Innovation & Tech

AI Is the Light Bulb — the Money Is in the Network Underneath

Over 128 million U.S. smart meters prove the point, says Anterix CEO Scott Lang: AI is the visible light bulb, but networks connecting infrastructure will decide who captures the value.

By Daniel Okafor

5 min read

AI is today’s light bulb
AI is today’s light bulbanalogicus / pixabay

What's News

  • More than 128 million U.S. electric meters — over three in four — are smart meters, per FERC
  • Edison's Pearl Street Station opened September 4, 1882 with 59 customers lighting about 400 lamps within a month
  • Harvey Hubbell patented the separable electric plug in 1904; its parallel-blade shape is still standard
  • Electricity took half a century to reach most U.S. homes; smart meters took two decades; Lang says the AI shift will be measured in years
  • Scott Lang is president and CEO of Anterix, which he founded around grid digitization in 2004

More than 128 million U.S. electric meters are now smart meters — over three out of four nationwide, according to the Federal Energy Regulatory Commission. That count anchors an argument from Scott Lang, president and CEO of Anterix: AI is the visible product everyone stares at, but the durable value sits in the networks that connect it to physical equipment.

Lang started making this case in 2004, when he set out to digitize the last mile of the electric grid — the wires and meters between the substation and the homes and businesses it serves. His bet was that internet protocol, the common language of the world's computers, would eventually connect the grid too. Once it did, every meter and sensor could talk to the software running it. "Plenty of smart people told me it would never happen," he writes. Utilities, like pipeline, rail, and water operators, plan in decades and cannot afford to get reliability wrong, so the real work was finding the handful of utility leaders willing to go first.

What does Edison's grid teach about AI?

When Thomas Edison switched on his Pearl Street Station in lower Manhattan on September 4, 1882, he was selling light. Within a month, 59 customers were using it to light about 400 lamps; by 1884 the station served more than 500 customers. In The Age of Edison, historian Ernest Freeberg shows that what remade American cities wasn't the light bulb itself but what it required: every building needed a wire, and every wire needed a power supply behind it.

Once that network existed, it outgrew its first purpose almost immediately. Electric light and the electric elevator let buildings grow taller. Factories ran through the night. Cities stayed awake after dark.

The unexpected uses followed. Early household appliances had nowhere to plug in, so people screwed fans and irons into ceiling sockets. In 1904, Harvey Hubbell patented a separable plug that made the connection safer and simpler; the parallel-blade shape still in walls today came from his design. Safety standards arrived the same way — Freeberg notes that insurance companies, tired of paying for electrical fires, pushed electric companies into agreeing on the first common rules.

Once the plug standardized, anyone could build against it. Refrigerators, radios, washing machines, and eventually computers all came through that one wire. The light bulb started it. The network built the twentieth century.

Why does the network beat the model?

For more than a century, the grid moved power while pipelines, rail lines, ports, and water systems moved physical goods. Increasingly, every one of those systems also has to move data. Every connected device — a car charger, a data center, a pipeline valve, a rail switch — is becoming a device that sends and receives information.

AI deserves the excitement it gets, Lang argues. But it only creates value when it can reach the equipment it is meant to run. For utilities, pipelines, and water systems, the network is no longer back-office plumbing; it is the platform every new capability will run on.

The network itself is also changing shape. Satellites are beginning to connect directly to devices on the ground, so a sensor on a remote pipeline can be reached from above. Lang predicts the most valuable networks will work on the ground and in orbit as one, and the licensed airwaves linking them will rank among the economy's scarcest resources.

Is this really bigger than the Industrial Revolution?

Lang says yes, and offers a mechanism. The Industrial Revolution gave machines the power to do physical work. What is happening now gives machines the ability to decide and act on work as it happens — software that spots a fault on a power line and reroutes electricity around it before customers notice. The same shift, he writes, is coming to pipelines, rail, ports, and water.

The pace compounds the stakes. Electricity took half a century to reach most American homes. The smart meter transition took about two decades. This one, Lang writes, will be measured in years.

What do the winners have in common?

Robert Metcalfe, who co-invented ethernet, observed that a network's value grows with the square of the number of devices on it. One fax machine is worthless; each machine added makes the others more valuable. Infrastructure networks work the same way, and their devices wear out and get replaced, while a well-built network keeps carrying whatever comes next for decades.

Nobody in 2004 was talking about AI. The utilities that built smart grid networks then are among the best positioned to use AI now. Pipeline, rail, port, and water operators face the same choice today, and Lang frames the downside concretely: operators that wait will retrofit — tearing open the walls to run wire that should have been there from the start, and paying more for it. The winners, he writes, plan for a network without edges, one that reaches a city substation and a remote pipeline, so they never have to bolt it on later.

Lang was called crazy in 2004 for saying the grid would run on a network. He expects the same reaction now to the claim that networks will run all the infrastructure the economy depends on. His closing comparison: Edison thought he was selling light; he was founding a century of infrastructure and innovation. Whoever builds the networks now, on the ground and in orbit, is founding the next one.

Original: ferc.gov

Share this article:

More from Daniel Okafor

Daniel Okafor

Show full bio

Correspondent covering money, funding and market shifts around entrepreneurship, startups and venture funding at Business Bearings.

2,391 articles

Related articles

« Previous article