AI
Essay "Tokens Too Cheap to Meter" sparks debate over AI cost collapse
Commenters question whether price declines are real or subsidized
3
Sep 22 6:44 PM
· 1d ago · 3 comments · 1 source · development 3 of 5
Several Lobsters commenters push back on the essay's premise, asking whether reported per-token price drops reflect true costs or are being subsidized by AI labs, and whether the industry's own claims can be trusted.
“How do we know that the price per token is going down? Are we just taking the AI labs word for this?”
roflmaoqwerty jyn.dev (blog author) Author of the essay roflmaoqwerty Lobsters commenter gulbanana Lobsters commenter timthelion Lobsters commenter
The whole story the bright band is this development · numbered dots are the others · click one to jump
Peak 6 pieces in one hour at Yesterday, 9 AM; 49 pieces over 39 hours (1 article · 6 posts · 42 comments)
Sep 22, 11 AM — 1 piece · 1 post — Hacker News 1 Sep 22, 12 PM — 1 piece · 1 post — Lobsters 1 Sep 22, 1 PM — quiet Sep 22, 2 PM — quiet Sep 22, 3 PM — quiet Sep 22, 4 PM — quiet Sep 22, 5 PM — quiet Sep 22, 6 PM — 2 pieces · 2 comments — Lobsters 2 Sep 22, 7 PM — 1 piece · 1 comment — Lobsters 1 Sep 22, 8 PM — 2 pieces · 1 post · 1 comment — Hacker News 1, Lobsters 1 Sep 22, 9 PM — quiet Sep 22, 10 PM — 1 piece · 1 comment — Lobsters 1 Sep 22, 11 PM — quiet Yesterday, 12 AM — quiet Yesterday, 1 AM — 1 piece · 1 comment — Lobsters 1 Yesterday, 2 AM — 3 pieces · 3 comments — Lobsters 3 Yesterday, 3 AM — quiet Yesterday, 4 AM — 1 piece · 1 comment — Lobsters 1 Yesterday, 5 AM — 2 pieces · 1 article · 1 post — Hacker News 1, Newswires 1 Yesterday, 6 AM — 1 piece · 1 comment — Lobsters 1 Yesterday, 7 AM — quiet Yesterday, 8 AM — 1 piece · 1 post — Mastodon 1 Yesterday, 9 AM — 6 pieces · 6 comments — Hacker News 6 Yesterday, 10 AM — 5 pieces · 5 comments — Hacker News 4, Lobsters 1 Yesterday, 11 AM — 3 pieces · 3 comments — Hacker News 2, Lobsters 1 Yesterday, 12 PM — 2 pieces · 2 comments — Hacker News 2 Yesterday, 1 PM — 1 piece · 1 comment — Hacker News 1 Yesterday, 2 PM — 2 pieces · 2 comments — Hacker News 2 Yesterday, 3 PM — 5 pieces · 5 comments — Hacker News 5 Yesterday, 4 PM — quiet Yesterday, 5 PM — 3 pieces · 1 post · 2 comments — Hacker News 1, Lobsters 1, Mastodon 1 Yesterday, 6 PM — 2 pieces · 2 comments — Hacker News 2 Yesterday, 7 PM — 1 piece · 1 comment — Hacker News 1 Yesterday, 8 PM — quiet Yesterday, 9 PM — quiet Yesterday, 10 PM — 1 piece · 1 comment — Hacker News 1 Yesterday, 11 PM — 1 piece · 1 comment — Hacker News 1 Today, 12 AM — quiet Today, 1 AM — quiet 1–2 3 4 5
4 PM yesterday 8 AM 4 PM now · 2:03 AM ET
What people said 3 voices · verbatim
T
LLMs are evaluated on benchmarks, which show this large drop in cost per intelligence: https://sanand0.github.io/llmpricing/intelligence.html This aligns with my experiences too.
R
How do we know that the price per token is going down? Are we just taking the AI labs word for this?
M
> once models are cheaper than tools Why wouldn’t a tool’s execution get cheaper on a similar curve to inference costs?
All 5 developments of Essay "Tokens Too Cheap to Meter" sparks debate over AI… →