1958: Jack Kilby demonstrates the first integrated circuit

On 12 September 1958, in a laboratory at Texas Instruments in Dallas, engineer Jack Kilby demonstrated a working integrated circuit—a single piece of semiconductor material that contained multiple electronic components connected together. The device was crude by later standards: a sliver of germanium with tiny wires and a transistor, resistor and capacitor fabricated on the same substrate. Yet it proved that the separate, hand-wired components that filled earlier computers could be miniaturized and unified.

Kilby’s demonstration solved a growing crisis known as the “tyranny of numbers.” As electronic systems grew more complex, the sheer quantity of discrete parts and soldered joints made machines unreliable, bulky and prohibitively expensive. By fabricating the entire circuit in one solid block, Kilby showed a path to denser, cheaper and far more dependable electronics. Robert Noyce at Fairchild would soon refine the idea with the planar process, but Kilby’s September demonstration is universally recognized as the first working proof.

The intellectual thread from that afternoon runs straight into modern artificial intelligence. Integrated circuits became the silicon chips that power every computer. Exponential improvements in chip density—later summarized as Moore’s Law—turned room-sized machines into pocket devices and, crucially, into the massively parallel processors required for neural networks. Today’s GPUs and TPUs, which train large language models and vision systems, are direct descendants of Kilby’s insight that computation could be etched into solid material rather than assembled from discrete parts.

Without the integrated circuit, the data centers that host contemporary AI would be physically impossible; the energy, space and failure rates of vacuum-tube or discrete-transistor machines would have halted progress decades ago. Kilby’s quiet laboratory demo on 12 September 1958 therefore stands as one of the indispensable engineering foundations of the entire AI era—an anniversary worth remembering each year as algorithms grow ever more ambitious on the silicon he made possible.