building on [[day1 david(entfinance)]], [[day1.5]] # 1. why tech adoption ## stylized fact1 difference between TPF surprisingly large TPF variance (differential across) reduced by competition ## stylized fact2 spread between frontier and laggard firms widening over time manufacturing and business service spead in time and spread opening over time ## stylized fact 3 growing ikmportance within and btw firms (wage) within firms that pays a lot there are cluster ## stylized fact 4 growing wage polarization across workers, largely driven btw firm dynamic --- firm dynamics slowing (right tails peaked) five forces (business school), most of the at a point in time is positioning evolution of manufacturing - changes of structure (managerial subsidiaries) world of changing technology adoption is one of the major factor behind this - productivity j curve -> underestimated productiviey --> overestimated productivity J curve: skewed msr of productivity growht after a major tech is introduced dip (learning; adoption) in productivity then exponential slope is hetereogenous across firms sorro paradox ( you see computer everywhere except productivity statics - ) measurement of intagngialb e growth tech -> level of technology dynamicss ---- # 2. basic tech foundation Y = (integ_0,1 alpha(i) y(i) rho di) 1/rho tasks ordered by comparative advantage cost minimization : firms chose cheaper input for each task i w/r = gamma_L(l)/ gamma_K(l) task assignment threshold under which to human and above which to capital - labor performr tasks i - capital performr tasks i effects of automation in equilibrium no creation of new tasks which type of tasks are more exposed to substitution automation vs task creation reinstatement effect: new tasks restore labor demand engogenous tech frontier: N = N(t) expands over time as a function of wages, feedback effects can amplify or dampen initial impacts reallocate btw capital and humans new occupations still very little explicit consideration for firm heterogeneity beyond task-based models: knowledge hierarchices - hierarchical structures with managers, knowledge is specialized and complementary across levels not efficient to let everyone learn complex skills tacit vs structured learning is living in individual jorge's q: centralized (temp. is controlled central), decentralized price (value local kwldg) david's q: asymmetric info. inference about my type, i don't wanna look bad three learning= 🧪experimenting, ⚖️scaling, 🔐codifying knowledge # 3. tech adoption across firms # 4. tech adoption within firms # 5. deep dive: ai # 6. conclusion