New Tools and Techniques
Shaun Sparks is Pennsylvania Public Utility Commission Legal Counsel and Chief of Staff, Office of Commissioner Ralph Yanora.
Suppose the people building the artificial intelligence (AI) boom are correct. Sam Altman of OpenAI describes a future where intelligence is metered like electricity, with the cost of intelligence converging toward the cost of the electricity behind it.
Dario Amodei of Anthropic holds that raw compute investment scales smoothly into artificial general intelligence (AGI). Nvidia’s Jensen Huang opines that reasoning-capable systems require roughly 100 times the computation of the prior generation.
If that is the business plan, the cost of the grid will become the defining cost constraint of intelligence products. Minimizing that cost will become a survival condition of intelligence providers. They will seek to shape cost recovery, rate design, and utility investment plans to deliver electricity on their terms. Expect them to bring capital and regulatory resources to do it.
If AI becomes the economic transformer its builders predict, regulators are obligated to accommodate public demand for the utility services AI will require as a critical input. The obligation to accommodate public demand and convenience is not optional and does not pause regulatory deliberation, but that obligation does follow the public interest – not the preferred terms of AI operators.