Project Description
This project explores the benefits of subsidising residential solar installations in New Zealand, comparing it to large-scale renewable energy projects like wind farms.
Currently, solar energy contributes less than 1% to the country’s energy mix. By offering subsidies, homeowners could reduce energy bills, become less reliant on the grid, and support sustainable energy production. A 30% subsidy would bring installation costs down to $5,600–$7,000, leading to annual savings of $1,500–$2,000. Over 20 years, households could save $30,000–$40,000, with a payback period of just 3–5 years.
While wind farms benefit energy companies, household savings from reduced wholesale electricity prices are minimal, at $100–$200 per year. Investing in home solar offers greater long-term savings and reduces grid pressure. If 50% of New Zealand’s homes adopted solar, it could lead to annual savings of $1.25 billion, and a total of $25 billion over 20 years.
Data Story
This project involved a lot of research to find the right data and craft a compelling story around it. Initially, we thought about building an app, but since New Zealand is so great for solar, we figured we could just ask, "Are you in NZ?" and, if yes, skip straight to the PR pitch about potential savings based on the weather—a bit too predictable.
The real focus was on gathering and working with data to tell a story. We used a range of AI tools, starting with something we humorously dubbed the Steve Wozniak AI model—Actual Intelligence, not artificial!
Once the data was sorted, we used AI to rewrite our text, making it more articulate and avoiding repetitive ideas, all while condensing the facts. We asked for a writing style loosely inspired by Taika Waititi with a dash of Jemaine Clement as a quirky scientist, so it stayed technical but fun. It took a lot of "try again" and "one more time" before we had our "Oh, I like that!" moment. That’s how the dialogue evolved, with the second draft often outshining the first.
We then used AI text-to-speech to narrate the script, which we dropped into a video editor. The final touch was image selection, where we stuck to Creative Commons or fair-use visuals.
https://data.mfe.govt.nz/data/?q=sunshine+2016 (The resolution of this and the small contrast between data points made it not really useful for build a solar viability app as the only places bad for solar is where people don't really live and the https://solarview.niwa.co.nz/ already does it all.
other sources referenced
https://www.eeca.govt.nz/insights/energys-role-in-climate-change/the-future-of-energy-in-new-zealand/
https://www.stats.govt.nz/indicators/sunshine-hours/
https://solarview.niwa.co.nz/
AI tools:
https://huggingface.co/spaces/finegrain/finegrain-image-enhancer?ref=therundown
https://huggingface.co/spaces/innoai/Edge-TTS-Text-to-Speech with the en-NZ-MollyNeurak-en-NZ (Female) voice.
https://chatgpt.com/
https://gemini.google.com/
https://www.hedra.com/app/stylize/bobble
https://aitestkitchen.withgoogle.com/tools/image-fx