Infrastructure · Trend

Open-weight models

Open-weight releases trail the closed frontier by months, reshaping access and sovereignty.

Trend strength 8/10

Momentum +2/q

Confidence high

Status rising

Forecast horizon

Open weights anchor national stacks; the debate moves to release norms and misuse liability.

Connections

Connections · 3

How this node ties into the rest of the map, and the evidence behind each link.

Open weights are the substrate for national stacks.

Open releases seed efficient downstream models.

Open-weight bases lower the cost of a domestic stack.

Signal sources

Dated facts from primary sources in this direction.

Training compute for frontier AI models grew ~4–5× per year from 2010 to 2024 — a trend Epoch projects to continue toward 2030.

Global data-centre electricity use is projected to roughly double from ~485 TWh (2025) to ~950 TWh by 2030 — about 3% of world electricity.

The price to query a GPT-3.5-level model fell from $20.00 to $0.07 per million tokens between Nov 2022 and Oct 2024 — a 280× drop in ~18 months.

DeepSeek disclosed its open-weight R1 reasoning model's final training run cost just $294,000; R1 was released under open weights on 20 Jan 2025.