Mar 12, 2026 · 43m · catalyst
AI scaling pathways: on grid, on edge, off grid, off planet
0:00 / 0:00
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this episode of Catalyst, host Shayle Kann and guest Jake Elder evaluate four pathways for scaling AI compute infrastructure—grid-connected hyperscale, edge computing, islanded off-grid microgrids, and orbital space facilities—assessing their technical bottlenecks, unit economics, and 10-year market viability.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Shayle holds 48% of the talking time here. How this is scored →
Shayle as informed peer 6.7 Guest teaching 3.8 Guest disagreement 1.3 Shayle pushing back 2.3
speaking balance: gold is Shayle, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 36:25 Questioning terrestrial public support for gas buildout
Elder challenges the premise that land-based scaling will outpace space by pressing Kann on whether society will actually tolerate hundreds of gigawatts of new terrestrial gas infrastructure.
Hardest push from Shayle ▶ 37:05 Challenging launch cadence versus clean terrestrial techKann pushes back on the orbital narrative by asking whether scaling terrestrial solar, batteries, geothermal, or nuclear is genuinely crazier than five fully reusable Starship launches every day.
Biggest teaching moment ▶ 31:48 Orbital thermodynamics and debris collision mathematicsElder walks Kann through the exact thermodynamic realities of heat dissipation in a vacuum and cites collision data showing a 4-square-kilometer orbital data center would be struck by debris hourly.
Shayle holds their own ▶ 7:20 Interregional transmission timeline reality checkKann demonstrates deep grid expertise by correcting common industry conflation, noting that while equipment orders take 5 to 7 years, interregional transmission line timelines in the US are essentially infinite.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Shayle as informed peer | Guest teaching | Guest disagreement | Shayle pushing back | Why |
|---|---|---|---|---|---|---|
| Constraints on Grid-Connected Hyperscale Data Centers | 7 | 3 | 1 | 2 | Kann and Elder discuss the primary constraints of grid-tied hyperscale data centers. Kann demonstrates strong sector familiarity by differentiating generation supply chain bottlenecks from the near-infinite timeline of interstate transmission buildout. | |
| Evaluating the Realities of Edge Computing for AI | 7 | 4 | 1 | 2 | Elder outlines three definitions of edge compute, and both agree latency is mostly a red herring for AI inference. Kann adds depth regarding levelized cost of compute and the organizational friction of managing hundreds of subscale sites. | |
| Opportunities and Hurdles of Off-Grid Power Delivery | 6 | 5 | 1 | 2 | Elder explains the technical reason off-grid data centers struggle: the grid acts as a massive shock absorber for power quality, inertia, and black starts. Kann contributes insights into cloud SLA histories and trade-offs of sub-five-nines reliability. | |
| Mid-Episode Sponsor Break: Fuel Cells, Corporate Energy, and Virtual Power Plants | 6 | 3 | 1 | 2 | After the mid-episode sponsor block, Kann and Elder evaluate geographic siting shifts and supply chain equipment constraints for off-grid architectures, noting modular gas reciprocating engines as a tactical workaround. | |
| The Physics, Economics, and Feasibility of Orbital Data Centers | 7 | 6 | 2 | 3 | Both analyze the viability of orbital data centers, debunking the near-term economic claims of proponents. Elder brings striking calculations regarding radiator surface areas and orbital debris strike frequencies, while Kann questions the Starship cadence required relative to terrestrial clean tech. | |
| Ten-Year Compute Allocation Forecast and Episode Conclusion | 7 | 2 | 2 | 3 | Elder shares a 10-year percentage forecast across compute modalities. Kann pushes back mildly against the edge computing allocation, arguing that off-grid terrestrial will take more market share given the lack of economic justification for edge. |
Statements from this episode (14)
Kann: On-grid, off-grid, edge, and space data centers will probably be built
“Massive data centers on the grid, massive data centers off-grid, small data centers on the edge, huge data center clusters in space. Each of these might get built. Actually, each of them probably will get built”
Elder: Interconnecting gigawatt-scale data centers takes 5 to 7 years in many markets
“How, how long can we, how much time will it take to build out the transmission capacity necessary to interconnect these mega sites, gigawatt scale sites to new power supply, ideally, you know, carbon free power supply. And in many markets, right, that's runnin…”
Kann: Lead times for gas turbines, transformers, and switchgear are 3 to 7 years
“The five to seven years is the timeline to get new gas turbines, if you're ordering them, and it's close to the timeline to get new transformers and other, and switchgear and stuff like that. Like, we're all in the, like, three to five or maybe seven year time…”
Kann: Almost none of the 50 GW data center pipeline is off-grid
“There's like, 50 gigawatts of behind-the-meter generation and development at data centers, right? And they, and some people have interpreted that to be, oh, 50 gigawatts of off-grid data centers getting built. It's actually close to zero of those that are true…”
Kann: Edge computing latency advantage applies to very few AI applications
“Not for zero applications, but for very few does it seem that you need such low latency that edge has a big benefit over, you know, the sort of, like, regional hyperscale model that we have today.”
Kann: 300 MW data centers have lower cost per FLOP than edge
“My guess is, you know, your 300 megawatt data center on a fully loaded levelized cost of flop is just gonna be cheaper.”
Elder: Stripe and Scale study found 1TW+ off-grid potential in Southwest
“There was this foundational study that came out about two years ago that was co-authored by Stripe and Paces and Scale Microgrids, and they found over a terawatt of opportunity in the American Southwest alone, with high levels of renewable development being ab…”
Elder: Early off-grid data center projects fail to maintain 90% uptime
“When we, we've heard some of the early data from some of the off-grid projects that have been built so far, the ANIC data suggests they're not able to stay above even 90% uptime yet.”
Kann: Grid constraints make sub-uptime off-grid data centers inevitable
“In a world where we're so constrained, on the grid side, it seems inevitable to be that that is going to happen to some degree, and that the engineering challenge is going to get at least partially solved.”
Kann: Data center development is rapidly expanding into West Texas
“Historically, you know, there were these tier one markets, like Northern Virginia, or Chicago, or Phoenix, or whatever, Atlanta, and they were where 90% of the demand for new data centers was going to be, and there's still that. But it is broadening out quickl…”
Elder: Off-grid setups partially bypass generation bottlenecks with modular equipment
“The off-grid option in some ways maybe shortcuts the actual supply chain bottlenecks on the generation equipment side, at least to some extent relative to the other options. But I agree with you, there's still a bunch of other pieces of equipment, transformers…”
Elder: ISS rejects under 100 kW heat with soccer-field-sized radiator
“I think the whole International Space Station, for example, rejects less than a hundred kilowatts of heat in total, and they have a radiator the size of a soccer field, right?”