folk-rnn (v1) tune, The Irish Show

Having apparently devoted my life to the preservation and promotion of artificial folk music, I bring you another gem-of-a-machine folk tune I have found by listening to The Endless folk-rnn Traditional Music Session. This one is titled by the network itself, “The Irish Show”: Screen Shot 2017-11-23 at 17.11.27.png

I don’t know what “show” the machine is referring to, but I want to watch it. This is close to a cracking good tune, despite the slight miscount in bars 8 and 16, and the strange second ending without a first. Let’s correct the miscounts, and remove that second ending altogether. I build my own second endings for the tune and turn, and change that G in bar 11 to F natural because I think this “show” should be a little stranger.

Screen Shot 2017-11-23 at 17.22.46.png


folk-rnn (v1) tune, “Sean No Cottifall”

As the world’s most sought-after interpreter of folk-rnn-generated tunes, I bring you another one I have found by listening to The Endless folk-rnn Traditional Music Session. This one is titled by the network itself, “Sean No Cottifall”:


This tune has a different structure from many of the others the model generates: AABCC, where the A and C parts are four measures long, and the B part is 8 measures. The A part sounds to me like a call to the dance floor. The B part is where the main portion of the dance begins. Then the C part reminds me of a hopak… for some reason!

I make some changes this little piece. I repeat the B part twice, and give it two endings to smoothen the transitions. I also flatten the third in bar 11 and play it against a C major. I really like the contrast that gives. Then I give the C part two endings, and make it go crazy.


My partner helped me realise the craziness of the hopak. Thanks Carla!

Music in the Age of Artificial Creation – An Illustrated Concert

Music in the Age of Artificial Creation (part of the 2017 Being Human Festival)

– An Illustrated Concert –

Nov. 20 2017 7 PM St. Dunstan and All Saints Stepney (

Tickets £5 (, or £10 at door

Machine learning has been making headlines with its sometimes alarming progress in skills previously thought to be the preserve of the human. Now these artificial things are “composing” music. Our event, part concert part talk, aims to demystify machine learning for music. We will describe how we are using state-of-the-art machine learning methods to teach a computer specific musical styles. We will take the audience behind the scenes of such systems, and show how we are using it to enhance human creativity in both music performance and composition. Human musicians will play several works composed by and with such systems. The audience will see how these tools can be used to augment human creativity and not replace it.

Programme includes:

– Richard Salmon plays artificial Bach chorales on the St. Dunstan organ
– Daren Banarsë ( and musicians play computer-generated tunes in the Irish style
– Luca Turchet ( interprets computer-generated tunes on his new “Smart Mandolin”
– Ensemble x.y ( plays “Two short pieces and an interlude”, a composition co-created by Bob L. Sturm ( and computer; and “Bastard Tunes”, a composition co-created by Oded Ben-Tal ( and computer
– Jennifer Walshe ( performs a new work interpreting computer-generated text

A wine reception will follow.

Music from the Ocean CD finally available

Fifteen years ago, I produced an experimental music album of over 58 minutes based on the sonification of near shore ocean buoy data. The album is rated 5 stars on Discogs (thank you to the two raters). It has been scrobbled over 1400 times (whatever that is). Swan Fungus rated it no. 17 in The Top 50 Records Of 2006. The album was being sold by Aquarius Records in San Francisco, but that shop is long gone. I have finally moved the remaining copies to the UK and am distributing them myself.


Buy Now Button with Credit Cards

The Seven Deadly Sins of AI Predictions – MIT Technology Review

“Performance versus competence” and “Suitcase words”. Did I mention the slides to HORSE2017 have been posted:

Mistaken extrapolations, limited imagination, and other common mistakes that distract us from thinking more productively about the future.

Source: The Seven Deadly Sins of AI Predictions – MIT Technology Review