Can AI Cure the World of Bad Butter Chicken?
By Perzen Patel
Perzen Patel is an Indian food expert who went from “can’t cook rice” in 2011 to caterer to food writer. Her stories and recipes will help you cook real Indian food at home. Her Substack is called Beyond Butter Chicken, and you can also find her on LinkedIn.
Every time butter chicken makes the news, my phone starts ringing. Just as I must eat hot, freshly fried pakora—ideally potato and chili bhajiya—on a cold, rainy day, I will have something to say if you tell me your favorite Indian dish is butter chicken.
Yep, I’m your friendly, neighborhood Spiderman butter chicken police.
Recently, BBC, NPR, Bloomberg, and the New York Times have reported on butter chicken. In India, where there are at least 5.8 million pending high court cases, the Gujral family, owners of Moti Mahal (home of butter chicken), have sued rival restaurant Daryaganj, accusing them of falsely claiming to have invented butter chicken.
Monish Gujral, the managing director at Moti Mahal, has said that “You can’t take away someone’s legacy.” His quote is littered across hundreds of publications covering this case.
Here’s where I start to “laumace.” My own word: when I laugh and grimace because I don’t know how to deal with the ridiculousness of something.
My question for Gujral: What legacy?
Murgh makhani versus butter chicken
Growing up in the 90s in India, I loved nothing more than going with my family to Indian restaurants, where they had an over-the-top, sitar playing in the background, walls decked out in red and gold, while we sat on colorful cushions and stared at the large stone elephant. Murgh makhani (a.k.a. butter chicken) with buttery garlic naan was always on our order if we went to one of these restaurants. And, like any sensible person with half a brain, I loved it.
At eight, I didn’t know the chicken was yummy because of fenugreek’s umami or that soaked cashews made the tomato gravy velvety smooth just as much as the cream did. But, I remember the earthy taste of marinated, tandoori chicken lazing around in the gravy. I remember wanting mum to tear a piece of her naan so I could wipe clean the last Murgh makhani left in the serving bowl. And I remember feeling like everything was right in the world as my full tummy lulled me to sleep in the back of Dad’s Maruti 500 on the way home.
Compare that to my first butter chicken at Pakuranga Plaza in New Zealand at age 15, as a fresh-off-the-boat immigrant. My homesickness was cured temporarily as I hungrily dipped my buttery naan into some very bright, orange butter chicken (maybe it was shiny from the lights in the mall?). It tasted familiar, but strange at the same time. Later, while trying to get the orange stains off my fingers in the bathroom, I felt like I had a 5kg rock resting at the bottom of my stomach that I needed to burp out.
Murgh makhani’s celebrity makeover into butter chicken reflects the journey of the migrant cooks who made the dish popular. In the 1950s, as Indians—mainly north Indians—began migrating to countries like the UK, the US, Canada, and Australia, they took their food traditions with them. The entrepreneurial types (blocked from job opportunities by companies telling them “You don’t have experience working here in <insert country>, but we won’t give you a job so you can get experience working in <insert country>”) opened their own curry houses and takeaway joints. First, their customers were other Indians craving home. Later on, locals came.
And so began Murgh makhani’s evolution into the sweet, excessively creamy, one-dimensional yet extremely popular butter chicken we know now. Taste matters. But not as much as money.
Most Indian restaurants (in New Zealand at least) serve the same 30 dishes no matter where you go, and butter chicken tops that list. Just in small little NZ, the term “butter chicken” has an average monthly search volume of 400,000 hits, and a Gisbourne takeaway—population, 38,000—made the news in lockdown for selling 1,300 plates of butter chicken in a single day.
Now a butter chicken version of everything you can imagine. Butter chicken enchiladas. Butter chicken tacos. Butter chicken pie. Butter chicken sushi made with biryani rice! And butter chicken pizza (which tastes good so long as you’re not dipping a cheese pizza into butter chicken gravy!)
Food blogger Kalyan Karmakar describes butter chicken as an “Olympic” brand within the Millward Brown brand dynamics framework. He explains that an Olympic brand has exponential awareness and consumption. It’s unlikely that another Indian dish will challenge butter chicken’s position in the near future.
If Monish Gujral wins his court case, he could close Moti Mahal and simply sue every place serving butter chicken for ruining Murgh Makhani. He’d become a trillionaire.
What does AI know about food?
Apparently, very soon, AI will help us learn just about anything we fancy. From curing cancer to unlocking world peace. Late at night the other day, I found myself typing into ChatGPT, “How can we rid the world of bad butter chicken?” Sadly, he? she? had no good ideas.
The challenge is not about cooking a better version of butter chicken or improving technique. The challenge is also not about using better ingredients like cashews, ghee and fenugreek leaves rather than just tomato paste, garam masala, and cream. And the challenge is certainly not about establishing some sort of butter chicken protection law.
It’s about embracing diversity.
India is a country of countries. The cuisine changes every 100km you travel; you could cook a new Indian dish every day until you die without running out of options.
If AI can’t directly cure the world of bad butter chicken, what role can it play?
Change begins at home
Food is subjective. The food you like may not necessarily be the food I enjoy. We’re also lazy. And busy. This is why, despite being surrounded by thousands of recipes, cookbooks, and food apps, we still cook the same 30 dishes. On repeat. Even though they bore us to death.
But one of the ways we can use AI to go beyond butter chicken is by discovering our taste. Evan Armstrong, who has written about using AI to discover cultural taste, explains that “knowing what to make is just as important as the ability to make it.”
The same applies to Indian cooking. If you cook Indian cuisine, you likely keep making butter chicken, tikka masala, and saag aloo on repeat because you know it’s something you and the family will enjoy.
Instead of asking Claude for a better butter chicken recipe, what if you used it to discover what you specifically love about the dish? Is it the rich, creamy mouth feel? Or perhaps you enjoy the mild tomato gravy? Maybe it’s the light smokiness.
Now, ask AI for five other dishes to replicate those same taste elements.
You could also use AI to improve familiarity with spices. Imagine you only have the most basic spices at home—cumin, turmeric, mustard seeds—and a head of cabbage to use up. What could you do? Ask Claude.
The role of AI in restaurants
In the UK, Londoners are moving away from stereotypes of Indian food. The city now has Chourangi, specializing in food from Kolkata; Trishna, which provides a taste of coastal cooking; and Gunpowder, where there’s no naan on the menu! UK-based food writer Mallika Basu says, “The pattern I’ve seen is that diners first experience Indian food at a curry house, then demand elevated, more authentic versions of those dishes and finally, they are ready to experience India’s regional nuances. It’s a journey.”
Could AI help fast-track this journey?
Food apps like Delivereasy and UberEats are already using AI to get you to order more food, more frequently. What if the single-minded sales focus changed to an educational lens? Maybe the next time someone orders butter chicken, they’ll see a side recommendation for Keralan Fish Curry or Awadhi Mutton Korma. Something similar enough to be approachable, but new enough to spark curiosity.
Whenever we go out, I’m the default family member who orders for the table. To cure my decision anxiety, I always ask the waiter for their favorite dish or scan the last five dishes on the menu, where the chef sneaks in the dishes they’d like you to try.
Waiters and restaurant owners could use AI to personalize their recommendations at scale. Instead of training waiters to “recommend” their most expensive dishes, they could suggest Mughlai Chicken to customers who love creamy dishes. Or recommend a new kind of Dahl to customers that order Dahl on repeat.
My favorite part of my catering business in India was partnering with hotels and restaurants to run Parsi food festivals. Not the ones who just wanted me to serve Dhansak and Patra ni Machchi—the butter chicken equivalents of Parsi food. Instead, I hosted festivals on rare Parsi wedding treats and another focused on a Parsi vegetarian breakfast.
AI could analyze sales data and assist our local takeaway to highlight regional or seasonal dishes. Imagine. At a takeaway where fried dishes are bestsellers, you might find a “Monsoon Flavors” menu. They’ve got your favorite “onion bhajji” but you can also try Bread Pakora, Vada Pao and Banana Bhajji’s now.
Or, at a restaurant where South Indian dishes perform better, chefs can use AI to introduce new South Indian dishes that still align with customers’ taste preferences. It can also then help curry house cooks, typically with limited knowledge of regional Indian food, find a good recipe.
Yes, radioactive orange butter chicken could be kept warm in a chafing dish for customers jonesing for their butter chicken fix. However, adventurous customers could go beyond their usual orders and start exploring India’s food diversity.
AI can help food media shift the conversation
If I see a recipe for Saagwala or Butter Chicken that mentions garam masala and no other spice, I intuitively know a white journalist or some Instagram celebrity wrote the column. Same with Dahl Makhani made with the wrong type of lentil (the title, “lentil curry,” is the clue). What we cook next Sunday can change depending on the dishes the mainstream media chooses to focus on. Who writes these recipes also matters.
To cover the butter chicken feud, the Asia New Zealand Foundation sponsored a NZ-based food writer to try the “original” Murgh Makhani at Moti Mahal. But during their taste test, rival Daryaganj won. I didn’t mind who the winner was until I read why.
It was because, “Their [Daryaganj] curry had boneless chicken chunks doused in a rich, sweeter butter curry. It was similar to those I’ve tried at Indian restaurants back in New Zealand.” This is not an isolated case. Author Nayantara Dutta succinctly shares my frustration. “In the West, people view Indian food through the lens of takeout". Since the media mostly echoes what the public wants (or thinks they want), the stuff that gets published has to be “the right level” of ethnicity. It can’t be “too brown” or “too white.”
But perhaps, AI can cure these writers and editors from their butter chicken obsession.
What if food editors used AI algorithms to detect patterns in user behavior, like the growing popularity of certain ingredients or seasonal recipes? Instead of 1281 columns on turmeric latte being the next super-food, they could commission an Indian food writer to explain why turmeric is always used in Indian marinades or create a listicle about Indian sweets featuring turmeric.
Larger outlets with bigger budgets could explore content personalization that challenges stereotypes about Indian food. By analyzing a reader’s browsing history and recipe interactions, the algorithm might suggest articles that are slightly outside their comfort zone but still aligned with their taste.
Dish NZ’s readers who have clicked on their Dahi Puri recipe more than five times might see an article about India’s ten favorite chaat dishes, an article discussing regional nuances between Dahi Bhalla and Dahi Vada or, if they are health food inclined, a video about Sweet Potato Makhana Chaat that’s more protein-rich. This stepping-stone approach would help readers explore new flavors or dishes in a way that feels approachable while broadening readers’ palates.
Who will win the butter chicken feud?
The high-court case about butter chicken isn’t the first food dispute in India. In 2018, the Delhi High Court heard a case where two kebab shops were vying for the brand name “Tunday Kebab,” a lightly-spiced kebab made of minced meat that has almost become synonymous with the city of Lucknow, India. Closer to home in New Zealand, it’s not officially summer until I argue with a friend or colleague from Australia about the origins of Pavlova.
I don’t know who will win the butter chicken feud. Given how slowly justice is served in India, my word laumace might have entered the Oxford dictionary before then.
Instead, I’m counting on all of us who are curious about AI. Collectively, we can use AI to make butter chicken less famous.
Maybe, just maybe, on the day the court announces the resolution restaurants will not make big vats of radioactive orange butter chicken. Nor will the media share another butter chicken recipe featuring chicken breast and garam masala.
And maybe, just maybe, you’ll laumace upon reading the headlines. Before heading into the kitchen to roll out some Aloo Paratha orwhip up a Parsi Akoori.