Week 15, 2016 - Connected devices; cars; rockets; encryption

By (8 minutes read)

Connected devices don’t do a lot when their servers are shut down, Tesla’s new car, encryption, and rocket landings.

When the IoT fails you

Smart devices of all sorts are heading our way, and while I mentioned a couple weeks ago how great the Amazon Echo seems to be this week brings a look at what happens when a product fails.

Nest, an Alphabet1 company, is closing down one of its products. Specifically, a home automation device that manages other devices. This is of course a decision they can make, but it brings about a situation similar to when Mailbox was shut down. Except, this time it’s not a free app that stops working, but an expensive device that becomes a paperweight. This isn’t the kind of behavior that inspires trust in Nest, and I think it really is a bad move towards their customers.

Instead of focusing on how terrible this is though, let’s have a look at how this came to be. Now, there are some signs that Nest isn’t in the best situation they’ve ever been right now, despite having built (and bought) several great products in the past. I don’t want to wade into the debate of whether Nest is really in big trouble (which I believe isn’t a major problem as long as Alphabet’s top trusts them), and I think this move is more emblematic of the entire Internet of Things or connected devices industry.

In many ways, that whole industry is currently in the “let’s throw things at the wall and see what sticks” stage of its life2. Compared to traditional computing devices, IoT devices are cheap. There are quite a few exceptions to this, and some of those exceptions are ridiculous, but it generally holds true.

These devices often have a single function, and at a time when you can get Raspberry Pie type hardware for the price of a coffee that means there really isn’t a big investment that needs to be made at the hardware level. Don’t get me wrong, I’m not saying it’s easy to design them, let alone design them well, and scaling production to meaningful numbers is a completely separate issue.

But if you combine the low hardware costs with cloud based infrastructure you can see how the costs for building something like this are minimal. But minimal is not nothing, and over time the fixed costs will stay up while your product might not be selling much anymore. And at that point a manufacturer needs to decide whether they want to support their loyal customers or cut their losses. Obviously, as it depends on those services staying up, in many cases this will mean that a device you bought will no longer work. Whether that’s your button for ordering new laundry powder, your thermostat, or any Kickstarter project you thought had great potential.

The thing with throwing stuff at the wall is that most thing don’t stick. And if you’re an early adopter that means you’ll get bitten by these failures3. That is sad, both for the people who created those things as well as for the ones who bought them, but there is also not much you can do about it. Everyone tries to lock in a part of the market, but at some point this will probably stabilize. When that happens a lot of these devices will become interchangeable and use common APIs.

Until then, using any of these devices remains a bit of a gamble. It’s just that what they promise is so compelling that you can’t stop buying them anyway.

The new Tesla 3

I’m not much of a car person, I neither own one nor feel like I should. That doesn’t stop me from being interested in Tesla though, so I did pay attention to Elon Musk’s announcement of their new Model 3.

The reason I’m interested in Tesla is because they’re currently at the forefront of the electric car market, and that is where cars seem to be headed. For better or worse, cars aren’t going away any time soon and while I’d love it if self-driving cars showed up soon that doesn’t seem to be happening either.

In the electric car market Tesla appears to be ahead of every other car manufacturer4, and by a pretty decent margin. Whether they stay ahead, let alone with a big margin, is something that we’ll have to find out when the traditional car manufacturers seriously start producing electric cars. In the meantime, Tesla has a proven track record and therefore has a head start.

So far all of their cars have been expensive, but the price has gone down quite a bit for the new model5 as they want to take on the mass market. Combined with Tesla’s good reputation this means people were excited.

That excitement was clearly proven by the fact that over 110.000 people throughout the world preordered the new model before it was even announced. And since then at least another 200.000 have joined them with a preorder. Just to be clear, these preorders aren’t guaranteed to become actual orders, but even so it’s a huge number for a product that you won’t be able to get for at least another year.

And it’s at least another year, as last year Tesla only built about 50.000 cars. For everyone who ordered a model 3, I really hope that they’ll be able to ramp that speed up or it’s going to be a long wait if you’re at the end of that queue.

As usual, Ben Thompson has a far better write up of what this all means over at Stratechery, but I also want to throw in one other thought.

Above I talked about the inherent problem with IoT devices. While Tesla’s cars aren’t quite like that, a big part of their value is derived from what else Tesla offers. This includes their software updates, but also their supercharger stations. At the moment Tesla is spending a lot of money on their R&D and infrastructure, meaning they’re not profitable. While I don’t see them being unable to raise funds when they require it, what would happen if they have to close down?

In some ways, not much, but it would be harder to charge your car (at least take longer) when traveling. Maintenance of the car would become a problem. It doesn’t need as much maintenance because the hardware is more robust with less moving parts, but if it’s needed who can do it? I think it’s interesting to consider these things, even when it’s an unlikely scenario. With most other car brands this wouldn’t be an actual problem if they go out of business. At least until there are a lot more electric cars on the road.


This wasn’t Elon Musk’s only success either this week. SpaceX, his rocket company (that definitely isn’t using electric engines), managed to land their latest rocket on a drone ship as shown in the embedded video below. I don’t have much to say about this, other than that it’s a great achievement.

And shortly after that, Blue Origin (owned by Jeff Bezos) showed once again that their reusable rocket strategy is working as well. Again, great to see this.

These kinds of successes make me believe that private space flight is moving along quite nicely. It’s an incredible thing, and I am certainly grateful to see that these people are spending so much of their life (the engineers) and their money (the financers) to make it happen.

Whatsapp encryption

Whatsapp turned on end-to-end encryption last week. Or rather, they turned it on the week before and only announced it last week.

From a security point of view this is a major step especially because of the number of people using the app. With over a billion people across pretty much every type of phone (but still not on my iPad) this means so much more messages are secure.

While iMessage is already pretty secure (and after the FBI dealings recently, I suspect that in June we’ll hear about improvements to that as well), Whatsapp surpasses that according to the EFF Secure Messaging Scorecard. Ticking boxes there isn’t necessarily the thing that counts the most, but it helps. Either way, it’s good to see security increasing in a time where it’s under attack from governments.

  1. I’m almost at the point where saying Alphabet instead of Google comes naturally. ↩︎

  2. A technique that for example Google and Amazon use a lot. ↩︎

  3. Yes, I too have spent a lot of money in the past on stuff that turned out not be even remotely useful or in too early a stage of its life. ↩︎

  4. By which I mean a combination of technology, manufacturing ability, and how well they are known. Which discounts the possibility if someone has produced a car that is better in every way, but isn’t actually available to consumers or nobody knows about it. ↩︎

  5. It’s not cheap, but a lot more affordable. ↩︎

comments powered by Disqus