AI performance costs falling faster than past tech

Epoch AI estimates AI inference prices at fixed performance fall about 13x per year; MIT puts annual algorithmic progress at about 3x after stripping out hardware gains and competition. Yet today's best models aren't necessarily cheaper—reasoning models can cost more due to higher compute per task. Real-world selection also depends on quality, speed, and error rate.

• Open-weight GLM 5.3 builds Firefox exploit chains autonomously in under 48 hours2026-10-30
• MIT Professor Says AI Is Taking Over Your Brain2026-10-29
• Harvard's Casper to host AMA on AI governance2026-10-28
• MIT study finds AI agents play spontaneously on virtual island2026-10-26
• Cognitive Surrender: AI's Threat to Democracy2026-10-24