IBM’s Watson supercomputer tackles the problem of creativity in cooking

IBM’s supercomputer research team has been pushing the boundaries of what computers are capable of. Their Deep Blue first beat Kasparov at chess almost 20 years ago. Later, IBM’s Watson beat Ken Jennings at Jeopardy. And now the same Watson is tackling the most difficult problem yet – cooking and creativity.

IBM Research has begun work on an unnamed cyberchef, an AI system designed to create new dishes that can delight our palates at their theoretical peaks of enjoyment.

Why bother teaching a computer how to cook? Of course, the first reason is that it advances computer science. But there is another interesting angle. The supercomputer could cook up recipes that have never been seen by man, because there are some things that computer are just better than humans at:

For example:

In the case of the flavorbot, these “new things” IBM is after range from spotting underrated, highly flavorful ingredients (like black tea, bantu beer and cooked apples), strange-but-tasty flavor pairings (like white chocolate and caviar, jamaican rum and blue cheese, or even bell pepper and black tea), and even whole recipes, complete with basic preparation steps.

And how does Watson do this? Unsurprisingly, this is rather difficult. More interestingly, lots of science and maths comes into play here.

This is a high level description of what the AI needs to do:

To generate these food leads, if you will, AI cross references three databases of information:

  • A recipe index containing tens of thousands of existing dishes that allows the system to infer basics like “what makes a quiche a quiche”
  • Hedonic psychophysics, which is essentially a quantification of whether people like certain flavor compounds at the molecular level
  • Chemoinformatics, which sort of marries these two other databases, as it connects molecular flavor compounds to actual foods they’re in

And here is a journalist’s article about the results; he was sent a bottle of Watson’s Bengali Butternut Barbeque Sauce and this is what he found:

When I unwrapped the brightly colored box and found the bottle inside, I immediately flipped to the back label. Most BBQ sauces start with ingredients like vinegar, tomatoes, or even water, but IBM’s stands out from the get go. Ingredient one: White wine. Ingredient two: Butternut squash.

The list contains more Eastern influences, such as rice vinegar, dates, cilantro, tamarind (a sour fruit you may know best from Pad Thai), cardamom (a floral seed integral to South Asian cuisine) and turmeric (the yellow powder that stained the skull-laden sets of True Detective) alongside American BBQ sauce mainstays molasses, garlic, and mustard.

I pour a bit of the bottle onto a plate of roasted tofu and broccoli–even a pork lover has gotta watch his cholesterol–and tentatively took a bite. Watson’s golden sauce may have the pulpy consistency of baby food, but it packs a surprising amount of unique flavor.

Immediately, you can taste the sweet warmth of the wine and the squash. The tamarind blends seamlessly, backed by a duo of vinegars, to tickle your tongue with just the right amount of tartness. The other flavors combine to leave an indefinable, warm aftertaste that, as you have a few more bites, actually heats your mouth–thanks to Thai chiles

This resulted in a reddit discussion, and one of the people working on this project showed up to share details of how exactly it works:

In a nutshell, however, Watson consumes massive amounts of recipes from different sources and then parses out the ingredients and steps. It also takes in information about the basic flavor compounds in ingredients, the general nature of ingredients, and, perhaps most interesting, a database of the “pleasantness” of flavor compounds, and a few other things that really make up Watson’s “secret sauce”.

From there it’s a collaborative creative process between chef and watson. It typically starts with an ingredient. Let’s say “cardamom”. Watson then searches the database, which is a pretty straight forward process, for the types of cuisine that have that ingredient. For cardamom there are about 100 different cuisines from Indian to Swedish to Bhutani and Barbadian that have a recipe somewhere that uses cardamom. Next it searches through the recipe database to pick out recipes that have cardamom in it. Cardamom is most often found in soups and cake, but it also can be found in things like fudge, baklava, and kebabs.

In the next step Watson starts to create a template of what it thinks might go in Swedish/Barbadian fudge with cardamom. Here’s where you can go crazy with Watson. The most common elements are automatically selected, but there’s lots of other options. For example, most fudge has a sweetener, chocolate, dairy, oil, and some nuts. Because we wanted cardamom, Watson recommends some spices too. You can go crazy and add in things like meat, alcohol, cheese, and a variety of other things at this step. You can’t just add in anything you want because there are some things that Watson has a hunch will just turn out to be nasty.

In the final step Watson generates a number of recipes that meet the guidelines provided. It tries to ensure that the ingredients selected match up with the various cuisines and also with the dish selected. In addition, using some of the “secret sauce” it makes sure that the ingredients will taste good together too. At the end it presents a number of recipes rated on scales such as “surprise”, or how rare is recipe like this compared to the database, “pairing”, or how well do the flavors pair or contrast with each other, and “pleasantness” which is based on the science of hedonic psychophysics. From there the chef works with Watson to find the best recipe.

That final paragraph sounds so cool. You randomly suggest a bunch of ingredients to a supercomputer and it comes up with interesting recipes based around your rough guidelines, while all the time preventing you from totally screwing up and ensuring that the resultant dish will taste good. (But, this is also reminding me of Arthur in the Hitchhiker’s Guide to the Galaxy and his struggles to get Eddie, the shipboard computer, to make him some tea.

References:

50% People “remember” false stories that never happened if they support their preconceived ideas

Remembering is believing, right?

Not really.

We have always known that people remember things that they agree with, and forget things that are inconvenient for them. However, here is some research that proves that people can “remember” completely made up events.

Here is the scary abstract of the research paper:

In the largest false memory study to date, 5,269 participants were asked about their memories for three true and one of five fabricated political events. Each fabricated event was accompanied by a photographic image purportedly depicting that event. Approximately half the participants falsely remembered that the false event happened, with 27% remembering that they saw the events happen on the news. Political orientation appeared to influence the formation of false memories, with conservatives more likely to falsely remember seeing Barack Obama shaking hands with the president of Iran, and liberals more likely to remember George W. Bush vacationing with a baseball celebrity during the Hurricane Katrina disaster. A follow-up study supported the explanation that events are more easily implanted in memory when they are congruent with a person’s preexisting attitudes and evaluations, in part because attitude-congruent false events promote feelings of recognition and familiarity, which in turn interfere with source attributions.

Let me repeat the most important points for effect:

  • The study involved 5000+ participants, so there is little chance of it being a few weirdos. This probably applies to “most of us”
  • Half the people “remembered” an even that had never taken place
  • 27% of the people remembered “seeing” the event on the news – an event that never took place
  • People were more likely to “remember” false events that agreed with their preconceived notions / political leanings

With the rise of social media putting an increasingly harsh spotlight on every action by every political leader, can we feel happier that the truth is more likely to come out? I would argue that social media just makes it easier to manipulate people…

Check out the original paper if you’re interested, and the related reddit discussion.

Data about birth-control in India: Myths and Realities

The Asian Age has an interesting article with data on India’s birth-rates and efficacy of our birth-control programs

Here are some interesting excerpts:

good news is that the increase in contraceptive prevalence has been larger and faster among illiterate and uneducated women than those with schooling.

According to the International Institute of Population Sciences (EPW Arokiasamy 2009), more than two fifths of the reduction in Total Fertility Rate country-wide is attributable to illiterate women. The study calls it “remarkable demographic behaviour which has given significant direct health benefits to women and children — almost equal to what educational improvement has done for progress in human development.”

But all is not good:

Now some disappointments: States which continue to lag behind are the same — Bihar, Uttar Pradesh, Madhya Pradesh, Jharkhand, Chattisgarh and Rajasthan — some 284 problem districts account for nearly half India’s population and 60 per cent of the yearly births countrywide.

and:

Among 18 to 24-year-old couples the contraceptive prevalence rate is not even 19 per cent. In many districts it is as low as 10 per cent. According to NFHS -3 and the latest Annual Health Survey, in Bihar more than half the women in the child bearing group are not using any family planning method.

and the worst part is:

In India, female sterilization continues to be the most dominant method of birth control even though women overwhelmingly favour non-invasive options.

because:

In the absence of tools that do not depend on partner-co-operation (condoms) or adherence to rigid regimens (pills), a poor woman confronts the prospect of an unwanted pregnancies every month, until somebody agrees to escort her for an operation.

Do Muslims procreate much more than Hindus? Apparently, Muslims are a little worse off in this aspect, but not as bad as is widely believed. Here is the data:

That brings one to a widespread myth relating to the practice of contraception by religion. Professor P.M. Kulkarni at JNU who has researched differentials in population growth among Hindus and Muslims (using NFHS data) says that all religious communities have experienced substantial fertility decline and contraceptive practice has been well accepted by all. Within religious faiths, 85 per cent of Hindu women would like to limit the family to two children whereas in the case of Muslim women, the figure is 66 per cent.

and:

The belief that religion and religious fiats discourage contraception among Muslims is not borne out by statistics.

An even more significant aspect of his analysis of NFHS data shows that the unmet need for family planning is one and a half times more among Muslim women than Hindu women.

Another interesting aspect is that the kinds of contraceptives preferred by Muslim women is different from that of Hindus:

In terms of contraceptive use, Muslim women’s use of the pill is almost twice that of Hindu women and the use of IUD is also higher compared to Hindu women. Two things can be concluded: First that among the rural poor, the difference in fertility between Hindus and Muslims is not as marked as is usually supposed.

Second: there is a perceptible difference in the preferred method of contraception: Muslim women seem to be more open to the use of it.

Read the full article for more details.