September 2026 Newsletter

Giant Projects

When I was coming along, we watched films of some of the huge projects the U.S. did, and the ones that stick out were the Hoover Dam, Golden Gate Bridge, and Grand Coulee Dam.

The Hoover Dam, which created Lake Mead—you know, where Elvis Presley boated around in Viva Las Vegas—cost $49 million at the time, which today would be about $1 billion.

The Golden Gate Bridge in San Francisco connected two points on the California coast, saving hundreds of miles in travel or a ferry ride through dangerous currents. It might be the most photographed bridge in the world; it’s an amazing setting. In the 1930s, it cost $35 million; today, it would be about $700 million.

Grand Coulee Dam cost $300 million, which is estimated to be about $700 million today, but the estimated cost to do this project today from scratch would be about $7.5 billion. Built on the Columbia River in central Washington State, it remains the largest electrical generator by initial specs of 6.8 megawatts in the U.S. That’s about seven nuclear power plants.

Hold On to Your Hats!

While all of the above was fantastic at the time, as of now, about $700 billion has been committed to AI data-center buildout in the U.S. It is estimated to grow to $5 trillion by 2030 and $30 to $50 trillion by 2050. What we are going through today in the AI space would be pre-Model T in the automobile technology timeline.

If you want to know where opportunity lives, just follow the money. All of this has a gargantuan effect on many, many industries. Electric demand from AI is estimated to be 67 gigs by the end of the decade. That’s 10 to 20 percent of total U.S. production. Nuclear energy, diesel generators, natural-gas turbines, fuel cells—anything and everything that produces electricity—is booming. Raw-material demand for AI data-center construction is surging. Demand for copper, steel, concrete, aluminum, and rare earths is booming.

No wonder SpaceX is accelerating its efforts for space-based AI compute. In order to be competitive with land-based centers, they have to reduce their launch costs by 70% or more. Elon and company think they have that handled with Starship development. From where I sit, that seems a long way off from the current four experimental launches per year. Given enough time, Elon’s mantra of “develop, build, push till it breaks, reiterate, build again, break again” engineering will work; how much time and money is up for debate.

Just recently, they announced plans for a spaceport on Pecan Island, Louisiana. This 18-mile coastal stretch will eventually house five launch complexes with two launch towers apiece, for a total of 10. Each launch pad will have its own power housing and fuel tank farm; basically, all five will be autonomous. Somewhere, an airport will be built for timely egress. From there, they hope to launch polar-centric orbits, which never run out of sunlight, to power their space-AI ambitions. Pecan Island faces south, which will provide an advantageous polar-orbit trajectory. First launch: 2029, according to SpaceX. The total for all this initially: $100 billion. This complex is estimated to provide approximately 10,000 direct and indirect jobs. No wonder all of the who’s who in Louisiana were on Pecan Island for the announcement.

Large projects are coming. I really don’t think the scale and size has really sunk in with most U.S. residents.

Looking at the date of this newsletter, it never hurts to remind ourselves that freedom has a price. Thank you to all first responders.

Thanks, Andy McClung, CFP®

Sources: Google.com/search; Ask Gemini

2026 Market Results – 09/11/26