[insert witty title]

This is really amazing, and also really creepy

I can’t even begin to imagine how long it took to make this using only a digital camera

Ok so this guide is perhaps the most important thing I have ever written, and contains information that can change pretty much any life for the better. How to dress as a robot.

Here is what you need to look like a terrifying metal killing machine:

• One big box (body)
• One small box (head)
• A lot of aluminium foil
• Packing tape
• Two CDs

The pictures should be pretty much self explanatory, just cut some holes in the boxes and wrap them in aluminium foil.

That’s it, all done! You are now (if you weren’t already) a soulless killing machine.

This particular costume was left at a friend’s house after a Halloween party. A few days later he held a second Halloween party for his 4 year old son and some of the neighbourhood children. To get the kids excited about the celebration he wore the robot suit to collect his son from nursery that afternoon. As expected the kids all loved it and freaked out. However, in the evening one of the parents arrived at the party without his daughter because she was “too scared to come, because of the robot”

Hells yeah!

When you want to present a message, the first thing people see, before even the words themselves is how they look. Consider, for example the following picture. Imagine it’s part of an advert for some clothes. What sort of shop would sell them? Would you expect the salesperson to be young or old?

Consider also a second example, our good old friend Comic Sans. The following sign is, to my eye, absolutely terrifying and I would not set foot in here in a million years:

also I would just be plain confused by:

I think it’s an uncontroversial fact that the appearance of a piece of advertising (be it billboard, website or powerpoint presentation) has a great effect on how we interpret the message.

A while ago I heard about a documentary called Helvetica, which explores where Helvetica came from, and how it is used today. On the face of it this sounds like the most mind-numbingly dull bit of TV ever, but I actually found Helvetica to be incredibly engaging, and it managed to weave an interesting and understandable story. After watching the documentary I noticed that Helvetica is everywhere, from warning signs to shops. If you keep your eyes open, it’s almost like a secret code (”Hello. We want to look modern and/or unobtrusive”). Here are a few photos I grabbed from the web in about five minutes:

Helvetica clearly and unobtrusively gives us information and commands. We’re surrounded by Helvetica to the point that we don’t even notice it any more. Below I have collected together a couple of clips from YouTube, showcasing Helvetica. Firstly, here is a short and pretty but completely uninformative trailer:

^{The font used throughout this trailer is not Helvetic, but Arial, Microsoft’s Helvetica-a-like font designed to avoid paying licencing fees. I guess this qualifies as probably the most subtle joke I’ve ever picked up on}

And here is a very nice two minute clip from the film itself. The two designers who are interviewed do a really good job of explaining how the choice of typeface can drastically change how people are going to perceive a message:

Watching this film has made me actually pay a bit of attention to typefaces in the world around me, and upon reflection I realize that somewhere along the line my mind took a real liking to the ‘Helvetica style’. To me seeing facts presented in this way brings to mind the sleek, clean lines of a Mac (where Helvetica is actually the default typeface in a lot of programs) and the cheap efficient design of Ikea (where Helvetica is used in almost everything). Grey Helvetica on a white background is pretty much the de-facto “clean, modern” look at the moment.

I would recommend Helvetica to anybody who has even a passing interest in aesthetics, it really made me think about how our emotional responses are tied to how information is presented, and taught me that when I give a powerpoint presenatation I can’t go far wrong by using Helvetica.

Mostly, though, I am glad that in watching this documentary I finally had chance to think through and articulate exactly why Comic Sans is a pile of shit.

edit: nearly forgot: http://www.helveticamovie.com

A few weeks ago now, the following excellent comic appeared on xkcd:

With that in mind note that none of the following are photoshopped, but they all seriously look like they are:

I was originally quite surprised that a Chinook helicopter could do this.

Then I realised, any helicopter can do this. Once.

This recent Intel advert has some pretty sinister racist undertones (overtones):

All I see in here is a white overlord standing over subservient black men. I can almost imagine the conversations that led to this:

Tech guy: “OK so the Intel Core 2 Duo is so powerful it runs white hot”
Advertising guy: “Got it. Intel Core 2 Duo. White… Power”

Just thought I’d keep everybody up to date with progress on the twitter haiku generator.

• The code is now stable (it ran overnight last night)
• I have decreased the number of false positives, although will not be able to completely eliminate them due to the bewildering number of abbreviations, foreign languages and typos that people use
• The code can now deal properly with a few common abbreviations: OMG (3 syllables), WTF (3 syllables), SMS (3 syllables). I’m still not sure if LOL should be one or three
• The code now outputs an html file with links to authors

Once I have built in code for the script to automatically upload the haiku to geometricrate.com I should be able to leave it running and generating a permanant stream of beautiful poetry for the world to see! Last night (actually just from midnight until 4am because I limited the number of requests it could make to Twitter in case the code went crazy) the script found over 160 haiku. They are all in this file. Here are a couple that stood out at a first glance:

By mazzi, Laura K. P-, and Wired Pi, on 21 Jun 2007 01:02:25

By Nikolai Hanse, Kim Feind, and David Crac, on 22 Jun 2007 00:36:33

By C, Claire Dahlin, and Claire Dahlin, on 21 Jun 2007 22:34:53

By Jen Blackledg, Bo, and Romulo Rodrigue, on 22 Jun 2007 03:06:43

If anybody happens to read the haiku and finds some that are either especially nice, or have the wrong syllable structure (should be 5-7-5) please post them in the comments. It’ll make my life easier!

Writing the haiku generator has really sold me on the usefulness of Python, the whole thing is contained in a couple of hundred lines of code.

For the past couple of evenings I have been doing two things: Coding (for fun, not work) and experimenting with Twitter. For those who don’t know:

Twitter is a social networking and micro-blogging service that allows users to send “updates” (text-based posts, up to 140 characters long) via SMS, instant messaging,

Basically you answer the question “What am I doing now” via text message and your message is broadcast for the whole world to see. The net result of this is people sharing completely uninteresting crap about their lives with the rest of the world (”i am hungover”, “eating beans for dinner”). Perhaps I shouldn’t mock too much, people in glass houses and all that.

Twitter has all of the features of the other social networking sites (except for messages over 140 characters in length), many people seem to make dozens of updates per day and have hundreds of fans. Even better: It has an RSS feed for the most recent updates.

I have written a little script that monitors the Twitter recent entries feed for haiku! (poetry, three lines per poem, 5-7-5 syllable structure). Within minutes I had gathered the following randomly generated snippets of people’s lives into poetry form:

And my personal two favourites:

which is actually rather beautiful. And

which is so perfect I don’t think anybody will believe it was actually randomly generated! Over the next couple of days my evening project will be to get the Haiku Machine installed on geometricrate.com and churning out an endless stream of randomly generated life-poetry.

Tech details: The incredible haiku machine is a Python script, based opn the Oblomovka Haiku scanner. It obtains the RSS feed from Twitter’s most recently updated page once every minute and then scans the posts against the Carnegie Pronunciation Dictionary in order to calculate the number of syllables in each sentence. It keeps going until it finds three sentences (not necessarilty connected) in the traditional haiku 5-7-5 pattern. It still needs a lot of polish, for example the Carnegie pronunciation dictionary doen’t tell us how many syllables there are in ummm, ummmm, ummmmm and ummmmmm, and when stupid internet acronyms are used (wtf, omg) the code tries to pronounce the word (whaa-tef, ohm-gee) and gets it wrong.

Further to a suggestion recieved at coffee, I posted the following instructions in a thread on a large web forum:

Open MSPaint, shut your eyes tightly and try your hardest to draw a cat.

It’s surprisingly tricky, and often amusing.

People seemed to take to the idea and within 24 hours had posted almost 300 blind cats! Yes, three hundred. I’m gradually importing them into the brand new, official, blind cat page:

The Blind Cat Machine

So far there are 115. I’ll add a nice browsable image gallery soon.

p.s. to everybody that posted one of the ‘original ten’ images, they have now been viewed about 10,000 times.

Late edit: So far today the gallery has attracted 2,500 pageloads from 500 unique people. That’s quite a lot.

Later edit: The blind cat gallery is now pretty much finished. It attracted over 3,500 visitors yesterday. An old post from this blog (Gold Farmers) is also featured on the frontpage of Monkeyfilter, which is bringing quite a bit of new traffic

In the comments of a recent post I challenged JRM to draw a picture of a cat whilst blindfolded. Never one to pass up a challenge he obliged, along with the following suggestion:

Perhaps you could turn this into a blogject- blind-cats!!

What a great idea! So far I have collected four five six seven eight nine eleven blindfold-cats:

JRM makes interesting use of perspective

Nat’s picture is probably the most recognisable as an actual cat

CMB did an upset cat

Gemma is not very good at this

Max managed to channel the very essence of Henri Matisse

JPS is probably not going to become a professional artist

JEG pretty much drew one of my mightmares

Andy used Adobe Illustrator to make this cheerful fellow

What the fuck is this, Rob?

Kat shows her artistic side

MAN almost managed to keep his face inside his head

Anybody else with too much free time should open MSPaint, close their eyes and draw a feline. It takes less than a minute and I’ll display all the works on this blog, along with links to artist homepages. You can use something like Imageshack for free image hosting and I can be contacted via the “Contact” page.

My request for more amusing, accidental plots has been answered by fellow Durhamite and all-around clever guy Andy (who maintains a very nice page of science quotes). He describes his graph as:

… a Poincare section of the phase-space of a system of three solitons in a harmonically trapped BEC. It shows islands of regular trajectories in an ‘ergodic sea’ of chaotic trajectories.

However, there is more to it than that:

…it also looks like there is a spider hiding there.

Excellent!

I have written before about how I really like it when science produces something aesthetically pleasing in a completely accidental way, for example (quoted from my earlier post on this subject):

… an index offset by one may cause a velocity field to look like Euclid possessed by the spirit of Jackson Pollock; an accidentally placed minus sign can change a spherical distribution into some sort of fractal pancake, or maybe a neat set of three-dimensional stairs…

This one isn’t a screwup, but rather an amusing alignment of cosmic proportions. A good friend of mine, PhilB, is currently analysing simulations of galaxy formation. When he colour coded the simnulated gas depending on whether it was star forming (green) or cold (blue) he found the following distribution:

A giant cosmic cat!

For the sake of reference the face is about four hundred and fifty thousand light years across. Thanks Phil

If anybody else has amusing (or just cool looking) plots then please send them my way (contact link in the main blog menu)

It has struck me now a few times when I have been coding and visualising data that sometimes the most aesthetically pleasing results occur completely accidentally. I have seen all manner of weird and wonderful stuff come out of coding errors: An index offset by one may cause a velocity field to look like the works of Euclid possessed by the spirit of Jackson Pollock; an accidentally placed minus sign can change a spherical distribution into some sort of fractal pancake, or maybe a neat set of three-dimensional stairs (hello RAC).

Here is an example of this sort of thing, in movie form.

The red/orange haze in the following movie represents the density of gas in a region of the universe about 6 million light years across. If this simulation were to run correctly the gas would slowly (over the course of ten or so billion years) collapse down into a rotationally supported disk. Some fraction of the dense gas in the disk should change into stars and as stars form they begin to undergo supernovae. Supernovae heat up the surrounding material, and prevent the whole thing from collapsing into stars. At the end we therefore expect to see a beautiful spirally disk, similar to our own milky way.

In reality there was a coding error on my part, so as soon as a single star forms it explodes with the force of 10,000,000,000 (10¹⁰, an easy typo to leave in some code) supernovae and rips the whole universe apart in a rather spectacular fashion:

6Mb DivX3 — explosion.avi

I’d really like to see nice things that other people have made through human error. Maybe we should try and make an “I fucked science up” exhibition.

I like street art. When I say street art I don’t mean the vulgar display people usually associate with grafitti, but rather the more subtle, little things that when you notice them put a smile on your face. Here are a few examples, all taken from streetsy.com