In prior innovation booms, new technologies replaced some old jobs, but tended to create new industries that were capital hungry and labor intensive, thus shifting labor demand into other physically expansive areas. Millions of humans were needed to build new industrial bases. The late 1800s were like this when mechanized agriculture halved labor demand, but workers shifted into mining, manufacturing, construction, railroads, etc. The dotcom boom was a bit different but mostly enhanced existing work via digitization- accountants used spreadsheets, retailers adopted e-commerce, designers used software, logistics were digitized and sped up, etc. In both these waves, labor demand *shifted* but humans remained central to the productive workflow. AI/automation feels different because it threatens to obviate not only repetitive clerical and some physical work, but also the *general purpose COGNITIVE layer of white collar economies*. Advanced economies are overwhelmingly composed of white collar cognitive-layer employment. This is wholesale different from prior booms. The key question now is: will AI augment existing workers or substitute for them? Even when prior innovation booms replaced some jobs, they tended to be sector specific (like farm workers replaced by harvesters and binders). These shifts tended to create brand new *physical* labor bottlenecks that created new demand: railroads needed track layers, automobiles needed mechanics and assemblymen, suburbanization needed construction workers, electrification needed grid buildout, etc. These were all new physically demanding positions where human labor remained key. AI threatens to slash labor demand horizontally across most or all sectors instead of specific ones, and its scaling laws are far less physically intensive once models are trained. Its scaling laws involve compute, semiconductors, datacenters, etc. Capital, not labor. This potentially raises capital demands with far weaker labor absorption. That is the substantive threat right now, that we obviate the labor side of output even as (potentially) aggregate productivity expands. We haven't faced quite this sort of transition before. We clearly seem to me, in these early stages, to be contending more with the substitutive effects than the augmentative effects of this transition. And that is why people are feeling a sort of existential dread about their career prospects in a way no one has really dealt with before in such broad scope.(𝐓𝐗𝐌𝐂)