Google Engineers Refused to Build Security Tool to Win Military Contracts

Many promising young researchers from around the world earn their stripes at US universities, and loads of them would be happy to stay and become Americans. Unfortunately, the US immigration system makes that difficult and lots of them get sent home.

The best part about that plan is they don't even have to develop their own AI weapons. They could deny that capability to adversaries just by allowing those workers to take up interesting & lucrative jobs in the American civilian economy.

https://www.bloomberg.com/view/articles/2018-05-31/u-s-shoul...

This question is an oxymoron.

Both the definition of open source (by the Open Source Initiative) and the definition of free software (by the Free Software Foundation) explicitly forbid discriminating against any person or a group of people for software to be considered as open source / free software.

Picture this: we're deep in the Cold War era, and Ronny Raygun is the president. The Star Wars initiative is in full swing. Nuclear mishaps and near-misses are still an all-too-frequent occurrence. Nena's red balloons are ready to fly. The President has just gone on the air and promised to "Outlaw Russia," and that "The bombing starts in five minutes." Some officers from the Pentagon approach your company's management and ask them to bid on a project that will involve launching dozens of military satellites equipped with atomic clocks, designed to locate any person, place, or thing on Earth at meter-scale accuracy.

Do you threaten to quit the company, lecture everyone who will listen about the potential for misuse of such dangerous technology... and miss out on the chance to contribute to GPS?

Or, how about a few years earlier, when they ask you to work on a colossal air-defense computer system using technology that hasn't even been invented yet. They want to use computers to vector fighters to intercept approaching Soviet bombers, and to target missiles to kill the ones the fighters miss. It's obvious to anyone not in uniform that the whole idea is insanely stupid, wasteful, and destabilizing. They want to spend $50 billion to develop a system that guarantees one whole 'nine' of uptime, that can be trivially jammed or disrupted, and that will literally be obsolete the day they turn it on since it can't deal with ballistic missiles.

So you tell them, "No, I don't want to work on SAGE. Peace out and fuck off." As it turns out, the system is indeed useless at its stated purpose. The spinoffs, on the other hand, not only revolutionize the entire IT industry, but create it. The next thirty years' worth of progress in computer architecture, memory and storage, networking, display and human interaction, and fault-tolerance is traceable directly to this boondoggle. All of these things would probably have taken decades longer to evolve without military funding and backing. Nobody dies as a result, but a bunch of your coworkers do get to play with a lot of cool toys.

Then there's the whole ARPAnet thing, funded by DoD to survive nuclear holocaust, asteroid strikes, plague, fire, and famine... any threat imaginable by anyone other than telecom-industry lobbyists.

Even the very wise cannot see all ends, and that goes for modern-day hippies too.

What kind of logic is that?

We don't necessarily need to match lock-step with terrible regimes to ensure our own security, and our own security is the only argument for matching them in the military field.

For example, if someone else uses gas or landmines, we can use masks and airplanes to defeat them, without having to resort to those things that can't pick their targets, and remain on the field, where children play, decades after the war ends.