On June 1, 2026, Sam Altman stood in a Saline Township field next to Governor Gretchen Whitmer (D-MI), a row of union leaders, and a preserved red barn, and broke ground on the largest economic investment in Michigan history. A $16 billion data center campus for Oracle and OpenAI. Whitmer called it a model for the nation. Altman said the site might be where cancer gets cured.
What almost none of the coverage led with: the same township voted the project down before any of this happened. Saline Township denied the rezoning. The developer and the landowners sued. The township settled, because, as one trustee put it plainly, it could not afford to fight. Residents protested at permitting meetings. The Michigan Attorney General moved to reopen the utility contracts. And then the shovels went into the ground. Per the project's website: "Responsibly designed from the earliest stages to protect Michigan's water, land, and natural resources, The Barn is a gigawatt-scale data center campus under development by Related Digital in Saline Township, Michigan, exclusively for Oracle and its customer OpenAI."
The Barn is a smart word choice that feels All-American and evokes the warmth of a petting zoo and autumn hayrides. Ditching the engineering-correct "data center" for "The Barn" is a masterstroke in communication and narrative shaping.
All of those stories are true at the same time. That is the entire point.
Here is the question every executive should sit with: when a project announces consensus, how do you read whether the consensus is real? Because the announced version of The Barn is union jobs, preserved farmland, and a governor's blessing. The contested version is wetland loss, diesel generators, a fast-tracked utility deal, and a community that was outspent rather than persuaded.
An executive who reads only the first version is uninformed. They are marketed to.
This is a job for SIGNAL™, the framework I use with clients to read, grasp, and understand a volatile operating environment: Sources, Inputs, Geography, Noise, Analysis, Loop. I have used it for years on exactly this gap, the distance between what a situation is announced to be and what it actually is, and on closing that gap before it costs a leader.
Start with Sources. The developer's website says low water use by design. Local environmental reporting says the permits authorize the destruction of roughly 9 acres of wetlands and the operation of 15 diesel backup generators. Both are accurate. The closed-loop cooling system genuinely sips water compared to older designs. The construction genuinely disturbs wetlands and draws groundwater. A single source gives you a slogan. Reading across sources gives you the truth, which is more defensible than either side's headline.
Then Inputs. Are you drawing on a full cross-section of information, from the voices backing the project to the ones fighting it, from domestic permitting records to international best practices and industry white papers? A narrow input set produces a confident, wrong picture. The discipline is in widening the aperture before you decide what you think.
Then Geography, which can be broad or literal, and in this case, it is very literal. It is the whole story. This is the Global Great Lakes, the region that powered the last industrial era and happens to hold exactly what the AI era is starving for: a utility capable of delivering 1.4 gigawatts of power, water, a cool climate, land, and a building-trades workforce. Michigan was not chosen at random. It was chosen because the bones are already here. The same Geography that makes the site valuable to Oracle makes the Saline River and the farmland valuable to the people who live there. Read the place, and you understand both the opportunity and the opposition.
Then Noise. The word everyone used at the groundbreaking was responsible. Whitmer, Oracle's CEO, and OpenAI's CEO all invoked the need for responsible development. Oracle's Clay Magouyrk admitted that a project at this scale has not been done before and that they "learn as we go." Critics fear precisely that: that learning as they go will leave irreversible impacts. So, responsible development is the most repeated phrase in the project and the least defined. That is Noise wearing the costume of a Signal. A slogan everyone can say is not a commitment anyone has made.
And the Loop, the part of SIGNAL that asks what this environment is teaching you for next time, now points somewhere larger than Saline. The backlash that produced one township's no vote has gone national in the months since. Analysts at Jefferies have started calling data centers the villain of the 2026 midterms as Americans link them to rising electric bills. At least eleven states have proposed restrictions or bans since late 2025, and Maine is moving to halt construction outright. The anger has also turned ugly: in April, an Indianapolis councilman who backed a data center found thirteen bullet holes in his front door and a note reading "No Data Centers," and the OpenAI CEO, standing in that Saline field, had a Molotov cocktail thrown at his home weeks earlier. None of that excuses violence, and the overwhelming majority of opposition is peaceful and legitimate. But it tells you the temperature. A project that treats community consent as a box to check is now building into the hottest political environment American infrastructure has faced in a generation.
All of this shows where executives consistently get it wrong on projects like this. They treat communications as a press release after the decision. The Barn shows the opposite. The communications failure happened upstream, at the township vote, when a 2,200-person community experienced a gigawatt-scale campus as something done to them rather than something built with them. By the time the videos played at the groundbreaking, arguing the companies were being thoughtful and careful, the trust deficit was already priced in. You cannot communicate your way out of a position you communicated your way into.
The world has changed. The way you need to explain it has not caught up. Most senior executives and founders still treat communications as a tactic for explaining decisions after they are made. The ones who win in an environment like this treat communications as a strategy that shapes how an audience understands a decision before, during, and after it is made, not just an explanation after the fact. The Barn will get built.
Whether it becomes the model Altman promised or the cautionary tale his critics predict will be decided less by the engineering than by whether the people around it ever believed they were partners. Reading the environment accurately is the work of SIGNAL. Mapping the audiences inside it is the work of TWIN™. Spreading the idea once both are clear is E-STOCK™'s work.
These are the disciplines I use every day. The builders did the engineering first. The next gigawatt should do the reading first, with a focus on winning the communications from day one.
Enjoy the ride + plan accordingly.
-Marc
You can always reach me @ marcaross@gmail.com
Two firms read the same filing and priced it a trillion dollars apart
SpaceX has filed an amended prospectus that sets up the largest initial public offering in history. The numbers are now fixed rather than rumored. The company will sell 555.6 million Class A shares at $135 each, raising roughly $75 billion, at a valuation near $1.77 trillion. That would make SpaceX the seventh-largest company in the United States on its first day of trading, ahead of Tesla and Meta, and would eclipse Saudi Aramco's 2019 debut as the biggest listing ever attempted. Trading begins on the Nasdaq on June 12 under the ticker SPCX.
Also, Morningstar's analysis suggests a fair value for the SpaceX IPO of $780 billion, less than half.
That is the story. Two serious institutions looked at the identical document. They arrived at numbers a trillion dollars apart, and a great many sophisticated people are about to commit capital without being able to say which is closer to the truth. That is not a pricing problem. It is an intelligence problem, and it is the clearest illustration this year of why reading the operating environment is a discipline, not a reflex.
This is where the SIGNAL Framework™ earns its place. The discipline of executive intelligence is not about collecting more inputs. It is knowing which inputs carry weight, where the noise is engineered rather than incidental, and how to separate the two before you act. The SpaceX offering is a near-perfect test, because the signal and the noise come apart cleanly once you decide to look.
Start with the real inputs. SpaceX reported $18.674 billion in 2025 revenue, up 33 percent year over year. Starlink alone accounts for roughly $11.4 billion of that and grew about 50 percent. The company launched 83 percent of all mass sent to orbit from Earth last year. Those are the load-bearing facts, and they are why even the skeptical valuation places a value of around $611 billion on the core launch and connectivity businesses. On the rockets and the satellites, the bull and the bear largely agree.
The disagreement lives in one place: the AI business folded in through the xAI merger. Morningstar values that piece at roughly $170 billion in probability-weighted scenarios and calls its economic moat indeterminate, which is analyst language for "we cannot tell you what this is worth, and neither can anyone else." Almost the entire trillion-dollar gap is a bet on an asset no one can model with confidence. The core business is knowable. The premium is a wager. Then comes the engineered noise on top of it: Musk comparing SpaceX to Tesla's IPO; a deliberately thin float of roughly 3 percent; a rule change positioning the stock for forced index-fund buying; Jamie Dimon walking JPMorgan's clients through the deal. None of it is illegitimate. All of it is designed to shape how the offering is understood before anyone reads the business clearly. The discipline is noticing that the loudest inputs are the ones built to be loud.
Here is the part that matters for the executive who does not run a rocket company. When you are the one being valued, the market runs this same process on you, and most of what it prices is not your fundamentals. So run it on yourself first.
Three steps, and you can start.
One: Separate your own story into the part that is knowable and the part that is a bet. SpaceX's knowable core is $611 billion. Its premium is an unmodelable AI asset. Your company has the same split. Name the revenue, the contracts, the operating facts a skeptic would concede, and then name the part of your valuation that rests on a story about the future. Be honest about which is which.
Two: Find the number the market will fill with a narrative rather than a model, because there is always one. For SpaceX, it is the AI premium. For you, it might be a pipeline, a category you claim to be creating, a regulatory outcome, or a founder's track record. That is the number that moves your valuation, and it is the one you cannot leave to others to define.
Three: Own that number's framing before someone else sets it. Morningstar set SpaceX's framing this week, in public, with a headline that traveled. The lesson is not that SpaceX should have stayed quiet. It is that the unmodelable number gets a story attached to it, whether or not you write the story. The companies that win decide they will be the author.
Reading the world accurately is the work of SIGNAL. Spreading an idea until the market prices it the way you want is the work of E-STOCK™, and the SpaceX roadshow is a textbook execution of it.
Most companies will never set their own valuation narrative the way Musk is. But every company is being read on the same terms, and the reading happens before the decision lands, not after.
Enjoy the ride + Plan accordingly.
-Marc
I help senior executives + founders sound like the smartest person in the room on what’s coming next — and know how to talk about it. | Founder @ Caracal Global + Brigadoon. | DET + WAS + EDI + LON
The Communicating Caracal Global | May 29, 2026
May 29, 2026
Communications at the intersection of globalization, commerce, and governments. Five situations where the story being shaped matters more than the facts on the ground, plus a win and a loss of the week.
The lead: The White House is running two foreign policies and reconciling neither
The most important communications failure of the week is a government that cannot give the same answer to the same question twice.
Over five days, the administration signaled a deal to end the Iran war was close, then, at a Cabinet meeting, Trump said that rising economic costs would not push him into a deal. Brent crude whipsawed from $144 to below $100 as the message changed. The New York Times described the president's posture on Iran as driven by mood and moment rather than any discernible strategy, and a PBS Frontline documentary built the same case from inside the war cabinet.
For executives, the lesson is not about Iran. It is about what happens to any organization that lets its narrative move with the mood of the room. When markets cannot tell a pause from a peace, they price confusion, and confusion is expensive. The discipline that protects a company here is the one most leaders skip: deciding what you will say, to whom, and in what sequence, before the event forces you to improvise. Strategic incoherence is now a planning variable at the highest level of the US government. Assume your stakeholders have noticed, and build communications consistency that the headlines are not giving them.
The reputational risk in your vendor stack is now a security risk
Reuters reported that US military personnel deployed to war zones are being targeted using commercially available location data. Read that sentence as a business leader, not a Pentagon analyst.
The global surveillance economy is no longer an abstraction in an academic white paper. It is happening in real time on the battlefield and in combat zones, and the data your vendors broker, sell, or aggregate is accessible to people you would never choose as employees, business partners, or customers. The partnerships that looked like growth in 2024 can now be a liability in 2026. The communications problem arrives the moment a journalist or a regulator connects your brand to that data trail, because by then, you are explaining a decision instead of having shaped how it was understood.
The defensive response is already visible: own your stack rather than rent exposure through third parties.
News broke that Kirkland & Ellis has committed $500 million over the next three to four years to develop its own proprietary artificial intelligence platform. According to Bloomberg, the investment represents one of the most ambitious technological bets ever made by a law firm. It also signals Kirkland’s strategy to control its own tech to outperform competitors in the legal industry’s competitive race to harness AI.
The move is to audit the story your data relationships tell before someone else tells it for you. Name the partnerships that would be hard to defend on a front page. Decide now what you would say about each one. The companies that get caught flat are not the ones with the worst practices. They are the ones who never rehearsed the answer.
The gap between your written AI policy and your real one is the story
VP Vance told Air Force Academy graduates the military should never let AI make life-and-death decisions. The Pentagon is moving forward with AI in war regardless. That gap between the stated policy and the operational reality is the entire communications exposure, and every company has its own version.
You probably have a written AI policy. You also have an unwritten one that is already different, because the line between "AI assists" and "AI decides" gets crossed in workflow design, not in a policy memo. Amazon scrapped an internal AI leaderboard after a senior executive told staff to stop using AI just for the sake of using it. Starbucks began factoring AI use into tech workers' bonuses. Corporate America is rationing AI as the compute bill comes due. Each of those is a company discovering that its real policy and its stated one have drifted apart in public.
Close the gap before a reporter finds it. The most defensible position is one where the policy you publish and the practice you run describe the same company. If they do not, the discrepancy is your next crisis, and you will be explaining it on someone else's timeline.
Rome just entered the AI conversation, joining Brussels in oversight
Pope Leo XIV used his first major encyclical, Magnifica Humanitas, to place moral concern rather than profit or efficiency at the center of the AI debate, comparing the technology's risks to the Tower of Babel. Anthropic helped unveil it. The Vatican is now an active voice in AI governance, alongside an EU AI-envoy role that remains unfilled, with the GCHQ chief calling AI an unstoppable force being weaponized in the gray zone between peace and war.
For any multinational, the takeaway is that the regulatory and reputational framework around AI is no longer set solely in Brussels. The moral frame is now being written in Rome, the security frame in London, and the market frame in Silicon Valley, and they do not agree. A company that tuned its AI messaging to a single regulator is now under-built for the conversation that actually exists.
Embrace the mantra 'where you sit is where you stand.' It will help you see how technology policy actually gets formed — and it is not only by those who issue fines and write laws. The audience that shapes your license to operate on AI now includes a moral authority with a billion followers. Speak to all the stakeholders who will shape and guide AI policy; it is far broader than what your compliance team has modeled.
Dynamic pricing was an operations decision. It just became a communications liability.
The New York and New Jersey attorneys general subpoenaed FIFA over World Cup ticket pricing, investigating whether staggered releases and dynamic pricing misled buyers. Strip out the soccer and the lesson is universal: the pricing mechanism that quietly migrated from airlines to concerts to sports to restaurants has finally drawn legal exposure, and the exposure is as much reputational as legal.
The failure here was never the algorithm. It was the absence of a story to go with it. Dynamic pricing optimizes revenue while generating resentment, and the companies running it have mostly decided to let the price speak for itself. That is a message, and it is the one customers heard.
If your business uses dynamic pricing, you are one investigation away from having to explain it. Decide now whether you can. The operators who survive this will be the ones who built a defensible narrative around their pricing before a subpoena required one. Silence is a position here, and it is the costly one.
Communications Win of the Week: The audience that owns the distribution
CBS suspended its takedown notices on bootleg YouTube uploads of Stephen Colbert's old public-access show, Only in Monroe, after public outcry: small story, large lesson. The instinct of a rights holder is to enforce. The smarter read was that the audience now sets the terms of distribution, and a company that fights its own fans for control of content loses the room while winning the case. CBS reversed fast enough to turn an enforcement misstep into goodwill. The win was not the content. It was the speed of the climbdown.
Communications Loss of the Week: Ferrari, defining its own launch as desecration
Ferrari unveiled Luce, its first all-electric car, with a Jony Ive design and a €550,000 price tag. The vehicle may be brilliant, but the communications were a rout. The company's own former chairman publicly called it the destruction of a legend and told Ferrari to take the prancing horse off it. When the most quotable critic of your launch is your predecessor, you have lost control of the story before the first car ships. The lesson is old and expensive: a launch is a narrative, and if you do not author it, the loudest insider will. Ferrari let the purists frame Luce as a betrayal, and now every review is litigating heritage rather than engineering.
Enjoy the ride + plan accordingly.
— Marc
Marc A. Ross | Chief Communications Strategist @ Caracal Global
You can always reach me @ marc@caracal.global.
The Communicating Caracal Global is a weekly memo applying the Caracal Global lens to the week's events, resolved on the communications stakes that decide how a company is understood by the audiences that matter most.
Caracal Global is a communications firm for global business, working at the intersection of geopolitics, commerce, and governments, specializing in Globalization + American Politics. Intelligence + Strategy + Communications.
