The Rise and Fall of the Cloud – Again with Tom Lyon

Tom Lyon begins by suggesting that if cloud computing is defined as outsourcing data processing to a company that owns the equipment, then the concept is nearly a hundred years old. He traces its origins to the 1930s, when IBM established service bureaus where clients could bring data to be processed using punch cards and tabulating machines, an expensive service akin to modern cloud offerings. This early period, marked by the Great Depression, saw basic arithmetic being outsourced, with computing often done by “human computers” before the widespread adoption of machines. The post-World War II era saw advanced punch card computations and a 1956 IBM consent decree that necessitated the creation of the Service Bureau Corporation, highlighting the significance of outsourced data processing even then.

The evolution continued into the 1960s with the proliferation of service bureaus, the birth of timesharing, and the emergence of software as a distinct business. The late 60s witnessed “go-go years” with the concept of a “computer utility” – a direct precursor to modern cloud computing – fueled by remote access, modems, and hard drives, leading to “irrational exuberance” and a subsequent “major depression” in the early 70s. This bust was exacerbated by a shift from services to software and the rise of the mini-computer. The late 70s and 80s brought networking innovations and the desktop era, with the “network is the computer” philosophy solidifying the idea of distributed computing, though general computing wasn’t yet fully within the network “cloud”. The late 90s dot-com boom saw the rise of ISPs and early Infrastructure as a Service (IaaS) providers like Loudcloud and TerraSpring, again characterized by “irrational exuberance” and ambitious data center plans.

However, this boom also led to a significant bust in the early 2000s, which Lyon attributes more to “telecom fraud” than just dot-com speculation. AWS launched in 2006, offering basic cloud services, just before the real estate crash. The 2010s saw AI “get real” with breakthroughs like Watson and AlexNet, propelled by GPU processing and big data. Today, in the 2020s, AI is experiencing “total irrational exuberance,” with an “insane” build-out of data centers, NVIDIA’s dominance, and concerns about creative accounting and fraud. Lyon warns of an impending “AI recession” driven by unsustainable growth expectations, massive infrastructure challenges (especially in energy and water), data sovereignty concerns, and copyright issues. While acknowledging the underlying value of AI, he suggests a period of “normalcy” is five to ten years away, similar to how previous busts eventually paved the way for future growth by leaving behind overbuilt but eventually useful infrastructure.