I Predicted SpaceX Q2 Earning $100B ARR, 2+ GW 2026- NEXT 18 Critical Months.
A 40% spot uplift + cheaper power/shell + wrappers can pull the crossover forward by a quarter or more versus a pure long-term $35–50B locked-rate case.
Risks remain execution on power reliability/permitting, utilization, and any broader rate compression, but the combination of build speed, ownership of modular generation, supplier wrappers, and elevated short-term monetization makes the de-risked path to positive cash flow in the second half of 2027 highly plausible under the assumptions given.
Q3 or Q4 2027 quarterly cash generation from already-completed, high-rate leased capacity begins to exceed the capex being spent on the next 1–2 quarters of ramp. Building GWs and the ~6–8 month payback cycle compounds.
2028 self-funding / strongly cash-flow positive but then AI in Space ramp. Pay the factories and launch facilities.
Each Starship load. Tight payback cycle. 4-5 Months. Build, prep, launch, move to right orbit, activate
Brian Wang is a Futurist Thought Leader and a popular Science blogger with 1 million readers per month. His blog Nextbigfuture.com is ranked #1 Science News Blog. It covers many disruptive technology and trends including Space, Robotics, Artificial Intelligence, Medicine, Anti-aging Biotechnology, and Nanotechnology.
Known for identifying cutting edge technologies, he is currently a Co-Founder of a startup and fundraiser for high potential early-stage companies. He is the Head of Research for Allocations for deep technology investments and an Angel Investor at Space Angels.
A frequent speaker at corporations, he has been a TEDx speaker, a Singularity University speaker and guest at numerous interviews for radio and podcasts. He is open to public speaking and advising engagements.
This is just my Armchair Guy take, but I still think that SpaceX’s primary revenue will still end up being from space applications and not language models.
I could see the best models ultimately ending up in the hands of the public, open-sourced, and the hardware becoming cheap because the models will be so efficient. Fight me lol (but not yet, I just ate).
I really see current “A.I.” becoming primarily decentralized and local, without requiring data centers and internet access until you want to look for something specific. I think fine-tuning will become efficient and simple, too. And, that corporate control over the public is going to diminish in a vast way that we can’t yet foresee. But, perhaps that’s just me putting the energy out there of what I want to see happen.
The AI acceleration plus the open weights/open models being quantized are giving me whiplash.
Trying to stay on top of what’s relevant and what’s changed is exhausting.
Will inexpensive agents plus good enough models cause investors to flee the funding of data centers for expensive frontier models? If so, where would the money go?
Or maybe frontier AIs continue to gain token spend fueled by defense contractors, internet security exploits, big Pharma/Materials Science, and the fear of foreign governments gaining an upper hand.
Who knows?
A year ago, no one saw open clawd coming. No one saw the rapid rise of multi-agent AI work multipliers. Recently, huge corporations pulled back on maximizing token usage because of the cost. They couldn’t justify the spend compared to measurable gain in productivity and/or profitability.
What if AI agents start building their own cheap frontier models without requiring humans in the loop? What will that do to AI monetization?
We are building electrical generation plants at break-neck speeds. We are building chip fab plants at break-neck speeds. We are rushing toward humanoid robots imbued with AI reasoning and autonomy. That sounds like a recipe for abundance.
But who knows?
Everyday, some new frontier model and some new use for AI agents is announced. What’s understood today is nothing like what’s understood tomorrow. Throw all previous forecasts out the door.
I’m just going to lay down and take a long nap….