Somebody somewhere posted a link to a blog post I wrote in August 2024 entitled A Little AI Magic. That’s become one of the top-five traffic destinations here.

In terms of AI, two years is a long time ago. When I wrote that post, Google had just re-enabled image generation after halting image generation in February of that year following controversy surrounding the generator’s insistence on racial diversity in all images.

Today, I decided to revisit the question of whether Microsoft’s Copilot or Google’s Gemini was the better choice for image generation.

For Copilot I selected the “Smart” setting – this utilizes GPT-5.1 – and clicked on the “Create an image” button. I then pasted in the prompt:

Generate image in style of emoji that represents Nichiren Buddhism

Here’s the result:

I then went to Gemini and selected “Thinking” mode and “Create images” and pasted the same prompt.

Here’s the result:

Personally, I prefer the “imagination” shown in Gemini’s image. The Copilot image is, well, boring. It’s accurate and useable but it doesn’t “represent Nichiren Buddhism” as much as label it.

So, while there was no comparison two years ago, the image generators in 2026 are far more sophisticated.

Now I shall digress because Gemini tonight lapsed into declaring that the Daimoku is “Nam-myoho-renge-kyo.”

Here’s the “thinking” that went into creating its image:

Analyzing Nichiren Buddhism
I’m currently focused on deconstructing Nichiren Buddhism to pinpoint its central symbols. I’ve identified the Daimoku (“Nam-myoho-renge-kyo“) and the Gohonzon (calligraphy) as critical. The challenge lies in condensing these elements into a single, readily interpretable emoji.

Embodying Nichiren’s Essence
I’m now zeroing in on embodying the core elements: Daimoku, Gohonzon, Nichiren, and the lotus. My strategy is to combine the lotus with a stylized scroll. The scroll will bear symbolic text in lieu of a detailed Gohonzon, and a friendly figure, representing Nichiren, will be added to the composition. The intent is to refine it with bold lines, vivid colors and a sticker appearance.

Verifying Visual Fidelity
I am now assessing the visual elements against the initial prompt. I am concentrating on how accurately the artwork reflects the specified style and whether key symbolic elements are present.

Analyzing Style Compliance
I am checking the image for alignment with the specified “emoji style,” looking at cartoon figures and strong lines. I am focusing on how well it visually captures Nichiren Buddhism. I also look at the monks and scrolls and whether the kanji ‘Nam-myoho-renge-kyo‘ is correctly written. I am preparing to share my findings.

I have been using a Gem – a reusable prompt – I wrote to generate images for the Daily Dharma since December of 2025. Not once has Gemini said that the Daimoku written representing the Japanese text using Roman letters – Romanji – begins “Nam” rather than “Namu.”

Gemini knows that seven Chinese characters make up the Daimoku – 南無妙法蓮華経 – and that the first two – 南無 – are rendered Namu in Romanji.

It is a phonetic transliteration of the Sanskrit word Namas, used to express deep respect and dedication. It does not carry the literal meanings of the individual characters, which are “south” (南) and “nothing/without” (無).

Why write “Nam” instead of “Namu”?

This is a very good question that touches on Japanese phonetics and how languages adapt foreign words.

The reason “南無” is romanized as Nam instead of the standard Japanese pronunciation Namu is due to a linguistic phenomenon called vowel devoicing or elision that occurs during the rhythmic chanting of the mantra.

Here is a breakdown of why this happens:

1. The Sanskrit Origin As mentioned before, the characters 南無 act as a phonetic transliteration of the Sanskrit word Namas (meaning devotion or bowing to). Ancient Chinese and Japanese monks chose characters that sounded similar to “Namas,” regardless of their literal meaning.

2. Standard Japanese Pronunciation In standard, isolated Japanese:

南 is usually pronounced Nan (or Minami).

無 is usually pronounced Mu.

Put together, the standard dictionary pronunciation is Namu.

3. The “Vowel Drop” (Elision) In spoken Japanese, the vowel sounds ‘u’ and ‘i’ are often “devoiced” (whispered or dropped completely) when they appear at the end of a word or between certain consonants.

Think of the common Japanese word for “is/to be,” desu (です). In normal speech, it usually sounds like “dess,” with the final ‘u’ being almost silent.

4. The Chanting Rhythm When chanting Nam-myoho-renge-kyo, the rhythm is crucial. It is chanted rapidly and rhythmically.

Trying to pronounce every syllable fully—Na-Mu-Myo-Ho-Ren-Ge-Kyo—is awkward and breaks the flow. Therefore, in the natural rhythm of the chant, the final ‘u’ sound in “Namu” is dropped, compressing the two syllables into one sharp beat: Nam.

Summary While the strict character-by-character transliteration is “Namu,” the romanization “Nam” is used because it accurately reflects the actual sound made when practitioners chant the mantra rhythmically.

I certainly hope this prioritization of the spoken Japanese over the literal Japanese doesn’t creep into the “thinking” that’s illustrating the Daily Dharma. I may chant “nam” but I want to read “namu.”