David Goodnight Austin Looks Past the Megawatts: Why AI’s Real Constraint May Be Delivery

 How much electricity AI requires is an important question, but David Goodnight Austin’s broader infrastructure lens suggests another question deserves attention: can the physical system deliver that electricity when and where it is needed? A power forecast alone does not build a transmission line, manufacture a transformer, connect a substation, or complete a data center. The difference between theoretical capacity and usable capacity can become an economic constraint. This is particularly relevant as artificial intelligence pushes companies toward larger computing facilities and faster infrastructure schedules. David Goodnight’s recent writing on AI and energy examines the relationship between technology and the physical systems supporting it. That theme can be extended into a practical lesson for investors and developers: infrastructure timing can influence technology timing. Projects with capital but no equipment, or equipment but no grid connection, remain unfinished. David Goodnight Austin brings attention to this less visible part of technological growth.