Chanos Labels Orbital AI Datacenters 'Snake Oil' with $600K/MW Costs, NRG Energy Set to Gain
Short seller James Chanos warns that orbital AI datacenters would incur annual electricity costs of $600,000 per megawatt—around 5–7% of revenues—and face higher launch, radiation, insurance and redundancy expenses. Investor Michael Burry proposes a $1 trillion nuclear power and grid expansion plan that could boost capacity for AI operations, positioning independent producers like NRG Energy to capture rising demand.
1. Chanos Criticizes Orbital AI Datacenters
James Chanos argues that putting datacenters into orbit would dramatically increase costs beyond terrestrial operations. He highlights annual electricity expenses of roughly $600,000 per megawatt—equal to 5–7% of AI deal revenues—and cautions that launch, radiation shielding, insurance and redundancy requirements would further erode any savings.
2. Burry's Nuclear Power Proposal
Investor Michael Burry recommends a $1 trillion investment in nuclear power and grid infrastructure to meet surging AI compute demand. His plan targets rapid licensing and construction of advanced reactors to deliver reliable baseload power at larger scale than current deployments.
3. Impact on NRG Energy
Independent power producers such as NRG Energy stand to benefit if AI datacenters drive a sustained rise in electricity consumption. Expanded nuclear capacity and enhanced grid partnerships could open new revenue streams for companies supplying large-scale power solutions.