The hard drive pictured here could hold up to 3/4 of one mp3 song. Though MP3 was not yet invented. 😉 Your terabyte smart phone can fit 200,000 songs, 17,000 hours of music.
__________
Nicholas Nouri
In September 1956 IBM shipped the world’s first commercial hard drive as part of the 305 RAMAC system. This “disk storage unit” weighed over a ton and occupied roughly 16 square feet. It held 50 aluminum platters, each 24 inches in diameter, and yet its capacity was just five megabytes – about five million characters, or the equivalent of 64 000 punch cards. The whole machine needed its own air‑conditioning unit. It seemed comically large to some observers, but it introduced random access storage and freed businesses from wading through mountains of paper.
A few months earlier, a different milestone quietly took place. At Dartmouth College, John McCarthy convened the Summer Research Project on Artificial Intelligence. He proposed that “every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it”. Many sceptics thought this ambition was fanciful.
Yet the meeting seeded a new field of research that would eventually produce today’s language models and robotics.
It’s worth pausing on how far we’ve come – and how far we still have to go.
Modern phones now offer 512 GB or even 1 TB of storage, roughly a hundred thousand times the capacity of the RAMAC drive. A recent Tom’s Guide piece even predicts that half‑terabyte storage may soon become the standard on everyday devices. We casually carry in our pockets what would have required a dedicated room and a team of engineers in 1956.
Artificial intelligence is experiencing a similar tension between clunkiness and potential. A recent report on AI hype points out that genuine breakthroughs – from protein folding to complex reasoning – coexist with inflated valuations and “AI washing”. Technical progress appears sustainable and that investment from traditionally conservative companies is accelerating, but reliability remains a key challenge.
When the RAMAC was unveiled, it was easy to laugh at a one‑ton contraption that stored a handful of megabytes. Today, that machine is a museum piece, and random‑access storage is invisible infrastructure underpinning our digital lives.
In the same way, the current wave of AI tools may feel overhyped, awkward or even unreliable. They are, in many respects, scaffolding – early prototypes that still require careful orchestration, safety work and clearer explanations of their value.
But history suggests that foundational technologies often look ungainly at first. The challenge is to engage critically with AI’s hype and risks, while recognizing that today’s clunky systems could become tomorrow’s invisible utilities.
