Megahacks, a review of “backhaul” and robot ethics

A week is a long time on the Internet

Last week we spotted a story saying that given current trends we were on track for 1 Billion account compromises in 2016...

Nek Minute! Yahoo announces that back in 2014 they lost at least half that many in a single breach!  

Even with daylight saving playing havoc with your internal body clocks, the astute of you will no doubt have noticed that it’s not 2014 anymore. It is in fact only 12 weeks until Christmas in 2016.

So there’s a couple of questions here. Did Yahoo know about this breach and choose not to notify users for two years? Or were they just unaware that such a massive amount of corporate data had been exfiltrated from their network? Either of those situations is a pretty big #FAIL in our books.

And we're not alone in thinking this. It looks like Yahoo is being taken to court in the US for showing "reckless disregard for the security of its users' personal information that it promised to protect." There are also some calls for the US Government to launch an investigation into the breach.

Why should you care? Apart from the fact that it’s terrifying how large this breach is (it’s like losing the details of EVERYONE in the European Union, 509,668,361 people), it might affect you too.

Spark is advising customers to change their passwords, because for some years now, they have been using Yahoo to provide email infrastructure to customers. Read more here.

All in all, it’s a bad look. Breaches aren’t getting smaller. It looks like our Deputy Chief Executive Andrew Cushen was right on track when he mentioned them in his 10 predictions for 2016...

The Internet’s on-ramp and regulatory traffic cops

We can think about New Zealand’s Internet service like a transport system - but instead of moving people and stuff around, it moves data. If you ever get stuck in traffic, it’s pretty clear that the experience of the network depends on how fast each bit goes. Like our roads, the network has local bits, like the streets or lanes near where you live and work. There’s higher-capacity core bits, like motorways, that join up these local bits into a national network. And then there’s the “on-ramps” which join your local service to the core network. That “on-ramp” is called backhaul, and it’s pretty important - a 10 lane hyper-expressway isn’t much use if you can’t get to it!

We have joined our friends at TUANZ to welcome a study of backhaul by the Commerce Commission (our watchdog for telecommunications and competition stuff). The last look at these issues was in 2008, so it’s about time. Since then, we’ve had a huge chunk of the UFB fibre rollout, and continuing increases in wireless speeds. It’s important to check that backhaul is fast enough, and is being supplied on fair terms to enable the potential of these faster networks.

Speaking of watchdogs…

We’ve seen a couple of regulatory growlings against Internet providers lately. First a local story - Trustpower’s been fined $390,000 for misleading pricing on its Internet and power packages. Lesson: don’t advertise “$49 a month” in massive writing, when the price is going to increase for the second year of a two-year contract. Secondly, following work by UK regulator Ofcom, Sky UK is facing a multi-million pound fine. The problem there is that they might have made it hard for customers to jump ship. That’s pretty serious if true - blocking customer choices is a big no-no for good, fair Internet services.

Robot ethics

A few months ago we spoke about when the bomb robot blew up the Dallas shooter. We said at the time that it might be a good idea to look at some ethical guidelines when it comes to designing and deploying robots. Well, turns out someone did.

The official standards body for the UK - the British Standards Institute - has developed a standard for “ethical design and application of robots and robotic systems.” This is pretty encouraging to see something dealing with ethical issues in advance, rather than waiting for problems to emerge. The authors are “scientists, academics, ethicists, philosophers and users” which sounds pretty good - issues in this area are broad and complex, so a range of views are needed. Unfortunately, even an electronic copy of the standard costs £158, which is money your average home garage robo-hacker would probably rather spend on parts. We’ll be carrying a hat around later to see if we can afford a copy.

Until then, the favourite discussion topic for robot ethics is self-driving cars. Cars are dangerous, as we can see in our road-death rates. By drawing on massive datasets about driving situations, and super-human robot reflexes, we can probably make self-driving cars way safer than human-drivers. Even so, robo-cars will be involved in accidents. In those situations, philosopher types love to debate whether cars should swerve to save the highest number of lives, or loyally protect their own passengers regardless.

Well, some of the people who make these cars don’t debate that at all. They have a simple answer: “slam on the brakes.” And also, if you’re asking who to save, you probably mucked up earlier - add that one to the dataset so every other similar car doesn’t make the same mistake. We say that sounds good, but we’re not going to give up our philosophy tea-breaks just yet. If you want to do the same, you can play with MIT’s online tool for these scenarios.

Speaking of robots…

Mr Robot actor Rami Malek wins the Emmy for Outstanding Lead Actor.