Weapons of hyperscale

Yesterday I wrote about the compromise of German government networks, an attack that seems to be the worst assumable accident for anyone thinking about cybersecurity and nation states. There's no such thing as secrets anymore. Unwittingly, Julian has won. 

Anne Currie, a self-avowed "greybeard" programmer wrote a related article in The Reg yesterday: Ethics? Yeah, that's great, but do they scale?

Her point is simple but powerful: off-the-shelf tooling for continuous integration/deployment and infrastructure-as-a-service have boosted "time to scale" by multiple orders of magnitude. This is, of course, great for companies that make stuff that we want. They'll improve it more quickly and get it to us instantly. But "we're getting close to the point where almost anyone could potentially affect the behaviour of a significant proportion of humanity." Great quote (I suggest removing "almost" and "potentially").

I think she's right. These tools are open for anyone. Before the last election I used to joke that if you give me a sufficiently large online advertising budget, I could swing any European election. I think that's still true. 

I'm not sure how often you interacted with companies today whose primary objective is to change your behaviour so that you stay longer, click more, consume more. I'd bet ten times or more since you got up this morning.

If you value freedom, that's a problem.

P.S. Case in point: the DDOS attack on Github. Look ma, no botnet!