alphabetting

2026-02-19 7:00

Commented: "Gemini 3.1 Pro"

the agentic benchmarks for 3.1 indicate Gemini has caught up. the gains are big from 3.0 to 3.1.

For example the APEX-Agents benchmark for long time horizon investment banking, consulting and legal work:

1. Gemini 3.1 Pro - 33.2% 2. Opus 4.6 - 29.8% 3. GPT 5.2 Codex - 27.6% 4. Gemini Flash 3.0 - 24.0% 5. GPT 5.2 - 23.0% 6. Gemini 3.0 Pro - 18.0%

2026-01-06 1:37

Submitted: "Debunking the AI food delivery hoax that fooled Reddit"

12 points1 commentswww.platformer.news

A “whistleblower” tried to corroborate his viral post with AI-generated evidence. This is how I caught him. PLUS: Grok's image-generation crisis, and the rapture over Claude Opus 4.5

2025-12-19 6:59

Submitted: "Performance Hints"

59 points1 commentsabseil.io

An open-source collection of core C++ library code

2025-12-16 12:48

Submitted: "Gemini 3 Pro vs. 2.5 Pro in Pokemon Crystal"

315 points92 commentsblog.jcz.dev

From 2.5 Pro's lighthouse loop to 3 Pro's "Zombie Phoenix" strategy against Red: a full breakdown of why Gemini 3 Pro is a superior long-horizon agent.

2025-10-16 4:49

Submitted: "VLLM TPU: A New Unified Back End Supporting PyTorch and Jax on TPU"

1 points0 commentsblog.vllm.ai

vLLM is a fast and easy-to-use library for LLM inference and serving.

Hacker News

alphabetting

5790

2021-04-20

Recent Activity

Commented: "Gemini 3.1 Pro"

Submitted: "Debunking the AI food delivery hoax that fooled Reddit"

Submitted: "Performance Hints"

Submitted: "Gemini 3 Pro vs. 2.5 Pro in Pokemon Crystal"

Submitted: "VLLM TPU: A New Unified Back End Supporting PyTorch and Jax on TPU"

HackerNews