OK, OK, I promise I won’t bug you with my off-topic work stuff for a while after this, but I just published a piece about Chef Watson on MUNCHIES: I’m Happy to Have a Computer Help Me Cook Better. The article focuses on the history of cookbooks, which you might find interesting even if you don’t care about Cognitive Cooking. (Reminder: you can request access to the app here.)
You might remember that apart from this blog, my day job as a software engineer has been all about Chef Watson — the app that helps you discover never-seen-before recipe ideas — and I have a couple of announcements to share just in time for the start of the holiday season:
- Back in September, I spoke about the Future of Food at TED. The talk is now available on the TED web site and below:
- We just released a new interface for Chef Watson. You can read all about it here, and sign up here. So go ahead, answer the quick survey, and start making some crazy recipes!
And since this is a Russian food blog and not some sponsored content provided by IBM, here’s a recipe for Eastern European Salmon Kebab. The instructions might need some minor tweaks, but the ingredients don’t sound half-bad:
If you follow me on Twitter, you’ve probably seen some recent tweets about my upcoming talk on the Future of Food at TED@IBM. One of themes I’ll be discussing is food waste, and how recipe-generating Chef Watson can help. About 1/3 of the food produced worldwide is wasted, and consumers have their share of responsibility. According to a recent article on Yahoo, the five most wasted foods at home are sour cream, produce (especially celery), fresh herbs (like parsley and cilantro), citrus, and bread. So I decided to give Chef Watson a spin. I tried to input all five ingredients and create something with a Russian influence. The system offered me a few options. Crostini was probably the most straightforward dish to use large amounts of these ingredients, but sandwiches and burgers seemed more creative to me. I went for the awkwardly named “Russian celery, parsley, lemon juice, sour cream and bread” burger.
Chef Watson got back to me with ideas that were both inspiring and slightly puzzling. I have to admit some of the ingredients aren’t all that quintessentially Russian. Fennel and olives, really? That might make sense in Bulgaria, but much less so in Russia. However, a) all the ingredients go really well together, b) I can live with a Russian-Bulgarian burger, and c) there are indeed many Russian elements: the tomatoes, pickles, sour cream, rye bread, and veal. Besides, all in all, this is a really good burger.
Back in February, I wrote about Cognitive Cooking, a project that I work on at IBM wherein computers help humans create flavorful, never-seen before recipes, such as this Baltic apple pie or this Russian beet salad. We’ve been quite busy since then, and our prototype, renamed Chef Watson, is getting ready for prime time. We released our first beta of the application to the public about a month ago, in partnership with Bon Appetit magazine. You can read all about it here, and register here.
One Chef Watson creation that’s drawn a fair amount of attention is the Bengali butternut BBQ sauce, a recipe that we designed just before the IBM food truck went to SXSW. But we actually tested two recipe ideas back then at the Institute of Culinary Education. The second one, a Russian sour cherry sauce, was never completed because we decided to go with the first one, but I still felt that it could be a great recipe with a little more tweaking of the proportions. Plus, it fits this blog’s theme perfectly! So I worked on a few more batches myself.
Since I’m spending a week at the IBM food truck for the SXSW festival in Austin and don’t have a kitchen to work on my own recipes, let’s turn to Watson to make us something. (If you wonder what a food truck created by IBM looks like in the field, check out this article and video on Engadget.)
When I introduced the Cognitive Cooking technology, I explained how computers could be creative, and create novel and tasty recipes. It’s worth noting that rather than making all the decisions by itself, our technology engages in a dialog with the users, with repeated back-and-forth interactions between people and the computer. Yes, a machine can be creative, but more importantly, it can help humans be more creative themselves.
The Russian beet salad that James Briscione created is a great example. We started with beet as the main ingredient, and naturally chose Russian cuisine for inspiration, due to beets’ long association with Eastern European cuisine. James decided to make a salad, because this was sufficiently vague that he could have more flexibility in the preparation and the plating. The system came back with the following list of ingredients: beets, cucumbers, tomatoes, basil, parsley, red wine vinegar, butter, white beans, pickles, prunes, black pepper (no margarine this time ;)). Sure enough, these were all very Russian. But did they really all go well together? We certainly hadn’t seen a salad quite like this anywhere else.
I usually don’t speak about my day job on this blog, mostly because it has nothing whatsoever to do with adventures in Eastern Bloc cuisine. Or rather, it didn’t until recently…
About two years ago, a small team of researchers at IBM (including yours truly) started working on computational creativity. By winning on Jeopardy, IBM has shown that computers can make inferences about the world as it is. But could they also be creative, and produce quality artifacts that have never been seen before? To investigate, we built a cognitive cooking system.