MikeTrendsTrends right now

Yhn BusinessEconomy first seen 1 d ago, last 1 d ago, peak #11

Why didn't we get GPT-2 in 2005?

A new essay asks why large language models like GPT-2 weren't built nearly two decades earlier, given that many core ingredients were available by the mid-2000s. The piece explores whether the holdup was compute, data, algorithms, or simply a lack of imagination about scaling, and readers are debating which factor mattered most.

Why now: It challenges assumptions about AI progress timing, sparking debate over what actually delayed the deep learning breakthrough.

GPT-2OpenAIDynomight

Open on hn →

Evidence

API: https://socialmediatrends-api.osmike.com/v1/trends/899472