Why Human Technology Is Slower Than You Think | Brian Greene

Progress is fake. Physicist Brian Greene dismantles the narrative of exponential technological acceleration, revealing that fundamental innovation has slowed dramatically despite subjective impression of rapid change created by information technology advances. Discover why computing progress is narrower than believed: Moore's Law applies only to transistor density not useful capability (doubling transistors ≠ doubling useful computation due to parallelism/bandwidth/power limits), Moore's Law now ending at 5nm approaching atomic scales where quantum tunneling prevents further miniaturization (industry shifted to 3D stacking/new materials providing continued improvement but not exponential rates), fundamental computing capabilities established in 1940s-1970s unchanged (modern computers just run same programs faster, no new categories of computation enabled), software progress even slower than hardware (most algorithms/techniques from 1950s-1980s, modern improvements mostly usability/scale not fundamental capability). Learn why AI predictions consistently fail: field began 1950s predicting human-level AI within generation (Herbert Simon 1965 predicted "within twenty years machines capable of any work man can do"), these predictions failed completely leading to AI winters when funding collapsed, pattern repeated every decade with new techniques hyped then disappointing (expert systems 1970s-1980s, neural networks 1990s, deep learning 2010s all promised AGI within decades), current large language models sophisticated pattern matchers not approaching general intelligence (make logical/mathematical errors revealing lack of understanding, can't learn from small examples or generalize robustly, no persistent memory/goals/agency beyond responding to prompts), 70 years into AI research still no artificial general intelligence despite consistent predictions it's 10-20 years away.