A prediction market on Manifold Markets asks 'Did DeepSeek lie about the GPU compute budget they used in the training...
A prediction market on Manifold Markets asks 'Did DeepSeek lie about the GPU compute budget they used in the training of v3?' As of the latest source (2026-05-28), the consensus probability of a lie is only 4.01%, indicating that the market largely believes DeepSeek's disclosed GPU compute figure. However, the mere existence of this question highlights ongoing skepticism around AI lab compute claims, particularly for DeepSeek's v3 model. While the low probability suggests no material change, the debate could resurface if new evidence emerges. Single source, low confidence. Source
signal brief
A prediction market on Manifold Markets asks 'Did DeepSeek lie about the GPU compute budget they used in the training of v3?' As of the latest source (2026-05-28), the consensus probability of a lie is only 4.01%, indicating that the market largely believes DeepSeek's disclosed GPU compute figure. However, the mere existence of this question highlights ongoing skepticism around AI lab compute claims, particularly for DeepSeek's v3 model. While the low probability suggests no material change, the debate could resurface if new evidence emerges. Single source, low confidence. Source
evidence
Decision support, not stock advice. This signal is research with cited evidence — not a recommendation to buy, sell, or hold any security.