Putting the A into AI

In a recent episode of Top Gear (S22E06), the hosts got going about driverless cars and how one day these cars are going to being making decisions about whether they should kill a bunch of kids or let you die instead. And it’s a funny little skit they do, but many people seem to believe that this is a current issue.

What I find surprising, is that no one is actually worried about these things, until someone starts a conversation about it, and then it’s fun to joke about, but in a serious manner. It’s the same thing we saw in the movie adaptation of iRobot, where Will Smith hates robots because one chose to save his life over a child’s, because the odds were better. It’s also why during the rest of the movie the robots try to imprison us, for the ‘better good’.

But it’s called Artificial Intelligence for a reason, mainly because it’s not real. It’s true that we’re doubling the processing capabilities of computers on a very frequent basis, but we’re still decades away from being able to get a computer to take in and process enough data to be able to make that kind of decision, and then it’s still only making a decision based on what we’ve taught it, there is no intelligence in the true sense of the word. Nevermind who gets to make those decisions anyway? Is there an ISO standard which dictates the hierarchy of people. For example: two 20 year olds > one 50 year old? No. Not any time soon anyway. And I don’t know if I want my car making these kinds of decisions in any case. I don’t want it to know that much.

The sad (maybe) truth is that it’s really really hard for a computer to quickly recognise different kinds of objects. It can quite possibly recognise obstructions, but to differentiate what these all are in any usable time frame is nigh impossible. More likely the car will see it’s about to crash, will slam on brakes and determine if there is an area to its left (for RHD cars) that it can swerve and avoid hitting something else. This is where we are currently. And it’s great, because as much as we as people can analyse, people make bad decisions. More often than not though, the car will avoid being in situations where it has to make this kind of decision. It won’t be tailgaiting like humans do, it’ll have a nice following distance that gives it time to react. And it can react quicker than us.

I’ve written about autonomous vehicles before, the ‘Yay’, but ‘ahh’ feelings I have towards them. The biggest factor against them at the moment is the lack of regulations governing them. And no one really knows how to react to this. There are too many uncontrolled variables, which I don’t see being solved in the next 10 years, not on public roads in any case.

Some of my posts (such as this one), I’ll have the idea for the post, sketch out one or two paragraphs, but then it sits as a draft for almost 3 months. And as time goes on, more and more people write and speak about the topic, such as the below 2 articles which have al come out in this time :) It is an exciting time to be in, and I look forward to seeing how the rate of penetration increases. I’m not as optimistic as others may be, but agree that it is the future, and hope to be able to be a part of it.

Elon Musk mentioned it in a recent key note address

Another very relevant link: Self-driving cars and the Trolley problem

Busy does not mean bad

Several years ago I read an article, can’t remember too much about it, but the one point that stuck with me is the generic response to “How are you?” is “Busy”. The author pointed out that that’s not a very good answer, and I’ve taken note of it every time I’ve given that answer since. A few weeks ago another article popped up on one of my feeds, with a similar message. And every time I give the answer (and I find myself doing it regularly) I kick myself, and then try expound.

But being busy isn’t a bad thing (I’ve even written about it before). The problem is when telling someone you’re busy, it doesn’t mean anything. Everyone’s busy, always. Except when you’re not, but that’s like never, right? I make time for the things I want to do (or feel compelled to do). It’s about prioritising. I do this at work all the time, however there it’s aligning your priorities with your manager’s.

Being ‘busy’ is an offhand comment we give to appease the masses. I don’t really care if you’re busy or not. I want to know if you’re well. If you’re having trouble. If you’re happy. If I can help. It’s become customary to accept being ‘busy’ as an answer; no one bats an eye. It’s like being fine. You’re as fine as you are busy.

If I’m busy I feel like I’m working effectively. And that kind of attitude carries over into my personal life. Don’t get me wrong, I spend a fair amount of time doing non-value adding things like browsing Reddit and watching Youtube videos, I have lots of time I could be spending ‘achieving’ something, but I still feel busy. As long as we’re not doing nothing, we’re busy. And so it adds to the lack of meaning in the word.

Are you busy because you have been given lots of work? Are you keeping yourself busy? Why are you busy?

So what is an acceptable way to reply? I get it, saying you’re busy or you’re fine has it’s places. It’s great in impersonal communication with acquaintances. You’ve only been asked out of politeness and custom, it’s a nice answer that doesn’t have to trigger an in depth discussion. But when it’s someone close to you, when it’s someone you care about, don’t say you’re busy, they don’t care if you’re busy.

DCIM101GOPROI really have no idea what pictures I can put in this post…

1996 Audi A4 (B5) 1.8

Today I sold my car. It was a 1996 Audi A4 1.8 (the naturally aspirated one). It was a great car that I am going to miss tremendously. Even though I’ve only driven it very rarely in the last year, every time I did I would remember why I loved it. It was voted SA Car of the year in 1996 and for good reason. Even though it’s 19 years old, the car still goes well, has many features, and in my opinion is still a good car by current day standards.

beach_audiThe car used to belong to my grandfather who looked after it with impeccable care. When he passed away my parents bought the car for me, and I’ve had it for the past 7 years. I received it at around 140,000km, and have put an extra 70,000km onto it since then. During this time it has only failed me once, and that was when the fuel pump packed in late at night on the way home from a show in Cape Town. Besides that I’ve only had to perform standard wear and tear replacements.

party_audiThe 1.8l engine put out 96kw when it was new, and has retained majority of that grunt. Although not over the top, it allows for comfortable cruising, and dropping a gear gives you all you need for overtaking or maintaining speed up a hill. The front-wheel drive, 5 speed manual car under-steered predictably, and minor tyre squeal from the 15″ tyres on stock mags would always let you know how close to the limit you were. Lean on hard-cornering was quite noticeable, but understandable. The car was absent of any ESP or traction control, but did have all-round disc brakes with ABS.

road_trip_audiThe radio was a standard cassette/radio, a CD shuttle in the boot was a later option which my car lacked. The radio only gave me hassles when the battery died and I didn’t know what the code was. I later also installed an RF transmitter into the dashboard to allow for some more modern audio input. The car originally unlocked with an IR remote. This had stopped working by the time I got the car, but the central working still functioned. I thus installed a garage remote receiver in the boot and hooked it up to the central locking system, thus returning remote locking to the car.

scenic_audiThe car is spacious with a large boot and even a full-sized spare wheel. The car had one airbag in the steering wheel for the driver, and other models came with a second for the passenger. 3 point seat belts for 2 of the 3 seats in the back, with a waist belt for the 3rd. The climate control still worked perfectly along with the rest of the electronics I haven’t already specified. In its latter years the catalytic converter was removed which gave the Audi a bit of a raspy warble.

nowhere_audiIt’s a great car, and can be picked up relatively cheap at current rates. Condition of the vehicle will vary a fair amount, and if proper maintenance hasn’t been seen to, you could be in for costly repairs. At 200,000km the clutch and gearbox still worked perfectly, and the engine wasn’t burning or leaking any oil.

New backwards compatible USB cable

Why hasn’t there been a USB cable developed that is reversible like Apple’s lightning connector? This is something that has bothered me for a while. There are a lot of jokes around the USB connector, my favourite being this one from SMBC.

20111004

But why can’t we have a reversible one. So I did a speed run through the 56 MB of documentation on USB 2.0, and can’t see any specific reason that this can’t be done. The only reason I’ve come up with is that there is a risk of shorting, especially if someone doesn’t conform to the standard. I also read on StackExchange, one person’s answer relating to USB specifications that the USB logo must be displayed on the top of the connector, something I feel can be easily overcome.

I have seen a prototype demonstrated before, and while nice, I don’t think it’s the simplest solution to the problem. I believe this solution makes use of a similar mechanism. Also I really feel a plug shouldn’t have moving parts.

So that leaves us with this:

I decided to test out Sketchup, having being spoilt with student version’s of Autodesk software throughout university, it was something new to try out. So excuse the poor animation and model.

2015-05-02 17.16.28

In any case, I took an old USB cable and shaved off the casing to test out the feasibility of this option. And my initial thought played true. If you have contacts on both sides of the plug, on the side not connecting to the USB port, the contacts short with the springs in the port housing, which in practice would result in an overload warning from your computer, and the USB port shutting down.

2015-05-02 17.15.22

So, I can see why no one has done this yet, but I still think there’s a better solution than what we currently have. And I’m not convinced that my method, with some minor adjustments, can’t work.

Some funny things I found in the standards document include maximum insertion and extraction forces, of 35 N at 12.5mm/min and 10 N at 12.5mm/min respectively. Also a cable pull-out specification, because who is going to try grip a plug, when you can just tug on the wire :)

note: this post was started in June 2014, but I’ve only now made time to tie everything together and publish, in May 2015