Posts Tagged: Apple

“Foolproof, and incapable of error”

Saturday 21st January 2012

Whilst I was thinking about and writing the previous post, a couple of extra things came to mind which I couldn’t really fit into the post. So I thought I might as well do a follow-up with a couple of extra observations. I did intend to write this earlier, but partly I was busy (more on that in a following post) and mostly I just didn’t get around to it.

1) The previous post was not so much about aeroplanes, but more about interfaces in general. Be that with machinery like a plane, or a device like a phone, or even infrastructure or services. And it struck me that one of the few organisations that consistently manages to create things with great interfaces is Apple. Not so much with their computers (I’m really not a big fan of MacOS, probably because I’m more used to Windows), but their iOS devices (iPhones and iPads) are really good examples of things which simplify tasks through good interface design.

It strikes me that if the computing business ever starts to go slack (!), Apple could do a good business out of consultancy; imagine if they applied their UI design skills to things other than making iPhones and iPads. This isn’t as daft as it sounds; some ex-Apple employees recently set up a business to make a better thermostat. That’s a specific example of someone applying the Apple approach to interfaces to a different type of product, and I’m sure there are other things which would benefit from the same approach.

2) For some reason, I also started thinking about 2001: A Space Odyssey (spoilers follow. Although, it’s a 40-odd year old book/film, so I guess most people at least vaguely know the plot. If you don’t, then go read the book and watch the film. They’re classics). The first – obvious – point is that a lot of the interfaces in that film do appear to tend towards simplicity. There’s loads of little things: the video phone booth that Dr Floyd uses near the start of the film, the tablets that Bowman and Poole use on Discovery, all the spaceship status screens look like they’re intended to be simple, and of course there’s HAL9000

On the topic of HAL, it occurred that his demise is pretty relevant too. HAL was programmed to help the crew, to convey information to them about Discovery and about the status of the mission. But before the crew left Earth the parameters of the mission were changed; this was secret, and the crew were not to be told until Discovery reached Jupiter. As the central computer, HAL knew the real purpose of the mission, but was not allowed to tell the crew. He was being asked to hide information, to lie. This ran counter to HAL’s programming – he was designed to give information, not to hide it – and because of that conflict he perceived there to be a problem. Which he then set out to rectify…

The point is, HAL failed because the people who defined his tasks for the mission did so incorrectly. The computer carried out its tasks as it saw best, but those tasks were in conflict with each other. And so the failure of the mission was the result of misuse of the computer. Now obviously the details in this and in the example in the previous post are very different, but in general, it’s the same fault: the computers behaved exactly as they were asked, the error arose from the way people were trying to use them.

And, really, how clever is that? That 40 years ago, people were thinking about how we’ll be using these ultra-sophisticated computers, and were (in a very broad sense) predicting some of the problems that we’re starting to see. Just makes me realise how great a job Clarke (and Kubrick, I think) did in writing that story, and how many ideas they’ve managed to pack into it. I’ve read the book many times already, but I really need to re-watch the film.

Posted In: GeekMoviesTechnology Tagged: | No Comments


Thursday 28th April 2011

I’m a little late to the party on this, but there’s been lots of coverage lately about iPhones “tracking” people. The original script that was released to look at this data was Mac-only, and at the time I couldn’t be bothered to hack it to work with Windows machines, so I couldn’t look at my own records. Happily though, other people have done it instead, so I’ve been able to have a look at the data from my own iPhone.

The above image shows every data point in the database, since I first got my iPhone last year. And from that, it looks like there’s some quite detailed data there. Certainly, from memory, the above map doesn’t seem to be lacking anything (although even at that level, I can see some anomalies, places I haven’t visited).

There are obvious concerns that arise from the concept of your phone logging everywhere you go. Clearly, from the data you can get a rough idea of someone’s movements. Each of those data points has a date and time associated with it, so from that you could build up a rough idea of where in the country I was on a given date. But can you tell more than that?

This all seems to come from WiFi and phone masts, so I assume that the most accurate information (in terms of location and time) would be in places with the most WiFi hotspots and phone masts. You can see from the image above that I’ve visited London in the last year. I went there for a day last August, so I thought I’d look to see how accurate the log is.

Now, London probably has the highest concentration of phone masts and hotspots in the country, so I assumed that the “tracking” data would be pretty accurate. In fact, it’s quite mixed. Right off the bat, there are two clumps of data points in areas that I didn’t visit (around South Kensington, and around Shoreditch). Additionally, there’s quite a spread of data points on the top-left of the map. If you look closely, they vaguely (very vaguely) track the train line that I used to travel in and out of London; and in fact the timings do coincide with those journeys.

The stuff in the middle seems to be more accurate. I had an interview that day at the Royal Academy of Engineering office, which is near to St James’ Park; there are a clump of points around that area, because I got there early and so walked around the park for a little while. I also walked upto the British Museum from there, and as you can see there’s a big trail in that general direction. During the day, I also spent some time in Hyde Park (time to kill…), so I find it quite interesting that there’s very little in that general area.

Now, looking at the data, I think it’d be pretty hard to work out exactly where I went, without already knowing it beforehand. Many of the data points have the same timestamp (I assume the phone fetches a bunch at the same time), and as you can see there are more than a few in places that I didn’t actually visit.

To expand on this, here’s a similar map showing a visit to Cardiff, last month.

In this case, I’d driven down to Cardiff to pick up my younger brother, and was there for about half an hour. The exact journey I took was West along Eastern Avenue (near the top of the map), south-east down Whitchurch Road and Crwys Road, and then onto the street my brother lives on; and then the same route in reverse about half an hour later. Looking at the data from my iPhone, it’s pretty much impossible to tell precisely where I went.

I’m not saying this data couldn’t be useful. It can paint a broad picture of where someone has been, and if that formed part of a larger investigation into the person then that information could be invaluable. But to gain access to this requires access to the phone, or the computer it syncs with, which isn’t always possible. And, as I say, it’s only a very broad picture. Looking at all of the data from my phone, there is no way that you could even work out where I live, for instance. If someone got hold of my phone and took this information from it, then it’d be fairly useless to them.

When details about this first surfaced, I was a little unnerved about it. However, having looked at it and seen the level of inaccuracy, I think that much of the negative publicity that Apple has received has been a little over egged. It’s a minor concern, especially given that mobile phone networks already record locational information as part of their service (and I’d bet it’s a fair deal more accurate than what’s stored on the iPhone). It’s a minor (and interesting!) mistake, which Apple are now fixing.

Posted In: Technology Tagged: | 2 Comments