What’s for dinner, Chef Watson?


It’s the question that people ask almost everyday. We’ve all stared into the abyss of our cupboards and fridges before pondering this.

But the act of cooking is a science in itself. Forming dishes out of dishes in a cupboard can be a potentially risky experiment. How can technology help us formulate new boundary-pushing revelations in the kitchen? Will this 3-day-old bolognese sauce compliment a random packet of udon noodles found in the pantry? Without the knowledge of past failures, your natural instincts may tell you that perhaps it’s best to keep them separated. But what if you were onto something?

IBM’s supercomputer Watson may be able to help you for all your future creations thanks to a collaboration between IBM and the US website Bon Apetit. Enter Chef Watson, a beta-stage cooking app that will be your new kitchen assistant.

Using the knowledge of what ingredients are likely to compliment other ingredients via items commonly associated together from the massive database of Bon Appetit’s recipe collection.

This recipe generating platform is currently in beta, but of course, I’ve to sign up and play around.

“Chef Watson is a web app built on IBM’s cognitive cooking technology and trained on Bon Appétit’s database of 9,000 recipes. With Bon Appétit’s approach to dishes and ingredient combinations, and Chef Watson’s own understanding of taste preferences and food chemistry, the app is able to combine ingredients to discover millions of novel recipes out of quintillions of possibilities.”

Combining knowledge of ingredients, dish composition, common pairings and styles. Food chemistry and the psychology of people’s likes and dislikes.

Creating your next dish is easy.

Simply enter “what” – an ingredient, a dish – anything.

There are three main outputs: ingredients, dishes (i.e. sandwiches, taco, seared) and styles (think category/cuisines).

It’s incredible smart. You can personalise search results with the categories:

  • Pinned it (saves combination of searched ingredients for you to explore new options in the future)
  • Have it (ingredients that you have on hand already)
  • Need it (shopping list)
  • Love it (for easy future reference of your favourite ingredients)
  • Hate it (exclude for all future dishes)
  • Saved it (your personalised recipe creations that you’ve wanted to save)

There is sliding scalebar at the top, “how crazy do we want to get?” (a dangerous option for someone like me that likes to err on the risky side of things). It allows the user to decide if they want to “keep it classic” or “surprise me”.

A simple entry of bacon brings up unlimited options.

As you enter an ingredient, the suggested ingredients, dishes and styles change based on database knowledge. As it changes, the different options are there to help you spark ideas and streamline the dish that you want to create. This is how Chef Watson helping you decide what to make for dinner!

On the side bar “Inspiration Station”, Chef Watson provides recipe suggestions using your search terms. It spits out one suggestion for each item that you add so that you end up with multiple options to choose from. Upon deciding your recipe, it brings you to a new page that makes it feel like you’ve done the hard work yourself – “discovered by Watson, created by you”.

Some clever food marketer which some background in psychology has definitely been involved. It’s a positive user experience.

Since it’s US-based, some cultural terminology brings up different results. Mincemeat refers to the citrus kind. If you want minced meat, you have to search “ground” beef/pork/whatever. A measurement converter between metric and imperial system would be handy too.

A disclaimer at the bottom reminds you to use discretion. “Remember that Chef Watson eats data, not real food. The ingredients and steps are suggestions, so be sure to use your own judgement when preparing these dishes. And, give us feedback to make the Chef smarter.” Sometimes Chef Watson gets it wrong. Just in case you blindly think that it’s a great idea to attempt to roll a whole raw turkey – mostly because the recipe is based off “Lemon- and Prosciutto-Stuffed Pork Loin Roast with Broccolini”. More or less substituting the pork loin roast with a 4-pound turkey.

Now if there was an actual Chef Watson that cooked food for you when you’re too lazy to cook…

Here’s a snapshot of a recipe that I created on Chef Watson – and trialed! It didn’t offer much in the way of presentation (so no photo) but it was tasty!

IBM have tested their own creations with professional chefs and you can see what they’ve already cooked up so far on Cognitive Cooking.

They’ve also just realised a cookbook called Cognitive Cooking with Chef Watson: Recipes for Innovation from IBM & the Institute of Culinary Education. It contains 65 recipes and is available in hardcover from April 14.

You can sign up here: https://www.ibmchefwatson.com/


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s