European Union must include a driver monitoring camera"

Escapade5160 — Wed, 08 Jul 2026 07:05:37 +0000

Fuck this

New comment by Escapade5160 in "GLM 5.2 and the coming AI margin collapse"

Escapade5160 — Tue, 07 Jul 2026 01:38:01 +0000

Most large orgs do not need to train end users. They just need to add glm-5.2 to their router and their in house harness will pick it up. Then slowly limit usage on anthropic models and people will swap willingly. It's a simple /model command in every harness.

New comment by Escapade5160 in "Claude Sonnet 5"

Escapade5160 — Wed, 01 Jul 2026 08:59:38 +0000

At that price you should just use glm-5.2. You get an Opus class model for 1/3 the cost.

New comment by Escapade5160 in "Micro-Agent: Beat Frontier Models with Collaboration Inside Model API"

Escapade5160 — Mon, 29 Jun 2026 21:43:48 +0000

So be it.

New comment by Escapade5160 in "Herdr: Agent multiplexer that lives in your terminal"

Escapade5160 — Mon, 29 Jun 2026 10:11:26 +0000

I'm a power user and it's not an issue for me. I just use Kitty with multiple windows and panes. I can jump between using hotkeys with no problem.

New comment by Escapade5160 in "HPV vaccine cuts risk of dying from cervical cancer before 30"

Escapade5160 — Mon, 22 Jun 2026 03:41:12 +0000

Can we stop calling vaccines jabs?

New comment by Escapade5160 in "I hacked into the worst e-bike and fixed it [video]"

Escapade5160 — Wed, 17 Jun 2026 01:48:27 +0000

It's in YouTube's best interest to only show users content they're interested in. Replace the word algorithm with users and you'll have a more accurate representation of how YouTube actually works. The reason those videos didn't get the love they deserved is because they're niche content, the 10 minute review videos appeal to a wider audience and therefore gain more traction.

New comment by Escapade5160 in "Grit: Rewriting Git in Rust with agents"

Escapade5160 — Wed, 10 Jun 2026 22:04:11 +0000

Particularly because LLM generated code is not licensable in any way. If you wrote it with an LLM you cannot own it.

New comment by Escapade5160 in "Claude Fable 5"

Escapade5160 — Wed, 10 Jun 2026 00:53:15 +0000

They gave everyone double usage to try it.

New comment by Escapade5160 in "Claude Fable 5"

Escapade5160 — Wed, 10 Jun 2026 00:52:34 +0000

It's crazy to release a model that just swaps you to another model when you ask it hard questions. Fable changes to Opus 4.8 when you talk about cybersecurity, biology, and a couple other categories. You still pay Fable input token cost though. Frontier models are stalling, this is anthropic trying to hype the market up. Now they're talking about stopping frontier model research. It's kind of strange how the moment they become the highest valued AI company, all of a sudden they're talking about everyone stopping frontier model development for "safety". They're just as corrupt as the rest.

New comment by Escapade5160 in "Ask HN: What are tools you have made for yourself since the advent of AI?"

Escapade5160 — Tue, 09 Jun 2026 09:19:24 +0000

is:unread -is:starred <-- go through your inbox and star what you want to keep, this filter will help you delete everything else unread. Add something like older_than:1y to also prune your Gmail from time to time.

New comment by Escapade5160 in "Microsoft Doubles Down on Controversial Quantum Computing Claims"

Escapade5160 — Wed, 03 Jun 2026 07:41:30 +0000

I feel they should call it Q.

New comment by Escapade5160 in "Expanding Project Glasswing"

Escapade5160 — Wed, 03 Jun 2026 00:30:52 +0000

This is the same gripe I have over any LLM vulnerability tooling. 95% of what gets flagged is something that if taken by itself could be a vulnerability. However, the path to execute that specific vuln, in that specific function, is impossible in that particular code base and it just makes noise.

New comment by Escapade5160 in "Malicious npm packages detected across Red Hat Cloud Services"

Escapade5160 — Mon, 01 Jun 2026 15:43:42 +0000

Can someone give a tldr on why this happens so much with npm ? I can't recall seeing this with any other package manager. Is npm just the default used these days and therefore sees this more often?

New comment by Escapade5160 in "Erin Brockovich made a map to track data centers around the country"

Escapade5160 — Wed, 27 May 2026 03:26:04 +0000

Ben Jordan did a fantastic piece on how harmful data centers are to the people living near them.

New comment by Escapade5160 in "Memory has grown to nearly two-thirds of AI chip component costs"

Escapade5160 — Mon, 25 May 2026 01:08:03 +0000

And four-fiths the cost of a consumer PC build.

New comment by Escapade5160 in "Show HN: Forge – Guardrails take an 8B model from 53% to 99% on agentic tasks"

Escapade5160 — Tue, 19 May 2026 22:09:42 +0000

I've been saying for a while that given a proper harness, small local models can perform incredibly well. When you have a system that can try everything, it will eventually get it right as long as you can prevent it from getting it wrong in the meantime.

New comment by Escapade5160 in "Show HN: Semble – Code search for agents that uses 98% fewer tokens than grep"

Escapade5160 — Sun, 17 May 2026 22:45:16 +0000

Setup hooks. Hooks are how your harness forces compliance with your own rules.