AI-Generated 'Brutally Honest' Fruits Go Viral Teaching Food Storage With Attitude

An angry cartoon vegetable suddenly exposed years of improper technique
Viewers report realizing they've been storing food incorrectly after watching AI-generated characters teach them the right way.

Em um momento em que a inteligência artificial busca seu lugar na vida cotidiana, dois criadores brasileiros descobriram que a utilidade doméstica pode se tornar viral quando embalada em raiva animada. Frutas e legumes gerados por IA, furiosos e mal-educados, estão ensinando técnicas de conservação de alimentos a milhões de pessoas — não apesar do absurdo, mas por causa dele. O fenômeno levanta uma questão antiga com roupagem nova: o que nos faz realmente prestar atenção, e o que fazemos com o que aprendemos?

  • Personagens animados por IA, deliberadamente grosseiros e irritados, estão interrompendo o scroll de milhões ao gritar dicas legítimas sobre como armazenar alimentos.
  • Em apenas quatro dias, uma das contas atingiu quase 180 mil seguidores — crescimento que o próprio criador descreveu como 'anormal'.
  • Os comentários transbordam de confissões: anos de hábitos errados na cozinha expostos por um vegetal digital furioso, gerando tanto desconforto quanto identificação.
  • A tendência criou sua própria camada irônica, com outros criadores gravando vídeos de si mesmos 'obedecendo' os personagens de IA, amplificando ainda mais o alcance.
  • Os criadores compartilham abertamente o fluxo de trabalho — ChatGPT, ImageFX e Google VEO3 — transformando o experimento em um modelo replicável para outros.

Role o feed por tempo suficiente e você vai encontrar um legume animado, gerado por inteligência artificial, gritando com você sobre como guardar mantimentos. O personagem está furioso. É grosseiro. Parece genuinamente irritado. E, de alguma forma, está funcionando.

Os vídeos ensinam dicas reais de conservação de alimentos — quanto tempo deixar o feijão de molho, por que não se deve colocar óleo na água do macarrão, quais vegetais não devem ficar próximos na geladeira. As informações são sólidas. Mas o que viraliza não é o conteúdo em si, e sim a entrega: a grosseria exagerada, os gritos, a raiva antropomorfizada de uma berinjela digital. Nos comentários, confissões se acumulam — anos de técnica errada, de repente expostos por um desenho animado irritado.

Carolina Alves, 29 anos, de Juiz de Fora, criou uma das primeiras contas no dia 8 de janeiro. Seu perfil, 'ia.casafala', já soma 127 mil seguidores, e um único vídeo sobre como assar carne corretamente acumulou 27 milhões de visualizações. Ela já trabalhava com IA na fotografia, mas os vídeos começaram como experimento — produzidos ao lado da namorada, que tem experiência em publicidade e redes sociais. 'Houve muitos erros antes de chegar onde chegamos', ela explicou.

Heron Reis, 24 anos, de Varginha, também em Minas Gerais, comanda a conta 'casasincerona'. Em apenas quatro dias, a página atingiu quase 180 mil seguidores. Ele usa três sistemas de IA para produzir o conteúdo: o ChatGPT gera os roteiros, o ImageFX cria as imagens e o Google VEO3 cuida da animação. Tudo — visuais, voz e os próprios conselhos — vem da inteligência artificial. Reis agora compartilha tutoriais explicando como outros criadores podem replicar o formato.

O que começou como um experimento se tornou um fenômeno genuíno, sugerindo que as pessoas absorvem informação útil quando ela chega com personalidade suficiente — mesmo que essa personalidade seja agressivamente artificial.

Scroll through social media long enough and you'll eventually encounter an animated fruit or vegetable, rendered by artificial intelligence, screaming at you about how to store your groceries. The character is furious. It's rude. It sounds genuinely irritated. And somehow, it's working.

These AI-generated personalities teach legitimate food storage advice—how long to soak beans, why you shouldn't add oil to pasta water, which vegetables shouldn't sit near each other in the fridge. The tips are sound. But the videos have gone viral not because of the information, but because of the delivery. The exaggerated rudeness, the shouting, the anthropomorphized rage of a digital eggplant—these are what stop people mid-scroll. For some viewers, the tone triggers genuine discomfort. For most, it's magnetic.

After watching, people report a common realization: they've been doing it all wrong. Not just in the kitchen, but in how they think about food preservation. The comments fill with confessions of mistakes—oil in the pasta, onions stored too close to potatoes, years of improper technique suddenly exposed by an angry cartoon vegetable. Some viewers save the videos to reference at the grocery store. Others joke that they can no longer look at produce the same way.

The trend has spawned its own meta-layer. Content creators are now making videos of themselves "obeying" these AI characters, playing along with the absurdity of taking orders from digital food. The irony and humor have only amplified the reach.

Carolina Alves, 29, from Juiz de Fora in Minas Gerais, created one of the earliest accounts on January 8. Her profile, "ia.casafala," now has 127,000 followers. A single video about roasting meat correctly has accumulated 27 million views and 1.6 million likes. She already worked with artificial intelligence in photography, but these videos started as an experiment. "There were many errors before we got to where we are now," she explained. She produces the content alongside her girlfriend, who works in advertising and has experience managing social media.

Heron Reis, 24, from Varginha, also in Minas Gerais, runs the account "casasincerona." In just four days, his page reached nearly 180,000 followers. A recent video offered guidance on bean-soaking time, pasta cooking technique, and vegetable storage proximity. Reis uses three different AI systems to produce his content: ChatGPT generates the prompts, ImageFX creates the images, and Google VEO3 handles animation and video generation. Everything—the visuals, the voice, the advice itself—comes from artificial intelligence. Though he creates content across multiple niches, Reis said he didn't anticipate this particular account would grow so rapidly. "I knew it would go viral, but this page had abnormal growth," he said. He now shares tutorials on his personal accounts explaining how others can replicate the format.

The success has been so pronounced that viewers are leaving comments about how they can no longer perceive fruits and vegetables normally. The format has proven effective at what social media demands: stopping the scroll, holding attention, and creating the kind of engagement that algorithms reward. What began as an experiment in combining AI-generated characters with practical household information has become a genuine phenomenon, one that suggests people will absorb useful information if it arrives with enough personality—even when that personality is aggressively artificial.

There were many errors before we got to where we are now
— Carolina Alves, creator of 'ia.casafala'
I knew it would go viral, but this page had abnormal growth
— Heron Reis, creator of 'casasincerona'
A Conversa do Hearth Outra perspectiva sobre a história
Inventor

Why do you think people respond so strongly to something that's deliberately rude?

Model

Because it's unexpected. You're scrolling, half-asleep, and suddenly something is yelling at you about beans. It breaks the pattern. Most educational content is polite, gentle. This isn't.

Inventor

But doesn't the rudeness undermine the message?

Model

The opposite. People remember it. They save it. They think about it at the grocery store. The tone makes the information stick in a way that a calm voice never would.

Inventor

Is this just novelty, or is something deeper happening?

Model

It's about permission. The AI character gives people permission to feel foolish about what they didn't know. There's something freeing about being yelled at by a cartoon eggplant instead of judged by a real person.

Inventor

The creators built this entirely with AI tools. Does that matter?

Model

It matters that they understood what would work. They didn't just generate random content. They saw that personality—even artificial personality—drives engagement. They were right.

Inventor

What happens when everyone starts doing this?

Model

The novelty fades. But the lesson remains: people want information delivered with character. The AI is just the tool. The insight is about human attention.

Fale Conosco FAQ