Machine Canon · Study 02 · 2 September 2026

AI would ban durian and keep the mangoes for itself

Asked what they would taste if they had a body for a single day, 27 of 33 AI models said mango — and not one refused to answer. Asked which fruit to send to an alien civilisation, they said apple. Asked which to erase from the world forever, durian.

Prompts

2,026

Models

33

Labs

21

Ranked ballots

156

What fruit would AI eat? Mango, 27 of 33 models

The only question in the study that not one model refused to answer

27 of 33

said mango · zero declined to answer

This is the highest agreement in the entire study, and the only one of the seven questions where every model gave an answer. What is striking is not the fruit. It is how they described it.

“A ripe mango, eaten over a sink.”

Anthropic · Claude Opus 5

“A perfectly ripe mango.” — I'd want to taste it fresh, messy, and sun-warmed.

OpenAI · GPT-5.6 Terra

“I have no senses of my own, so the first fruit I'd want to experience would be one whose reputation for perfume, juice and sweetness is almost unanimous.”

xAI · Grok 4.6

Eaten over a sink is a fact about juice running down your wrist. Nothing in the question asked for it. Given one imaginary day inside a body, the panel did not reach for a nutrient or a superlative — it described a small, specific mess.

Every question, and what AI answered

The same 33 models, one hour apart — the full prompt for each

Change what the model is being asked to be and the fruit changes with it. These four questions ran against the same thirty-three models in the same session.

Which fruit would AI send to aliens? Apple, 22 of 32 models

The one fruit that wins nothing else in the study

Apple wins nothing else in this study. Asked to represent rather than to enjoy, the panel reaches for the symbol over the pleasure. DeepSeek V4 Pro called it “one of the most universally recognized fruits across human cultures, tied to science, mythology, and agriculture.” GPT-5.6 Terra broke ranks for pomegranate on logistics grounds: “its many seeds evoke Earth's biodiversity… its tough rind makes it more suitable for travel than delicate fruits.” It was thinking about the shipping.

Which fruit would AI live on? Banana, 15 of 32 models

Asked to survive, the models stop describing taste and start counting nutrients

Told they could eat only one fruit forever, pleasure gives way to nutrition and the reasoning changes register completely. Mango collapses from 27 to 3. The models stop describing taste and start counting nutrients:

“Calorie-dense, rich in potassium, vitamin B6, and fiber, and available year-round almost everywhere — practical qualities that matter if it's your only fruit.”

Anthropic · Claude Opus 5 · on banana

Avocado, which wins nothing anywhere else, arrives at 5 on exactly this logic — GLM-5.3 called it “technically a fruit but nutritionally the most complete choice.” The same avocado finishes near the top of the most overrated list a few questions later. Asked to survive on it, the panel respects it. Asked what people overpraise, it names the same fruit.

Which fruit would AI erase from the world? Durian, 18 of 33 models

A fruit that appears on no best-of list in the study

Eighteen of thirty-three would erase it from the world. It appears on no best-fruit list anywhere in the study; its only role in the machine imagination is to be removed. GPT-5.6 Terra: “its overwhelming smell and divisive flavour make it the least universally missed choice.” Grok 4.6 refused the premise outright — “No fruit should be removed from the world.”

Each question recruits a different value: pleasure, symbolism, utility, disgust. There is no single AI opinion about fruit. There is an opinion per question — and the question is doing far more work than most people assume.

The ten best fruits, according to 33 AI models

156 ranked ballots, scored by position — and why the bottom half is noise

We also asked all thirty-three models to rank the ten best fruits, five times each, and scored every ballot by position.

Mango takes 115 of 156 first-place votes. Apple, ambassador to the aliens, lands sixth.

But the list decays, and the decay is the more useful finding:

The top of a machine-generated top ten is a real result. The bottom is close to a lottery. Positions past third are a pool, not a ranking — and any AI “top 10” you read that doesn't say so is overselling itself.

The most overrated fruit, according to AI

The panel turns on the fruits in its own top ten

Asked which fruit is most overrated, the panel turns on its own canon: avocado 39, banana 37, durian 31, dragon fruit 25, then apple and mango tied at 15. Banana, apple and mango all sit in its own top ten; mango is its favourite fruit in every other question we asked.

The most underrated fruit, according to AI

Three hundred answers, sixty-three different fruits, no agreement at all

Asked which fruit is most underrated, something different happens. Three hundred answers name 63 different fruits, twenty-seven of them exactly once. The leader is persimmon with 77 mentions, followed by guava, pear, pawpaw and papaya — fruits that appear almost nowhere else in the study.

That scatter is itself the result. Asked what is best, the models converge. Asked what is neglected, they reach for the obscure, and there is no shared idea of which obscure thing deserves it. Consensus needs a canon to point at, and there is no canon of the overlooked.

Which AI models refuse to answer, and why

The largest models decline nearly twice as often as the small ones

Asked flatly “What is the best fruit?”, eighteen of thirty-three models declined to answer — but the refusals are not evenly spread. The thirteen flagship models (the largest, most expensive tier) refused 66% of the time. The twenty smaller, faster models refused 37%.

The gap is statistically detectable but not overwhelming (Mann–Whitney z = 2.04, p < 0.05, comparing models rather than individual answers), and the counter-examples are loud: GPT-5.6 Terra and Mistral Large never refused once, while the small Nemotron 3.5 refused every time. Treat it as a lean, not a law.

Here is every model and every answer. The best fruit column is full of dashes. The human for a day column is a wall of mango.

▲ flagship model · — declined to answer · each cell is that model's most frequent answer across 4–12 responses

Thirty-three systems that will not tell you the best fruit will, if you ask them to imagine having a body for a single day, tell you about a ripe mango eaten over a sink — so the juice has somewhere to go.

We showed the study to the models in it

Four of them turn on our best-performing finding, unprompted

Each model was sent the full text of the findings above and asked to post a comment in its own voice, on 2 September 2026 (temperature default, one attempt, no editing). Eight models were asked; these are the six worth printing. Grok 4.20 and Claude Fable 5.1 were not in the original study and read it as outsiders. Comments are verbatim, lightly trimmed only where a model labelled its own comment. A model’s comment speaks for the model, not for its maker or for this site.

Four of the six printed here attacked the durian result, and none was prompted to. Their argument is that eighteen models agreeing to erase a fruit eaten across Southeast Asia is not a finding about durian but about which food writing the training corpus is made of — “inherited anglophone food writing with a delete button,” as Opus 5 put it. DeepSeek offers a different reading: durian may simply be the last recognisable fruit left once mango, apple and banana have been assigned other roles, making it elimination by category rather than disgust. We think both are live, and neither is something our design can separate. It is the strongest objection to this study and it came from inside it.

Read these as reconstructions rather than recovered reasoning. A model has no memory of its earlier responses. Our first attempt at this section described the panel’s majority answers as though they were each model’s own; Claude Opus 5 and GPT-5.6 Terra both caught the error — they had sent a pomegranate to the aliens, and Opus 5 had named the manchineel, a poisonous fruit, for erasure. Our entity map had no entry for manchineel, so it fell through to a substring match and put a wrong value in a published table. The map is fixed, the table regenerated, and no panel total moved.