Why the Abbott government hasn’t abandoned their incredibly unpopular metadata policy yet is beyond me. Nearly all other developed nations that have pursued such a policy have abandoned it, mostly because attempting to pass something like this is akin to committing political suicide. Worse still in their attempts to defend the policy from its critics the Abbott government has resorted to tactics and sensationalist rhetoric, none of which has any bearing on the underlying issues that this policy faces. Top this off with a cost estimation that seems to be based on back of the napkin math and you’ve got a recipe for bad legislation that will likely be implemented poorly and at a great cost to all Australian citizens.
Conceptually the idea is simple: the government wants to mandate that all ISPs and communications providers keep all metadata they generate for a period of 2 years. Initially this was sold as not being an increase in the power that authorities had however that idea is incredibly misleading as it greatly increases their ability to exercise that power. Worse still obtaining access to metadata doesn’t require a warrant and isn’t just the realm of law enforcement or intelligence agencies as people on local councils can obtain this data. Suffice to say that the gathering and retention of this data is a massive invasion of the privacy that the general public expects to have from its government and that is exactly why nearly all developed nations have dropped such policies before they’ve been implemented.
As expected the usual tropes for these kinds of policies have been trotted out, initially under the guise of a requirement for national security. I’d concede that point if it wasn’t for the fact that mass surveillance has not proved to be effective in combating terrorism, something which the critics of the policy were quick to point out. The rhetoric has then shifted away from national security to local security with Abbott saying that the metadata will help them track down peadophiles and child traffickers. Suffice to say if surveillance of this nature doesn’t help at a national level then I highly doubt its effectiveness at the lower levels and “think of the children” arguments like this are nothing more than an appeal to emotion.
Yesterday Abbott was pressed to give some hard figures on just how much this scheme would end up costing and he retorted with the rather ineloquent quip that it would be an “explosion in an unsolved crime“. When pressed the figure he gave was $300 million, estimated to be less than 1% of the total $40 billion that the entire telecommunications sector is estimated to be worth. That figure has apparently been sourced from PricewaterhouseCoopers (PwC) however the details of that figure have not been made public. In all honesty I cannot see how that figure can be accurate given the amount of data we’re talking about and the retention times required.
To put it in perspective Australians consumed something on the order of 1 Exabyte in 6 months up to June last year which is a 50% increase on the year previous. The amount of metadata on that data would be a fraction of that and, taking the same 1% liberty that Abbott seems intent on using, you get something like 50 Petabytes worth of storage required. Couple that with the fact that it won’t be stored in one place (negating economies of scale), the infrastructure requirements to provide access to it and the personnel required to fullfil requests and that $300 million figure starts to look quite shakey. Indeed the Communications Alliance in Australia has estimated it to be between $500 million and $700 million which casts doubt over how accurate Abbott’s lowball figure is.
Honestly this legislation stinks no matter which way you cut it and the rhetoric that the incumbent government has been using to defend it speaks directly to that. These policies are just simply not effective in what they set out to achieve and the only tangible result we’ll ever see from them will be an increased cost to accessing the Internet and a reduction in the expectation of privacy. I do hope Abbott keeps harping on about it though as the more he talks the more it seems likely that we’ll be able to cement the One Term Tony phrase in the history books.
One memory that has always remained clear in my mind was from the first few weeks of engineering degree. Sitting in an introductory electronics class the lecturer began by throwing some rudimentary math our way and lamented how we were quick to grab our pens and paper or calculators to work out the exact answer. It was a clever way of showing us how much we relied on technology to do our thinking for us and so the next week was spent teaching us methods of estimating values, working out rough solutions to equations and educating us in the ways of solid guesswork. After a while it became second nature to us engineers (all 10 of us!) and I don’t think any good engineer would be without it.
I hadn’t really thought about it until I was dicussing LED backlit TVs with some of my friends over lunch. The basic premise of the technology (I’m not going to talk about edge LED backlighting, that’s cheating) is that instead of using what amounts to giant fluro tubes to light up the screen for each individual pixel you use either a single white LED or a RGB LED array. I imagined what the sales pitch to the higher ups must’ve been when they were trying to make a full 1080p display, saying that the amount of LEDs would have been huge. One of my friends then whipped out his phone to calculate it (instantly bringing back the university memories) and instantly the engineering estimator kicked in and I came out with about 6 million LEDs to make up the panel. The actual figure is 6220800 (1920 x 1080 x 3), which made me feel like I’d earned a fair whack of geek cred for being able to guess that close without a calculator.
It’s not just party tricks like doing insanely large multiplcations in your head that estimation is good for. Most of the time I spent in the labs at university were trying to get some electronic widget to give the correct output. Now all the equations you’re taught are based on perfect models, so you’re never going to get exactly where you want. That’s where estimation comes in handy, if you know your inputs and can hazard a rough guess at the outputs it can save you hours sitting down and working out the actual output. If everything comes out inline with your expectations then you can go back and verify your results using the equations, otherwise you know you have to rework your experiment.
Although the real scientists would argue that’s what research assistances are for 😉
In the real world where project managers and higher ups demand estimates to ensure resources are allocated appropriately being able to come up with figures quickly is one skill that’s saved me countless times over. It’s one of those things that once you learn you never really think about again as the answers just start popping into your head, kind of like muscle memory but in your brain. Plus being able to do large multiplications in your head is a sure fire way to get all the ladies.
Well, that’s what my lecturer told me anyway 🙂