fbpx

Human Progress Logo Human Progress

01 / 05
When the Mac Was a Munition

Blog Post | Computing

When the Mac Was a Munition

The US government classed the Power Mac G4 as a super-computer

Louis Anslow —

This article was published at Pessimists Archive on 1/23/2024.

This week the Macintosh turned 40

In 1999 Apple introduced the Power Mac G4, a ground breaking Personal Computer, not aesthetically – like the iMac before it – but mathematically: it had 1 GIGAFLOP of computing power.

Image of Apple Power Mac G4, a personal computer

Crossing this threshold meant the US government classified the device as a munition and potential weapon of war, and so was subject to strict export controls. The cold war era 1979 Export Administration Act that set these rules, was written to prevent powerful computers falling into communist hands.

Article talking about American technology falling into communist hands

A GIGAFLOP in 1979 was cutting edge, powerful and scary – unlike today when we carry around thousands of GIGAFLOPs in our pockets: the iPhone 15 Pro has 2000 GIGAFLOPs, while iPhone 12 has 1000.

In a classic example of Apple’s marketing genius the munition classification was used in its marketing, it dubbed the G4 a ‘Personal Supercomputer’ and ran a TV ad featuring tanks surrounding the device. Steve Jobs would say at its unveiling that they got the idea after being informed of the restrictions by the US Government government officials.

Gif of an Apple commercial based on the Power Mac G4
(Watch the full ad here)

The narration begins:

For the first time in history, a personal computer has been classified as a weapon by the US government . . . the Pentagon wants to ensure that the new Power Macintosh G4 does not fall into the wrong hands.

The punchline?

As for Pentium PCs, well, they’re harmless.

Companies lobbied for the out-dated law to be changed as Moores Law saw US computer companies rapidly approach crippling export restrictions on products. (This mimicking a similar parallel battle about encryption at the time.) And they succeeded. The truth was, the GIGAFLOP threshold had been increased by President Clinton in prior months – but was not set to go into effect until January 2000 – until then Apple had export restrictions, but did have a headline grabbing marketing gimmick.

Image of an article from CNN talking about Apple's attempt to get the export restrictions on their product lifted

The Power Mac G4 is so fast that it is classified as a supercomputer by the U.S. government, and we are prohibited from exporting it to over 50 nations worldwide.

Steve Jobs

This intriguing chapter in computing history is pertinent today, particularly as calls to regulate AI and once again treat the computing power used to train them as munitions, grows louder.

In 1979 1 GIGAFLOP was deemed weapons-grade computational power. By 2000 it was in a consumer product – but restricted by an outdated law. Now, Apple’s M3 Mac – boasting 4000 GIGAFLOPs – is exported globally.

In 2023, President Biden’s AI executive order introduced a new benchmark for reporting AI trained on hardware exceeding “1e26 floating point operations or 1e23 integer operations.” Perhaps we’ll look back in 50 years and laugh how we treated once impressive levels of computing power as threatening.

Assuming AI hasn’t killed us all by then.

Reuters | Science & Technology

Microsoft Readies New AI Model to Compete with Google, OpenAI

“Microsoft is training a new, in-house AI language model large enough to compete with those from Alphabet’s Google and OpenAI, the Information reported on Monday.

The new model, internally referred to as MAI-1, is being overseen by recently hired Mustafa Suleyman, the Google DeepMind co-founder and former CEO of AI startup Inflection, the report said, citing two Microsoft employees with knowledge of the effort…

MAI-1 will have roughly 500 billion parameters, the report said, while OpenAI’s GPT-4 is reported to have one trillion parameters and Phi-3 mini measures 3.8 billion parameters.”

From Reuters.

Phys.org | Computing

First Experimental Proof for Brain-like Computer with Water and Salt

“Theoretical physicists at Utrecht University, together with experimental physicists at Sogang University in South Korea, have succeeded in building an artificial synapse. This synapse works with water and salt and provides the first evidence that a system using the same medium as our brains can process complex information.”

From Phys.org.

Live Science | Computing

Intel Unveils Largest-Ever AI “Neuromorphic Computer”

“Scientists at Intel have built the world’s largest neuromorphic computer, or one designed and structured to mimic the human brain. The company hopes it will support future artificial intelligence (AI) research.

The machine, dubbed ‘Hala Point’ can perform AI workloads 50 times faster and use 100 times less energy than conventional computing systems that use central processing units (CPUs) and graphics processing units (GPUs), Intel representatives said in a statement. These figures are based on findings uploaded March 18 to the preprint server IEEE Explore, which have not been peer-reviewed.”

From Live Science.