Towards the end of last year, the JHB Incogvino team got to hang out with the geeks at IBM who were showing off their latest Watson trick. Watson, if you haven’t heard about it before, is the first stab at Artificial Intelligence. A very, very clever computer that can do some pretty interesting things by crunching enormous sets of data.

The Chef at The Griffin telling the story of what it's like to get his instructions from a computer...

The Chef at The Griffin telling the story of what it’s like to get his instructions from a computer…

You may remember when Watson won Jeopardy.

From Wikipedia:

Watson is an artificially intelligent computer system capable of answering questions posed in natural language,[2] developed in IBM’s DeepQA project by a research team led by principal investigator David Ferrucci. Watson was named after IBM’s first CEO and industrialist Thomas J. Watson.[3][4] The computer system was specifically developed to answer questions on the quiz show Jeopardy![5] In 2011, Watson competed on Jeopardy! against former winners Brad Rutter and Ken Jennings.[3][6] Watson received the first place prize of $1 million.[7]

Watson had access to 200 million pages of structured and unstructured content consuming four terabytes of disk storage[8] including the full text of Wikipedia,[9] but was not connected to the Internet during the game.[10][11] For each clue, Watson’s three most probable responses were displayed on the television screen. Watson consistently outperformed its human opponents on the game’s signaling device, but had trouble responding to a few categories, notably those having short clues containing only a few words.

Enter Watson’s next trick. It can cook. Which led to a fascinating experience… First of all, pairing wine to a bunch of recipes generated by a computer must have been interesting in itself, but probably the most interesting part of “Dining with Watson” was how hard it was for the Chef (in this case from The Griffin in Illovo) to put together these dishes. Some of them went against every trained cooking bone in his body. And yet… they tasted great!

Watson’s recipe generation repertoire is generated off a couple of factors.

  • Primary ingredient
  • Primary style (eg. Indian, Spanish etc.)
  • Should the recipe be completely unique

Unique recipes is where it gets a bit crazy. Combinations of ingredients and styles that never, ever would make sense in your head. But I guess that’s the fun bit. And how you end up with Swiss Thai Quiche… or Belgian Bacon Pudding.

I’m going to post each recipe separately, but definitely have a look at the wine pairing below. Then follow the links below and try your hand at some computer generated cooking. Let us know how it goes!

IBM Watson’s recipe for Swiss Thai Quiche

IBM Watson’s recipe for Deboned Loin of Lamb with Pinenut Crust

IBM Watson’s recipe for Belgian Bacon Pudding



3 comments on “IBM’s Watson can cook (quite well actually)…

  1. Pingback: Deboned Loin of Lamb with Pinenut Crust. A recipe from IBM's Watson #CognitiveCooking - Incogvino

  2. Pingback: Swiss Thai Quiche. A recipe from IBM's Watson #CognitiveCooking - Incogvino

  3. Pingback: Belgian Bacon & Mushroom Pudding. A recipe from IBM's Watson #CognitiveCooking - Incogvino

Leave a reply