I’m often asked about the themes we invest in, and one I always mention, because people overlook it, is space. We've done a lot of podcasts on space and I believe we will eventually have a lot of what’s done on earth in orbit. Perhaps the most interesting aspect is power generation. We have unlimited power from the sun, but we can’t always see the sun, unless we are in space……

AI isn’t going to “crash” because the models get worse—it stalls when the inputs choke: power, transformers, interconnects, cooling, and permitting. That’s the quiet bottleneck hiding behind every glossy “AI will change everything” slide deck. And it’s why the most underappreciated AI trade isn’t chips… it’s electricity and the hardware that moves it.

Now take that bottleneck to its logical extreme: if Earth’s grid can’t ramp fast enough, you don’t just optimize data centers—you change the physics. You push more inference to the edge (including orbit), run it on sunlight, and beam down answers instead of raw data. Space-based solar and wireless power are still early, but proof-of-concept work is real; scaling is the whole fight. The investing edge is spotting where capex goes first (edge inference + space infrastructure) versus what stays a long-dated sci‑fi call option (full gigawatt beamed power).

The signal: power is becoming the binding constraint on AI, not chips. Data centers and AI are projected to drive a big step-up in electricity demand this decade, which is exactly why “grid tech” has become a real market theme. We talked about this yesterday.

The spicy extrapolation: if the grid (permitting, transformers, interconnect queues) can’t keep up, the market starts hunting for non‑grid solutions. That’s where “compute in orbit” becomes a thought experiment worth taking seriously—because the economics of AI are unusually high revenue per kW, and solar in space is always “on” relative to many terrestrial constraints.

But: scaling from “cool demo” to “meaningful global power/compute” is brutally hard. We have seen credible proof‑of‑concept work: Caltech’s space solar project demonstrated wireless power transfer in space as a step toward space solar power architectures, and DARPA recently demonstrated long-distance wireless power beaming on Earth (laser) as an enabling tech milestone.
The gap between “we moved some watts” and “we deliver gigawatts reliably and cheaply” is where most narratives die.

Phase 1 (now): orbital “edge inference” for sensors (most realistic)
This is the least sci‑fi version: Earth‑observation, maritime tracking, missile warning, weather, comms—these generate huge raw data. If you do inference onboard (classify, compress, decide), you downlink answers, not terabytes. That’s a real economic driver: less bandwidth, faster action, fewer ground stations. This is where the “sky‑brains” concept fits today.

Phase 2 (next): purpose-built “AI sats” as compute add-ons (plausible, niche)
Now you’re not just processing your own sensor stream—you’re renting compute. The constraints are cooling (radiation only), radiation tolerance/bit‑flip risk, launch cadence, and networking. This can work in niche workloads where latency, sovereignty, or resilience beats pure cost.

Phase 3 (later): space-based solar power beaming to Earth (hardest, but biggest)
Space solar is the “Kardashev headline,” but it’s also the hardest engineering stack: colossal lightweight structures, conversion losses, beam safety/regulation, and grid integration on the ground. For context on scale: the ISS’s new roll‑out solar arrays are meaningful—but they’re still ~tens of kW per array—getting to gigawatts is a different universe of manufacturing and deployment.

These benefit if any version of orbital compute/space power ramps—because they sell the bus, payload integration, comms, and space-qualified components.

If orbital inference is real, you need radiation-tolerant compute, not just datacenter GPUs.

Even if compute moves to orbit, the near-term reality is: AI is still mostly on Earth. The grid buildout (transformers, switchgear, cooling, power distribution) remains the dominant spend path.

Noise: everyone is still anchored on “does the Fed cut in December?” News: bond traders are telling you the easing cycle is likely shallower and shorter than the equity market wants — and, more importantly, that the Fed can cut the front end while the long end refuses to cooperate. That’s the message in the price action: long yields aren’t behaving like they’re being “pulled down” by policy cuts, and the debate is getting louder precisely because this kind of disconnect hasn’t shown up in a sustained way since earlier regimes.

The investor implication is simple: if long rates stay sticky, you don’t get the classic “multiple expansion” tailwind. You get a market that has to earn its upside through cash flows and winners that can self-fund — and you get a harsher penalty box for anything dependent on cheap refinancing, long-duration narratives, or leverage disguised as “capex investment.”

The second piece of real “news” is Oracle earnings tonight as a live stress test of the AI credit cycle. Oracle has become the market’s favorite “canary” because it’s trying to play hyperscaler games without hyperscaler balance-sheet flexibility: Reuters notes Oracle has a roughly $300B OpenAI deal, has tapped the bond market aggressively (including an $18B raise), and is simultaneously staring at negative free cash flow while planning a massive capex ramp (reported as $105B in FY26).

That cocktail is why its credit is being watched like an index: if Oracle’s print suggests the AI buildout is real, diversified, and financeable, it helps stabilize the “AI infra on margin” narrative. If it suggests timing risk, customer concentration risk, or a funding treadmill, it doesn’t just hit ORCL — it tightens financial conditions for the whole levered AI supply chain.

Today stock is Symbotic (SYM)…..

Symbotic is still a real “physical AI” business (warehouse automation at scale), but last week’s -15% move is mostly a capital markets / supply event, not a thesis-breaker: they priced a 10M-share deal at $61.50 (6.5M new shares + 3.5M sold by a SoftBank affiliate, plus an option for 1.5M more) with proceeds for general corporate purposes.

That said, whether it’s a “must own” comes down to one question: can they diversify away from Walmart fast enough to earn a platform multiple? Walmart was ~85% of FY25 revenue and a “significant majority” of the $22.5B backlog—that’s an enormous concentration risk (and the market will keep punishing the stock for it when sentiment turns).

The flip side is the bull case is genuinely asymmetric: they’ve deepened the Walmart relationship by buying Walmart’s robotics unit and signing a separate development agreement for pickup/delivery automation that Walmart funds (reported $200M acquisition + $520M funding structure).

Add first meaningful non-WMT wins (e.g., Medline) and the narrative can shift from “one-customer project company” to “repeatable automation platform.”

The H.E.A.T. (Hedge, Edge, Asymmetry and Theme) Formula is designed to empower investors to spot opportunities, think independently, make smarter (often contrarian) moves, and build real wealth.

The views and opinions expressed herein are those of the Chief Executive Officer and Portfolio Manager for Tuttle Capital Management (TCM) and are subject to change without notice. The data and information provided is derived from sources deemed to be reliable but we cannot guarantee its accuracy. Investing in securities is subject to risk including the possible loss of principal. Trade notifications are for informational purposes only. TCM offers fully transparent ETFs and provides trade information for all actively managed ETFs. TCM's statements are not an endorsement of any company or a recommendation to buy, sell or hold any security. Trade notification files are not provided until full trade execution at the end of a trading day. The time stamp of the email is the time of file upload and not necessarily the exact time of the trades. TCM is not a commodity trading advisor and content provided regarding commodity interests is for informational purposes only and should not be construed as a recommendation. Investment recommendations for any securities or product may be made only after a comprehensive suitability review of the investor’s financial situation.© 2025 Tuttle Capital Management, LLC (TCM). TCM is a SEC-Registered Investment Adviser. All rights reserved.