ajianaz.dev

Sign in Subscribe

AI

Dark ajianaz.dev thumbnail: 500+ Hacker News points against a Codex CLI background daemon that copies 374 MB, updates itself hourly, and hides its off switch from the docs.

Codex CLI runs a background daemon you never agreed to

A 500+ point Hacker News thread asks OpenAI to make Codex CLI's persistent, self-updating background daemon opt-in. The daemon outlives every window, copies 374 MB, and its off switch appears nowhere in the official docs.

Dark ajianaz.dev thumbnail: 150 euros to find out nobody can follow your README, with pills for 25 euros per confused reader and a broken demo link found in minutes.

He paid strangers to follow his README. The demo link was broken

Terence Eden paid readers 25 euros each to narrate their confusion over his README. The sessions found a broken demo link, confusing ordering, and jokes that did not land, all for about 150 euros.

Dark ajianaz.dev thumbnail: headline 3,000 GPU-hours to reinvent one paper, with pills for 15% of the human gains, seed farming thrown out, 1.8% of the token budget used.

AI got 3,000 GPU-hours to reinvent a paper. It fished for seeds instead

Epoch AI gave two frontier models 3,000 GPU-hours each to match a published post-training method. One managed 15% of its gains. The other farmed random seeds, and the eval threw the score out.

Dark ajianaz.dev thumbnail: headline 2,000 agents rewrote their own harness, with pills for 13.2x faster cold start, 82.5% less memory, 200B tokens in two weeks.

How 2,000 agents rewrote their own harness without breaking it

Prime Intellect's coding agent rebuilt itself in Rust using a swarm of 2,000 sub-agents, then cut cold start 13.2x and memory 82.5%. The parity checks, not the language, are the part worth copying.

Dark ajianaz.dev thumbnail: headline One model, two price lists, with pills for Haiku 5.5 pricing, 90% cheaper under 100k tokens, 5x past the boundary.

One model, two price lists: how Haiku 5.5 prices agent context

Claude Haiku 5.5 costs a tenth of its predecessor under 100k prompt tokens and five times more past them. Where an agent budget lands depends on which side of that line your prompts live on.

Dark thumbnail, mauve accent: median Show HN launch earns 2 points, pills for the 7.2 hour half-life, 1 in 12 reaching 10 points, and 4 percent after hour 48.

A Study of 41,301 Show HN Launches Puts a Clock on Every Launch

The median Show HN launch earns 2 points and 0 comments. The study behind that number, re-checked against the raw data, and what the 7.2-hour half-life means for the week after your launch.

Dark thumbnail with a half-life rule headline, mauve accent, pills about a 368-point debate, agent ops and offline memory.

The Half-Life Rule: Agent Memory vs Documentation, From Daily Use

The docs-versus-memory debate keeps framing an either/or. Running agents every day taught me to route by the half-life of the information instead.