#12: Science cocktails | Hi-tech fortunes? | Secrets in plain sight
art | science | digital culture | design | creativity | words | tech
science | art
The Cocktail Accelerator
I haven’t sent an email in a couple of weeks as I was busy preparing for a trip to Switzerland where I was making science-themed cocktails. I was invited over based on some previous science cocktail projects I had done, including one from Science Hack Day San Francisco in 2012.
This time around, I did a cocktail performance/experience in the social hub connected with the World Conference of Science Journalists in Lausanne. It was a really fun event and people seemed to really enjoy the experience, and even liked the cocktails. I did the project in collaboration with a few other people: Matt Bellis (physicist), Mira Ederer (mixologist), Rae Cooper (graphic designer), and Virginia McVeigh (cocktail consultant).
Here is the menu:
Each cocktail has scientific themes built into the recipe and used scientific principles, effects, or props to make and present them.
Here is the Milky Way, consisting of visible and dark matter in the correct proportions, a somewhat swirling galaxy on the surface (although you see the main effect during the preparation), and a black hole at the centre of the galaxy:
I’ll write more about the project and process in my paid part of the newsletter where I share more of my private in-process work. You can check that out if you’re prepared to pay for what is effectively a cup of coffee a month. (It’s only paid as I don’t want to expose all my workings completely publicly and a small buy-in indicates a commitment to a relationship based around the work.) I’ll have more photos and info there. Just in case you want to do that, here’s a button!
art | tech
Artificial fortunes
Just what can do-it-yourself technology combined with artificial intelligence and machine learning achieve? Is it belief-based computing? From the Believe it Yourself project website:
The fast spreading and ease of access to Machine Learning and Cloud computing has brought to a boom of experiments and excitement around our ability to build machines that make sense, learn, measure and predict the world around us.
Moreover, with enough examples, we can train a tool or a ‘machine’ to recognize or quantify pretty much anything we want. ‘Beauty’, ‘Hotdog-ness’ or the more problematic ‘Criminal-ness’ and ‘Sexual orientation’ can be now measured within a few frames, based on a model, a probability, determined by a set of arbitrarily collected data. Subjective judgments and biased datasets can easily be turned into objective measures and potential truths, which will then be embedded in devices around us.
But what if we would train machines to measure even more unmeasurable, personal and culturally driven things? If we gather enough samples could we detect signs that prove and detect our superstitions? and can we use that to build tools and devices that reflect our own beliefs?
BIY™- Believe it Yourself is a series of real-fictional belief-based computing kits to make and tinker with vernacular logics and superstitions.
It’s an art project dressed up to appear like real technology, which is most of the point. Nicely executed with just the right amount of authentic work combined with a dose of trickery, it’s a comment on belief in technology versus belief in ideas.
digital culture | tech
Secret codes hidden in plain sight
I’ve been fascinated by secret codes ever since I was given the “The Official Spy’s Guidebook” as a child. It’s recently been re-released with a new cover but all the same content I loved (especially the pocket spy kit on p.52, at 30s in the video below).
One of the types of codes I’ve always found interesting is steganographic code. This involves incorporating a code in another medium in a way that doesn’t seem to change that medium but can be extracted if you know the secret. A fairly common example is the use of copyright watermarks in images. You can’t see the watermark in the image (if it’s done well) but you can extract it from the raw data.
Steganography has been on the rise since content copying became rampant on the internet. One place where content seems to be copied without permission, but with no real way to detect it, is song lyrics. A range of services provide lyrics to songs, including Google’s basic search function. If you search for the lyrics of a song, you don’t even need to click through to a website to see them: Google just displays them.
The company Genius Media specialises in song lyrics and has claimed that Google has been taking their transcriptions without permission. To prove this, they showed how they had encoded a secret message in the use of apostrophes (see highlighted apostrophes below). It’s often hard to notice, but there is a difference between straight ' and curved ’ apostrophes as far as a computer is concerned—it’s as much of a difference as between the letters a and b to them.
Genius Media chose to use either straight or curved apostrophes in a certain order to code a message that could be read back if you knew it existed. They found that Google’s lyrics had indeed used the pattern of apostrophes in which they had encoded the phrase “red handed” using Morse Code. The story was reported by the Wall Street Journal but the full article is behind a paywall. You can click through though and see a video of how the encoding works in the article preview.
As far as I can tell, there has been no resolution of this case yet, but I suspect there will be a legal battle that lasts a while.