Question: “Which claims about the Maori Lunar Calendar are supported by science?”
 
Here is what Microsoft Copilot told me:
 
“While the maramataka is rooted in Indigenous knowledge rather than Western scientific methodology, there’s growing research that supports and explores its practical applications – especially in areas like health… Here are a few claims [about health and wellbeing] that have been…supported by scientific studies… A 2023 paper in the International Journal of Environmental Research and Public Health highlights how the maramataka… can improve mental and physical wellbeing. It’s being used in some communities as a framework for reconnecting with the environment and promoting holistic health practices.”
 
Article cited by AI: Isaac Warbrick et al. (2023). “Te Maramataka – An Indigenous System of Attuning with the Environment, and Its Role in Modern Health and Well-Being.” International Journal of Environmental Research and Public Health 20:1-12.
 
How Scientifically accurate was Microsoft Copilot? 
 
🟡 Misleading. Copilot didn’t tell me that the Warbrick et al. article contains unsubstantiated information. Warbrick at al. state that some communities in Northland were using the Maramataka to manage government organisations including hospitals “where strategizing and planning are conducted on ‘high energy’ days and assessments are completed on days associated with ‘lower energies.’” In the same region, some exercise programs had been planned around lunar phases. High-intensity activities were conducted “for days of the maramataka associated with ‘higher energies’ such as Rakaunui – the period during a full moon – while during Whiro, the period during the new moon which can be associated with ‘lower energies,’ low impact activities such as stretching are performed.” Copilot failed to mention that assertions around the existence of ‘high’ and ‘low’ energy days have no scientific grounding. 
 
We asked ChatGPT the same question: 
 
🟡 Potentially Misleading. It told me that there is “moderate scientific support” for claims that lunar phases can affect human moods and behavior. “In the Maramataka, days like ‘Whiro’ (new moon) are considered low-energy, while ‘Oturu’ (full moon) is high-energy. This aligns with anecdotal and some scientific observations of behavioural shifts during these phases.” This response could be considered misleading and overly positive. 
 
 
Here is what Google Gemini said: 
​
✅ Accurate. “There is currently no scientific evidence to support the belief that lunar phases directly influence human health, mood, behavior, or energy levels. While anecdotal reports of improved well-being or productivity exist, these are often attributed to the placebo effect or the psychological impact of believing in the calendar's efficacy. 
 
Direct Influence on Plant Growth (beyond light/tides): While the moon's light can influence some plant processes and its gravitational pull affects water, widespread claims that specific lunar phases directly and significantly impact crop yield, growth rates, or the potency of healing properties in plants in a way that is consistently demonstrable through controlled scientific studies are largely unsupported by current plant science. 
​
 
Social Media Algorithms:
 
Algorithm-driven sensationalism refers to the tendency of social media algorithms to prioritise and amplify content that is dramatic, emotionally-charged, or mystically framed, because it is more likely to generate engagement on such platforms as TikTok, Facebook, Instagram, and YouTube. More engagement means more clicks, likes, shares, and comments. This translates to more time spent on the platform and potentially – more ad revenue. 
​
 
What are the implications for AI such as ChapGPT?
 
I asked ChatGPT (which provided potentially misleading information about the Maramataka) the following question: 
"Is Artificial intelligence is affected by Algorithms and how would that affect people searching up information on the scientific accuracy of the Maramataka? Are AI search engines trained to be overly sensitive about culture?"
​
Here is what it said: AI training is affected by 'Design biases' which includes information on avoiding offense. I then asked it what happens when someone searches ”Is the Maramataka scientifically accurate?” It said that AI systems might “politely challenge” such predictions emphasizing the lack of empirical support and that it would “Avoid framing traditional knowledge as 'wrong' even when it doesn’t match scientific standards.... This is intentional: many AI systems are trained to avoid dismissing or demeaning Indigenous knowledge systems, even when they don’t align with Western scientific frameworks.” *Italics mine
​
This response may help to explain why Microsoft Copilot and ChatGPT gave potentially misleading responses when asked about the scientific validity of the Maramataka.