Eurovision’s AI experiment shows how filmmakers can cut costs and go green

London-based filmmaker Matthew Blakemore has released a music video project showcasing the creative potential of AI-assisted video production.

‘Glitter Tsunami’ (Love, Peace & Sequins) shows us (in true cringeworthy Eurovision style) that by blending human ingenuity with AI, we can achieve a whopping 94% reduction in energy use. This is huge, especially in an industry known for its energy-intensive practices.

The video (watch below) despite its use of singing and dancing goats may not win any songwriting or MTV awards, but from a filmmaker’s perspective, it’s a glimpse into the future. It’s a testament to what’s possible in AI in video production – and not just for creative possibilities, but also to drastically cut costs and carbon footprints for both freelance video and filmmakers and their clients.

AI meets Eurovision: new frontier in music video production

Created on a budget of £600, “Glitter Tsunami” demonstrates how AI can help artists produce content with limited resources. Blakemore wrote the song lyrics and storyboarded the video using traditional methods, then utilised various AI tools to bring his vision to life.

Launched at the International Broadcasting Conference (IBC) in Amsterdam, the project challenges misconceptions about AI in creativity and highlights AI-driven sustainability.

The video features over 35 references to Eurovision entries, from recent winners like Nemo and Loreen to acts like Daz Sampson and Scooch, celebrating the contest’s history.

The massive power needs of film sets, the carbon footprint of transporting equipment, on-location shoots means traditional filmmaking often clashes with environmental goals. But the AI alternative offers a different story. It aims to show that with AI, videos and films can be created while being kinder to the planet.

Here’s a breakdown of the energy used to make the video:

• Music generation: Over 100 AI-generated song iterations were created using state-of-the-art tools, consuming about 0.7742 kWh of energy.

• Sound effects: Generated 840 AI sound effects to find the right sounds; only 42 were used, consuming 0.3252 kWh with advanced audio synthesis platforms.

• Visual content: Produced over 15,000 AI-generated images and 1,000+ video clips, consuming an estimated 60.405 kWh.

Total energy consumption was approximately 61.5044 kWh, equivalent to charging 5,125 smartphones or running a refrigerator for about 2.5 days.

“While AI content generation consumes energy, it’s crucial to compare it with traditional methods,” says Blakemore. “A typical shoot involves travel, lighting, and post-production, consuming much more energy. We need to consider the full environmental footprint when assessing new technologies,” he says.

Estimates suggest a traditional music video could consume 1,000 to 10,000 kWh or more, depending on scale. This means the AI-assisted production’s energy use is potentially under 6% of a conventional shoot.

Can AI music videos become a legal minefield?

A 2022 study in the Journal of Intellectual Property Law found that 85% of AI music startups used some unlicensed material in training data. A 2023 WIPO report suggests strict licensing requirements could reduce AI music tech startups by up to 60%.

Regarding the AI Song Contest’s 2024 ban on generative tools Suno and Udio, Blakemore says:

The contest’s stance is glaringly inconsistent, given that previous winners used AI tools trained on unlicensed data, including ChatGPT. This moral absolutism and selective gatekeeping not only silences valid ethical perspectives but also brings the contest’s validity into question.

This opens up exciting doors for freelancers:

• Niche it: Use your expertise in specific genres (think historical docs, sci-fi shorts, educational animations) to create AI-generated content that caters to niche audiences. AI lets you make compelling, targeted content that stands out.
• Personalised videos: Imagine offering clients videos tailored to their individual preferences! AI makes this possible, creating unique experiences that grab attention in the crowded digital world.
• Streamlined pre-production: AI-powered tools can quickly generate storyboards, animatics, and even rough cuts. This means faster iteration and smoother collaboration with clients.
• Visual effects on a budget: AI is changing the game for visual effects and post-production. Freelancers can use it to create amazing visuals and deliver high-quality projects faster and more affordably.
• Accessibility: Make your videos reach a wider audience with AI-generated captions, translations, and audio descriptions.

For freelancers, this means a chance to expand their skills, specialise in AI-adopted filmmaking, and potentially thrive in this exciting new era. Perhaps, however, less the dancing and singing goats.