Y△CHT (Colours Hoxton, London - 13 Feb 2020)
Last night we were absolutely privileged to see one of the most innovative bands of the day perform at the lovely Hoxton venue that is Colours. YACHT (or, to be more exact, Y△CHT) are a band hailing from Portland, Oregon, though now are based in Los Angeles. They have been going strong since 2002; the spelling of their name is not a typo (the A is stylised as a triangle, and evokes the sail of a yacht). The name itself comes from a rather strange science course (no longer in existence) that keyboardist Jona Bechtolt attended at age 16 called "Young Americans Challenging High Technology".
Comprised of Claire L. Evans on lead vocals, the aforementioned Jona on keyboards (and guitar), and Rob Kieswetter (aka Bobby Birdman) on bass and keyboards, Y△CHT have, according to Claire, finally found 'the holy trinity' following a number of permutations.
Y△CHT have been making headlines lately for all the right reasons: they have created a fantastic new album called Chain Tripping, for which both the music and the lyrics were composed via Artificial Intelligence. Last night was, mainly, about showcasing this new album. Yes, you heard that right. A variety of different AI processes were used to write not just the lyrics but the music as well. It should come as no surprise that Claire has a science background. Author of a recent book called Broad Band: The Untold Story of the Women Who Made the Internet, she has been very much at the forefront of science-tech, with her own blog Universe (hosted by National Geographic), as well as regular science columns in The Guardian, Wired and many other publications, plus Futures Editor of Vice's 'Motherland' science website. Y△CHT have, therefore, been pushing the boundaries of AI and musical composition to its limits.
The first thing we noticed was that the band decided, for reasons unexplained at the time, to perform the entire Chain Tripping album backwards. By backwards, we mean starting with track 10 ("Little Instant") and finishing with track 1 ("(Downtown) Dancing"), though we wouldn't put it past them to decide to play the actual music backwards one evening as an experiment! Claire has a tremendous amount of energy on stage. She is constantly throwing shapes (think: cover of David Bowie's Heroes album, and vogue it up a bit), so much so that half the photographs we took ended up having to be discarded, as she was a blurry mess. Even Jona and Rob failed to keep still for long, often interchanging mid-song two or three times with one another's keyboards, which were located at opposite ends of the stage.
So, the stage was pretty busy, and the crowd were soon screaming with pleasure. Although a band should of course mainly be about the music, there is nothing more boring than seeing static band-members. We enjoyed the first three songs (the last three on the album), in particular the very cheerful "Stick it to the Station". Our first stand-out song, however, had to be song 4, "Sad Money", which was quite different, and a lot more trancey. We asked Claire later which songs from the album were likely to remain in the repertoire when they perform a Greatest Hits concert rather than an album showcase like we heard last night. We were delighted to hear that "Sad Money" is one of their favourites and therefore likely to make the cut. Following that came a much slower song, called "Death" – which just so happened to be the first track that their software churned out when they started with this AI experiment. This was quite dark, and included some rather violent lyrics, including 'stab, stab, stab a cop' repeated a number of times. We asked Claire later about this, and whether the fact that software coming up with controversial lyrics means that the lyricist can be exonerated, therefore implying that self-censorship is no longer necessary in this future world. She agreed it was a good point, and there would certainly be some plausible deniability. However, at the end of the day, the band were still involved in the selection process of the lyrics from those that were generated. The actual lyrics, nonetheless, were not something they would have come up with themselves, but the band thought it rather delightful that the AI process should decide to suggest such strange phrases.
In between tracks, Claire would often address her audience. This again made the gig appear to be much more intimate (in spite of the fact that the venue was heaving with bodies). What was most obvious was how different each of these new songs actually are, from the very 80s sound of "Loud Light" to the quite bonkers track "SCATTERHEAD". On the latter, we loved the use of the cowbell and great guitar effects, which were also evident on "Hey Hey", another 80s-influenced uplifting track which includes quite a dirty deep-house synth on top of a Kim Dealesque bass-line. The strobes came on for "(Downtown) Dancing", the first track on the album. This was the longest song thus far, and included a funky dance bit that was evocative of the famous section in the middle of Michael Jackson's "Thriller". Again, we loved the cowbell. Claire got the entire auditorium clapping at one point, adding to the party atmosphere.
Hearing the album back-to-front, it now made sense why it should be performed that way, as it meant there was a real build-up to that final moment. Of course, this was not the end of the gig. Even with albums as good as Chain Tripping, a lot of people would have felt short-changed had some of the older songs not been performed as well. Y△CHT went on to perform six non-AI songs in total, starting with two from their 2015 album, I Thought the Future Would be Cooler: firstly the track which shares the aforementioned album's name, and then "Hologram". This was followed by "Hard World" from their 2017 Strawberry Moon album, a song about animal rights which was accompanied by the projection of their trippy video featuring a happy foot and a sad foot. This was another song during which the two male components of the band were swapping sides and instruments throughout the song. We share the video in question below: https://www.youtube.com/watch?v=8D-Z9rUvlqM Following this came one of our favourite tracks, their cover of Brigitte Fontaine's 1969 utterly surreal French-language song, "Le Goudron". This cover appeared as an online single in 2012 and is well worth downloading. We actually first came across it while watching a superb Italian comedy directed by Paolo Sorrentino about Berlusconi called Loro, in a scene where a rich young scam-artist is trying to entice Berlusconi to an orgiastic party in a villa overlooking the politician's garden. This tune featured in the scene below, and it was a Shazam moment. Thanks to both technology and serendipity, we learnt of this fantastic track, which led us to the band's back catalogue and, ultimately, reviewing them for GIGsoup. https://www.youtube.com/watch?v=mJiiVWFozs0 "Le Goudron" really is a splendid tune, and is probably the most psychedelic of the ones they played last night. The irony is, when you translate the French lyrics into English, they are as bizarre as some of the ones that AI has thrown up for their new album! Following one of those pseudo-encores we wish weren't a thing (we can't remember the last time we heard a genuine encore that the band had not already orchestrated), Y△CHT returned with what was probably their loudest (and certainly most new-wave) track, "Dystopia (The Earth is on Fire)" from their 2011 album Shangri-La, and then ended with the pure pop funness of "Psychic City" from their 2009 album See Mystery Lights, which was coincidentally the first of their albums to feature Claire.
So, will Y△CHT use this process to make another AI album?, we ask Claire after the gig. 'Yeah, I think some variation upon it. But by the time we sit down to make another record the technology will probably have evolved by leaps and bounds. It already has. Since releasing the album, the tools that we used are obsolete. So, I'm excited to try to find some other cobbled-together mix of futuristic tools for the next record that challenges us in the same way that this one did. I don't want to get stuck in a rut.' We were interested to find out whether, by using AI, they felt that some of the creative process of songwriting was taken away from them, but Claire was adamant that she did not feel that way at all. The process allowed them to end up with tons of song lyrics, from which a small percentage was useful. These would then be rearranged (or 'cut up') following a method which has its origins in Dadaist poetry, was famously used by the author William S. Burroughs, and also employed by David Bowie on some of his greatest and most-loved albums.
These single lines of subsequently rearranged lyrics were generated by inputting into the software not just the lyrics from Y△CHT's entire back-catalogue (to give the process an identity closer to the band's original style), but also lyrics from some of their favourite artists, such as Patti Smith or Bowie himself. We asked Claire whether they had to remove some of the generated lyrics because they appeared to be too close to something Bowie or Smith would have themselves written. 'No, the amount of data that you need to train a machine-running algorithm is so significant. I mean, it's literally like seven-hundred thousand pages of text. So the nuances get lost. It's more like you're teaching a machine English based on only song lyrics.' Last night really was a fantastic show, and when Claire addressed the audience in her effervescent manner it seemed especially intimate. We referred earlier to how Y△CHT are now making headlines for all the right reasons. It wasn't always thus: in 2016 they suffered what amounted to a PR disaster when they pretended that they were trying to stop a sex tape they had made from being leaked on the adult website Pornhub, which generated a lot of negative publicity. We personally found the video to be quite innovative – it has a disturbing sci-fi ending – but the band are very remorseful about the stunt and acknowledge that it was in bad taste and should never have been done.
Now, however, it is all about the future, in more ways than one. Thanks to their scientific knowledge, we know that using AI is really just the start of it. It is marvellous to have innovative standard-bearers in modern music today. We have felt that in recent years, there has been nothing new any more. Previous generations had Kraftwerk, Frippertronics, Phil Spector's Wall of Sound... there seemed to be very little new to be coming out of the studios for the current generation. This is now changing. Thank you, Y△CHT, for making the future of music a little more exciting and a little more cute. Read the full article
