<div dir="ltr"><div dir="ltr"><div class="gmail_default" style="font-family:arial,helvetica,sans-serif"><span style="font-family:Arial,Helvetica,sans-serif;background-color:transparent">On Tue, Jun 30, 2026 at 9:20 AM BillK via extropy-chat <<a href="mailto:extropy-chat@lists.extropy.org">extropy-chat@lists.extropy.org</a>> wrote:</span></div></div><div class="gmail_quote gmail_quote_container"><div dir="ltr" class="gmail_attr"><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><i><font face="georgia, serif"><span class="gmail_default" style="font-family:arial,helvetica,sans-serif">> </span>Anybody Who Thinks Orbital Data Centers are a Good Idea Is Suffering<br>
from AI Psychosis, Experts Argue<br>
<br>
"It really seems like anyone with some renders and a white paper<br>
written by someone being gassed up by an overly agreeable AI can get<br>
VC funding these days."<br>
By Victor Tangermann Published Jun 30, 2026<br>
<br>
<<a href="https://futurism.com/artificial-intelligence/orbital-data-centers-ai-psychosis" rel="noreferrer" target="_blank">https://futurism.com/artificial-intelligence/orbital-data-centers-ai-psychosis</a>><br>
Quote:<br>
But the emerging consensus is that the plan isn’t just overly<br>
ambitious; it would require several technological revolutions to<br>
overcome glaring engineering challenges.</font></i><br></blockquote><div><br></div><div><font size="4" face="tahoma, sans-serif"><b>I asked GPT<span class="gmail_default" style=""> to read the article and give me his (or hers or its) </span>opinion of i<span class="gmail_default" style="">t:</span></b></font></div><div><br></div><div><p class="gmail-PDq2pG_selectionAnchorContainer"><font size="4"><span class="gmail_default" style=""><font face="arial, helvetica, sans-serif"></font><b style=""><font face="tahoma, sans-serif">GPT:</font></b><font face="arial, helvetica, sans-serif"> "</font></span>The Futurism article and the accompanying <strong>Real Engineering / IEEE Spectrum</strong> video make a number of good engineering points, but I don't think they establish their strongest conclusion—that orbital data centers are <em>inherently</em> unworkable. What they really show is that <strong>certain very ambitious first-generation proposals appear extremely optimistic.</strong> That's a much narrower claim.<span aria-hidden="true" class="gmail-PDq2pG_selectionAnchor"></span></font></p>
<p><font size="4">The strongest rebuttal is not "they're obviously feasible." It's that the critics repeatedly blur together three separate questions:</font></p>
<ol>
<li>
<font size="4">Is the physics possible?
</font></li>
<li><font size="4">
Is it practical with foreseeable technology?
</font></li>
<li><font size="4">
Is it economically superior to terrestrial data centers?
</font></li>
</ol>
<p><font size="4">Those are very different questions.</font></p>
<hr>
<h2><font size="4">1. The article often argues against an extreme design rather than the general concept</font></h2>
<p><font size="4">The Futurism article focuses heavily on Starcloud's enormous proposed 5-GW orbital facility, quoting calculations that it would weigh over 100 million kilograms and require radiator areas measured in square miles. <span class="gmail-"></span></font></p>
<p><font size="4">But that's similar to arguing:</font></p>
<blockquote>
<p><font size="4">"The Airbus A380 is too difficult to build, therefore airplanes are impossible."</font></p>
</blockquote>
<p><font size="4">The engineering challenges generally scale with power.</font></p>
<p><font size="4">There is a tremendous difference between</font></p>
<ul>
<li><font size="4">
a 5 MW demonstrator,
</font></li>
<li><font size="4">
a 50 MW commercial platform,
</font></li>
<li><font size="4">
a 500 MW facility,
</font></li>
<li><font size="4">
and a 5 GW megaproject.
</font></li>
</ul>
<p><font size="4">Many technologies mature through exactly this progression.</font></p>
<hr>
<h2><font size="4">2. Cooling is difficult—but not a showstopper</font></h2>
<p><font size="4">The video correctly states that space has no air or water to carry heat away.</font></p>
<p><font size="4">This is absolutely true.</font></p>
<p><font size="4">However, it does <strong>not</strong> follow that cooling is impossible.</font></p>
<p><font size="4">Every spacecraft already rejects essentially all of its heat by radiation.</font></p>
<p><font size="4">Examples include:</font></p>
<ul>
<li>
<font size="4"><span class="gmail-hover:entity-accent entity-underline gmail-inline gmail-cursor-pointer gmail-align-baseline"><span class="gmail-whitespace-normal">International Space Station</span></span>
</font></li>
<li>
<font size="4"><span class="gmail-hover:entity-accent entity-underline gmail-inline gmail-cursor-pointer gmail-align-baseline"><span class="gmail-whitespace-normal">James Webb Space Telescope</span></span>
</font></li>
<li>
<font size="4"><span class="gmail-hover:entity-accent entity-underline gmail-inline gmail-cursor-pointer gmail-align-baseline"><span class="gmail-whitespace-normal">Juno</span></span>
</font></li>
</ul>
<p><font size="4">The physics is straightforward:</font></p>
<p><font size="4">electrical power in</font></p>
<p><font size="4">↓</font></p>
<p><font size="4">computation</font></p>
<p><font size="4">↓</font></p>
<p><font size="4">heat</font></p>
<p><font size="4">↓</font></p>
<p><font size="4">radiators</font></p>
<p><font size="4">The issue is simply radiator size.</font></p>
<p><font size="4">Large radiators are expensive.</font></p>
<p><font size="4">Large radiators are fragile.</font></p>
<p><font size="4">Large radiators increase launch mass.</font></p>
<p><font size="4">Those are engineering and economic problems—not violations of thermodynamics.</font></p>
<hr>
<h2><font size="4">3. The critics sometimes underplay one genuine advantage: continuous solar power</font></h2>
<p><font size="4">One of the biggest advantages of space is easy to overlook.</font></p>
<p><font size="4">Ground data centers must deal with</font></p>
<ul>
<li><font size="4">
nighttime
</font></li>
<li><font size="4">
weather
</font></li>
<li><font size="4">
seasonal variation
</font></li>
<li><font size="4">
transmission losses
</font></li>
<li><font size="4">
grid congestion
</font></li>
</ul>
<p><font size="4">A satellite in a dawn-dusk Sun-synchronous orbit can receive sunlight almost continuously.</font></p>
<p><font size="4">That eliminates:</font></p>
<ul>
<li><font size="4">
batteries
</font></li>
<li><font size="4">
peaker plants
</font></li>
<li><font size="4">
transmission infrastructure
</font></li>
<li><font size="4">
land acquisition
</font></li>
<li><font size="4">
local opposition
</font></li>
</ul>
<p><font size="4">Whether those savings outweigh launch costs is uncertain.</font></p>
<p><font size="4">But they are real benefits.</font></p>
<hr>
<h2><font size="4">4. Launch cost trends matter enormously</font></h2>
<p><font size="4">Many arguments implicitly assume today's launch economics.</font></p>
<p><font size="4">But launch costs have been falling for decades.</font></p>
<p><font size="4">If fully reusable launch systems become routine, the economics change dramatically.</font></p>
<p><font size="4">Suppose launch cost eventually falls another factor of ten.</font></p>
<p><font size="4">Many designs dismissed today suddenly become plausible.</font></p>
<p><font size="4">That doesn't prove orbital data centers win.</font></p>
<p><font size="4">It means the answer depends on assumptions about future launch costs rather than physics.</font></p>
<hr>
<h2><font size="4">5. Radiation is serious—but modern computing already handles errors</font></h2>
<p><font size="4">The video argues that radiation-induced bit flips would produce catastrophic AI hallucinations.</font></p>
<p><font size="4">That's somewhat overstated.</font></p>
<p><font size="4">Current computing already deals with bit errors using</font></p>
<ul>
<li><font size="4">
ECC memory
</font></li>
<li><font size="4">
redundant computation
</font></li>
<li><font size="4">
error detection
</font></li>
<li><font size="4">
checkpointing
</font></li>
<li><font size="4">
majority voting
</font></li>
</ul>
<p><font size="4">Radiation-hardened spacecraft have done this for decades.</font></p>
<p><font size="4">The question is cost.</font></p>
<p><font size="4">Consumer AI accelerators would almost certainly require additional protection.</font></p>
<p><font size="4">That reduces performance.</font></p>
<p><font size="4">Again:</font></p>
<p><font size="4">engineering tradeoff,</font></p>
<p><font size="4">not impossibility.</font></p>
<hr>
<h2><font size="4">6. Maintenance is probably the strongest criticism</font></h2>
<p><font size="4">Here I think the critics have their best argument.</font></p>
<p><font size="4">Today's AI hardware becomes obsolete astonishingly quickly.</font></p>
<p><font size="4">Many companies replace GPUs every<span class="gmail_default" style="font-family:arial,helvetica,sans-serif"> </span></font><span style="font-size:large;background-color:transparent">2–4 years.</span></p>
<p><font size="4">Launching expensive hardware into orbit that cannot easily be upgraded is a genuine disadvantage.</font></p>
<p><font size="4">However, SpaceX's entire long-term strategy assumes something critics often ignore:</font></p>
<blockquote>
<p><font size="4">frequent, inexpensive launches.</font></p>
</blockquote>
<p><font size="4">If launching replacement modules eventually becomes almost routine, orbital servicing begins to resemble replacing racks in a terrestrial data center.</font></p>
<p><font size="4">That is admittedly speculative—but it is precisely the capability companies like SpaceX are attempting to create.</font></p>
<hr>
<h2><font size="4">7. The article assumes Earth workloads</font></h2>
<p><font size="4">This is probably the biggest conceptual weakness.</font></p>
<p><font size="4">Not every workload needs enormous bandwidth to Earth.</font></p>
<p><font size="4">For example:</font></p>
<ul>
<li><font size="4">
Earth observation
</font></li>
<li><font size="4">
military reconnaissance
</font></li>
<li><font size="4">
satellite image analysis
</font></li>
<li><font size="4">
planetary science
</font></li>
<li><font size="4">
asteroid mining
</font></li>
<li><font size="4">
lunar operations
</font></li>
<li><font size="4">
Mars communications
</font></li>
</ul>
<p><font size="4">In many of these cases, the raw data already originates in space.</font></p>
<p><font size="4">Instead of downloading petabytes,</font></p>
<p><font size="4">process them in orbit,</font></p>
<p><font size="4">and transmit only the final answers.</font></p>
<p><font size="4">Several recent academic papers argue that <strong>communication</strong>, rather than computation, is the fundamental bottleneck and propose semantic compression and workload-specific architectures to make orbital computing practical. <span class="gmail-"></span></font></p>
<p><font size="4">Those use cases are much stronger than trying to replace every terrestrial AI cluster.</font></p>
<hr>
<h2><font size="4">8. Latency limits what belongs in orbit</font></h2>
<p><font size="4">One point that deserves emphasis is that orbital data centers are unlikely to replace conventional cloud computing.</font></p>
<p><font size="4">Round-trip latency to low Earth orbit is only a few milliseconds in vacuum, but the complete network path—including ground stations, routing, and optical links—adds delay and complexity. As several analyses note, networking and communication become the defining constraint. <span class="gmail-"></span></font></p>
<p><font size="4">That means orbital facilities are best viewed as a complement to Earth-based infrastructure, not a universal replacement.</font></p>
<hr>
<h2><font size="4">9. History argues against declaring engineering ideas "impossible"</font></h2>
<p><font size="4">Many technologies were once widely dismissed because they seemed economically absurd:</font></p>
<ul>
<li><font size="4">
reusable orbital rockets
</font></li>
<li><font size="4">
mass-produced electric vehicles
</font></li>
<li><font size="4">
private orbital launch
</font></li>
<li><font size="4">
satellite internet constellations
</font></li>
</ul>
<p><font size="4">None violated physics.</font></p>
<p><font size="4">They simply required advances that critics underestimated.</font></p>
<p><font size="4">That doesn't mean orbital data centers will succeed.</font></p>
<p><font size="4">It does mean we should be cautious about moving from "very difficult" to "inherently impossible."</font></p>
<hr>
<h2><font size="4">My overall assessment</font></h2>
<p><font size="4">I think the critics are <strong>mostly right in the short term</strong>, but they overstate their conclusion.</font></p>
<p><font size="4">I would separate the claims like this:</font></p>
<div class="gmail-TyagGW_tableContainer"><div tabindex="-1" class="gmail-group gmail-TyagGW_tableWrapper gmail-flex gmail-flex-col-reverse gmail-w-fit"><table class="gmail-w-fit gmail-min-w-(--thread-content-width)"><thead><tr><th class="gmail-last:pe-10"><font size="4">Claim</font></th><th class="gmail-last:pe-10"><font size="4">My assessment</font></th></tr></thead><tbody><tr><td><font size="4">A 5-GW orbital AI data center can be built economically today</font></td><td><font size="4">Almost certainly false.</font></td></tr><tr><td><font size="4">Cooling, servicing, and radiation are major engineering challenges</font></td><td><font size="4">True.</font></td></tr><tr><td><font size="4">Space data centers will replace terrestrial ones within a few years</font></td><td><font size="4">Very unlikely.</font></td></tr><tr><td><font size="4">Orbital data centers are physically impossible</font></td><td><font size="4">No.</font></td></tr><tr><td><font size="4">Small or specialized orbital compute platforms could become viable over the next couple of decades</font></td><td><font size="4">Plausible, especially for workloads where the data already originates in space.</font></td></tr></tbody></table></div></div>
<p><font size="4">In other words, I think the evidence supports <strong>"far harder and more expensive than proponents currently suggest"</strong> much more strongly than <strong>"fundamentally unworkable."</strong> Even critics generally acknowledge that the obstacles are economic and engineering in nature, rather than violations of physical law<span class="gmail_default" style="font-family:arial,helvetica,sans-serif">."</span></font></p></div><div><font size="4"><br></font></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><br>
</blockquote></div></div>