ARCHIVEArarat Turns Brandy Glasses Into Emotion Displays in Belarus
How Ararat's 'From Heart to Heart' Made Emotion Recognition Part of the Ritual in Belarus Ararat's...
Oracle Innovation Trains a Neural Network to Write Eurovision
Oracle Innovation's 'Douze Points to Artificial Intelligence' is a 2019 experiential campaign created by Israeli agency Allenby in which a neural network trained on hundreds of Eurovision songs generated thousands of melodic and lyrical fragments that were then curated into a fully performed original track called 'Blue Jeans & Bloody Tears.'
The campaign launched in Israel in May 2019, timed precisely to coincide with the Eurovision Song Contest 2019, which was held in Tel Aviv — marking the first time Israel had hosted the contest in four decades. That cultural moment gave Oracle Innovation a rare platform: a nation already primed for Eurovision fervour, primed equally for a demonstration that artificial intelligence could engage with one of pop culture's most rule-bound, emotionally loaded musical formats. The brief, as executed by Allenby Concept House, was not to produce a novelty stunt but to stage a genuine creative inquiry — could algorithmic music, shaped by machine learning trained on the Eurovision canon, produce something a human audience would actually feel?
The answer took the form of 'Blue Jeans & Bloody Tears,' a song whose subject matter — a disillusioned love affair between a human and a machine — mirrored the very process that produced it. The campaign sits at the intersection of the electronics and technology industry, using Oracle Innovation's AI and cloud capabilities as both the tool and the message. In Israel, where the technology sector had established itself as a global force in enterprise software and deep-tech development, the campaign spoke directly to an audience fluent in both the cultural language of Eurovision and the professional language of neural networks. Clapboard rates this campaign 92 out of 100.
Allenby Concept House built the campaign's strategic logic on a single, potent cultural overlap: Israel in 2019 was simultaneously the host of the Eurovision Song Contest and one of the most active technology ecosystems in the world. The agency identified Eurovision not merely as entertainment context but as a structured data set — decades of songs governed by consistent emotional and compositional conventions — making it an unusually tractable domain for AI-generated music training.
The strategic insight was that Eurovision's rule-bound nature, its melodic repetition, its lyrical archetypes of longing, triumph, and heartbreak, would actually make a neural network's outputs more legible and more emotionally resonant than a less codified musical genre might. By training the model on hundreds of Eurovision songs, the system could generate thousands of melodies and lyric lines with a recognisable internal grammar, giving the human curation team — led by Amir Shoenfeld and including Avshalom Ariel, Ben Scheflan, Nim Shapira, and Eran Hadas — enough material to assemble a song that felt genuinely structured rather than randomly assembled.
In the broader technology landscape that contextualised the campaign, the cloud and AI infrastructure market was expanding rapidly. According to Canalys, worldwide cloud infrastructure services revenue reached $27.5 billion in the third quarter of 2019 alone, representing thirty-seven percent year-over-year growth. Synergy Research estimated full-year 2019 cloud infrastructure spending at approximately $96 billion, with annual growth of around thirty-eight percent for public infrastructure-as-a-service, platform-as-a-service, and hosted private-cloud services, according to Synergy Research. Oracle Innovation operated in direct competition with platforms including Amazon Web Services and Microsoft Azure, both of which were investing heavily in AI-native developer tools during the same period. Against that backdrop, demonstrating AI creativity through Eurovision — Israel's most visible cultural stage in 2019 — was a strategically coherent act of brand positioning rather than a peripheral experiment.
Creative director Nimrod Shapira led the project for Allenby Concept House, and his articulation of its purpose remains the clearest window into its intent. As quoted by CTech and Calcalist in May 2019, Shapira stated: 'I could not pass on the opportunity to use machine learning algorithms to explore human creativity.' That framing — exploration rather than replacement — defined every creative decision that followed.
The neural network's role was generative and abundant. It produced thousands of melodic units and lyric lines drawn from its training corpus of hundreds of Eurovision songs. The curation team's role was editorial: selecting, combining, and shaping those outputs into a coherent composition. Song production was handled by Avshalom Ariel, with song curation credited to Amir Shoenfeld alongside Ben Scheflan, Nim Shapira, Avshalom Ariel, and Eran Hadas. The resulting track, 'Blue Jeans & Bloody Tears,' was performed by Izhar Cohen, Israel's winner of the Eurovision Song Contest in 1978 — a choice that collapsed the distance between Eurovision's human history and its algorithmic future into a single performer. Avshalom Ariel also contributed to the project as a performer.
The visual and experiential layer of the campaign was directed by Karni and Saul of Sulkybunny, with title design by Adam Blufarb. The production documentation, as archived on adsoftheworld.com and clapboard.com, positions the campaign as a fully integrated experiential work: the song was not a demo or a proof of concept left in a laboratory but a finished, performed artifact released into the same cultural arena as the Eurovision Song Contest 2019 entries it had studied. The thematic conceit — a love story between human and machine, rendered in the emotional vocabulary of Eurovision — gave the AI-generated music a conceptual frame that elevated it from technical demonstration to creative statement.
No verified performance data for the 'Douze Points to Artificial Intelligence' campaign is publicly available in terms of reach figures, engagement metrics, or commercial outcomes. What the campaign's structure does make analytically legible, however, is the quality of its strategic and creative architecture — and that architecture is unusually strong.
The decision to anchor an AI capability demonstration to Eurovision Song Contest 2019 in Israel gave Oracle Innovation a cultural moment with natural amplification. Eurovision, as a media event, generates sustained press coverage across entertainment, technology, and cultural commentary channels simultaneously. A campaign that offered journalists a genuine news angle — artificial intelligence producing an original Eurovision-style song, performed by a real Eurovision winner — was structurally positioned to earn coverage that a conventional product demonstration would not. As documented on adsoftheworld.com and across Allenby's own published records, the campaign was submitted to international creative archives, indicating the agency's confidence in its creative standing.
The casting of Izhar Cohen, Israel's 1978 Eurovision Song Contest winner, gave the campaign a human credential that functioned as editorial validation — the AI-generated music was deemed performable by a figure with decades of credibility in the format the neural network had studied. The song's subject matter, a disillusioned romance between human and machine as reported in campaign documentation on adsoftheworld.com and clapboard.com, ensured that the concept and the content were unified rather than decorative. In a category — enterprise technology — where brand campaigns frequently divorce product capability from emotional resonance, 'Douze Points to Artificial Intelligence' achieved the rarer outcome of making the product the story.
'Douze Points to Artificial Intelligence' is an experiential campaign created by Allenby Concept House for Oracle Innovation in Israel in May 2019. A neural network trained on hundreds of Eurovision songs generated thousands of melodic and lyrical fragments, which were curated into a fully performed original song called 'Blue Jeans & Bloody Tears.'
A neural network was trained on hundreds of Eurovision songs. The system produced thousands of melodies and lyric lines through algorithmic music generation. A human team including Amir Shoenfeld, Avshalom Ariel, and creative director Nimrod Shapira curated and combined those outputs into the final performed track, as documented on adsoftheworld.com and clapboard.com.
Izhar Cohen performed 'Blue Jeans & Bloody Tears.' Cohen is Israel's 1978 Eurovision Song Contest winner, making him a historically resonant choice for a song produced by a neural network trained on the Eurovision canon. His involvement gave the AI-generated music a direct connection to Eurovision's human performance tradition.
Nimrod Shapira served as creative director at Allenby Concept House and project lead for the campaign. As quoted by CTech and Calcalist in May 2019, Shapira said he could not pass on the opportunity to use machine learning algorithms to explore human creativity, framing the project as an inquiry rather than a stunt.
Eurovision Song Contest 2019 was held in Tel Aviv, giving Israel its highest-profile cultural moment in decades. The contest's rule-bound musical conventions made its song catalogue an unusually structured training set for a neural network. Oracle Innovation used the alignment of Israel's host-nation status and its technology sector to stage the campaign.
Clapboard awarded the 'Douze Points to Artificial Intelligence' campaign a score of 92 out of 100. The campaign was created by Allenby Concept House for Oracle Innovation in Israel in May 2019 and is categorised as an experiential work in the electronics and technology industry, archived on clapboard.com.
.webp)
Creative Collaboration Platform: The Backbone of Modern Creative WorkWhat is a creative collaboration platform?Clapboard defines a creative collaboration platform as the operational core for today’s distributed creative teams. This is not a collection of digital collaboration tools stitched together; Clapboard’s system
Understanding the Creative Production MarketplaceWhat is a Creative Production Marketplace?Clapboard defines a creative production marketplace as a structured ecosystem where brands, agencies, and production talent converge to execute video, advertising, and branded content at scale. Unlike the legacy agency or freelanREAD FULL ARTICLE
The Urgent Need for a Creative Production Operating SystemWhy Traditional Creative Production Models Break DownClapboard sees the creative production landscape shifting under pressure. Content velocity is no longer a buzzword—it’s a mandate. Modern advertising demands relentless output across dozens of channels, each wREAD FULL ARTICLE
