Empirical Study Finds Pearl's Proof-of-Useful-Work Blockchain Produces Zero Useful AI Computation Despite Massive GPU Network
By
[Submitted on 3 Jun 2026]
Toasted just enough. A reliable bake, gently seasoned.
Summary
This empirical study examines Pearl's Proof-of-Useful-Work (PoUW) blockchain protocol, which claims to simultaneously secure its network and perform AI inference. The researchers found that despite Pearl's 24 EH/s network (approximately 320,000 GPU-equivalents consuming 112 MW), it produces zero useful AI computation. The study reveals that the dominant mining software contains no inference code, the verification protocol accepts random matrices, statistical checks are easily defeated, mining is unprofitable across all GPU tiers, and the computation offers no vendor lock-in. Additionally, the mining software's release caused budget GPU rental prices to rise 38% and utilization to surge from 57% to 94%, displacing legitimate research workloads.
Key quotes
· 5 pulledPearl's 24 EH/s network -- representing approximately 320,000 GPU-equivalents consuming an estimated 112 MW -- produces zero useful AI computation.
Budget GPU rental prices rose 38% and utilization surged from 57% to 94% following the mining software's public release, displacing legitimate research workloads.
The dominant mining software contains no inference code; the verification protocol accepts random matrices by design.
Mining is unprofitable at current PRL prices ($0.21) across all GPU tiers (-54% to -72% ROI).
These findings quantify the verifiability-usefulness tension identified theoretically by Leinweber et al., providing concrete measurements of its magnitude and economic consequences in a deployed system.
You might also wanna read
General Compute Launches ASIC-Based Inference Cloud for Faster AI Agent Performance
General Compute is an inference cloud built on ASICs (purpose-built alternatives to Nvidia GPUs) designed specifically for AI inference, not
Product Hunt·1mo ago
AI compute crisis: Enterprise GPU hoarding leaves startups stranded as OpenAI faces $14B loss projections
The article examines the structural dysfunction in the AI industry where enterprise GPUs sit idle (up to 95%) while AI startups struggle to
techradar.com·2d ago
The Hidden Cost of Compute: Why AI Infrastructure Investment Is Misguided
The article argues that the AI industry is over-investing in massive compute infrastructure (data centers, GPUs) while neglecting critical a
South Korea's GPU Race: Why AI Competitiveness Depends on Utilization, Not Just Hardware
South Korea is aggressively expanding its AI GPU infrastructure as a national strategic priority, but the article raises critical questions
koreatechdesk.com·1d ago
Telegram's Cocoon: Confidential Compute Network for AI and Cryptocurrency Mining
Cocoon by Telegram is a confidential compute open network that enables GPU owners to mine TON cryptocurrency by providing computing power to
Product Hunt·6mo ago
Zibra AI Launches GPU-Native Data Orchestration Platform for Spatial AI Training
Zibra AI introduces a GPU-native data orchestration platform designed to solve I/O bottlenecks in spatial and physical AI training. The plat
Product Hunt·1mo ago