This is a transcript of a talk I presented at Readercon 22 on Friday, July 15. Images are included as they appeared in the slide show. I had submitted the following description for the conference program guide:
Cooking has always been based on science, but the connection was made explicit with the 1984 publication of Harold McGee’s revolutionaryÂ On Food and Cooking. Chefs like Ferran Adria and Heston Blumenthal consider their research laboratories to be just as important as their kitchens in the development of new dining experiences, and have embraced the use of hydrocolloids, liquid nitrogen, and other agents to create foods that can only be described as science-fictional. With the recent publication ofÂ Modernist CuisineÂ and the ready availability of immersion circulators, gels, and “meat glue,” an ambitious home cook can experiment with methods that would have been out of reach even five years ago. How far can science take us in the kitchen? We’ve clearly moved beyond “astronaut food,” but are some of the more outlandish predictions SF has made about food within reach? We’ll look at examples–both old and new–of the extremes to which cooking can be pushed.
In 1979, while still a student at MIT working toward a biology degree, I took a summer job working at the General Foods Technical Research Center in Tarrytown, New York. Although it owned the Birdseye vegetables brand – easily the healthiest food it sold – most of GFâ€™s products were either convenience foods or coffee.
The research groups were organized like a nutrition label: fats & oils, carbohydrates, flavors, and proteins, which is where I was assigned to work.Â The three big proteins products were Jell-O, Cool Whip, and Shake â€˜nâ€™ Bake. Jell-O brand gelatin dessert (sorry, that’s now a conditioned reflex with me) is more than a century old. This photo is an ad from 1908.
I was involved with many projects in the Proteins group:
- We had to alter the formulation for Cool Whip to accommodate a new source of sodium caseinate, the milk protein that was the toppingâ€™s main component.
- I worked on improving the viscosity of gelatin when it was subjected to prolonged high heat, using formaldehyde as the strengthening agent. Formaldehyde was an approved food additive, present in denture adhesives, and apparently still not classified as a carcinogen due to heavy lobbying from chemical companies.
- I tested â€œbacon analogue,â€ a soy-based bacon substitute that cooked like bacon. The red and white â€œphases,” which corresponded to the fat and lean layers, had different rates of contraction, which would cause the edges to curl when heated. Just like real bacon! It looked like bacon, cooked lie bacon, and tasted like hot cardboard. Artificial bacon bits are still made from soy.
- The most interesting project was the â€œshrimp analogueâ€ being developed by a visiting Japanese research fellow. Egg albumin (whites) was mixed with shrimp flavoring and freeze-dried. The resulting powder was extruded through a cheese puff machine, soaked in water, and then pressed into shrimp-shaped molds. It was an exact recreation of the texture, but theÂ taste still needed work.
Two things happened to me when I returned to MIT after my summer at GF: The first was I discovered the world of speculative fiction that existed outside of my limited library of Niven, Clarke, Asimov, Herbert, and Tolkein.Â I donâ€™t need to mention how influential Harlan Ellisonâ€™s Dangerous Visions was; itâ€™s one of the reasons why thereâ€™s a Readercon. Harlan Ellison was the Guest of Honor (GoH) at Readercon 11.
I read through â€œThe Best SF Stories,â€ an Nebula Awards anthology compiled by SFWA, a one-stop somewhat representative history of the field. RogerÂ Zelaznyâ€™s Chronicles of Amber, at least the Corwin cycle, struck me as the first fantasy books in which characters actually ate regular food instead of nuts & berries, game, or magical bread. Corwin wouldnâ€™t start a hellride without eating a well-described breakfast.
In 1980 Bantam published Fundamental Disch, a collection of Thomas Dischâ€™s short stories (with a forward by Readercon 2 GoH Chip Delany). The first story in the collection is â€œDescending,â€ published in 1964. It was the first time I had seen a character defined both by the contents – or lack thereof – of his kitchen…
Catsup, mustard, pickle relish, mayonnaise, two kinds of salad dressing, bacon grease, and a lemon. Oh yes, two trays of ice cubes. In the cupboard it wasnâ€™t much better: jars and boxes of spice, flour, sugar, saltâ€”and a box of raisins!
An empty box of raisins.
â€¦ and also by his shopping list.
â€¦ A jar of instant and a 2-pound can of drip-ground coffee, a large tin of corned beef, packaged soups and boxes of pancake mix and condensed milk. Jam, peanut butter, and honey. Six cans of tuna fish. Then he indulged himself in perishables: English cookies, and Edam cheese, a small frozen pheasantâ€”even fruitcake. He never ate so well as when he was broke. He couldnâ€™t afford to.
Disch originally wanted to use gigot dâ€™agneau – leg of lamb – instead of pheasant, but his editor (Damon Knight?) thought it was too recherchÃ©.
In his 1989 novel Look Into the Sun, James Patrick Kelly (Readercon 19 GoH – yes, I am contractually obligated to point them out), describes a food substance called Vitabulk: cheap to produce, nutritionally complete, and with an almost indefinite shelf life. People carried kits with them (which I always imagined as similar to the â€œflavor packetsâ€ that come with instant ramen) so they could vary the flavor of the stuff, which tasted â€œlike insulation.â€ At a dinner party for his wealthy friends, the protagonist serves wieners and potato salad, a meal that was greeted with this reaction:
The guests were in various stages of gustatory ecstasy. The fare was not at all unusual for the wealthy; they ate at least one natural meal a day and meat or fish once a week. For others, forty-five grams of USDA guaranteed pure beef frankfurter was an extravagance: Christmas dinner, birthday treat.
This idea of mass-produced â€œpeople chowâ€ appeared again in Norman Spinradâ€™s Little Heroes, in which anyone could get a free serving of â€œkibbleâ€ dispensed from a vending machine. It was nutritionally complete, had a long shelf life, and required no preparation.
A friend who’s an entertainer was constantly forgetting to eat. He’d get cranky and only then would he think to eat something. When he asked us how to solve the problem, we all suggested that he keep stashes of food bars (Power Bars, Cliff Bars, etc.) around so he always had access to food. Being somewhat of an extremist, he decided that a few bags of monkey chow would work just as well. I spent a few days talking him out of it, citing the damage to his teeth and the havoc all of the fiber would wreak on his gut as the two primary reasons why a primate diet was contraindicated.
As it turns out, a few years ago someone decided to try living off nothing but monkey chow and water for a week, a feat he chronicled in a blog called The Monkey Chow Diaries. I can only guess that ZuPreem brand is better than Purina.
If you read enough SF stories set in the future, you reach the inescapable conclusion that people just donâ€™t care about food, as long as they have something to eat. Preparing food was drudgery, something to be avoided or made as painless as possible. These notions, popularized in the post-war â€™50s, gave rise to the convenience food industry I found myself working for.
As recently as last week – in Rule 34Â by Charles Stross (Readercon 21 GoH) I read a prediction about â€œcultured meat extruders,â€ machines which sat in the kitchen and made â€œchickenâ€ or other â€œmeatsâ€ from tissue cultures, much in the same way we have bread machines. (Although they donâ€™t so much manufacture the bread from raw feedstock as process ingredients.)
In The Wonderful Future That Never Was, a collection of predictions collected from the Popular Mechanics archives, the future of food was laid out over the course of a few decades:
- Dairy derived from kerosene (1928)
- All frozen food (1937)
- Made from cereal grass (1940)
- â€œIn A.D. 2000, cooking as an art is only a memory in the minds of old people.â€ (1950)
- Served in concentrated or pill form (1928)
Weâ€™ve been spared kerosene-derived dairy, but we can walk through supermarkets where we could buy entire weeksâ€™ worth of meals from the freezer cases. Not much of our food supply contains grass, but soy and corn have become ubiquitous ingredients of almost any processed food.Â The prediction for 2000 isnâ€™t quite as dire as presented here, but itâ€™s not that far off the mark, either.Â And weâ€™ll get back to the food pills.
For all of its advances and improvements, â€œfood scienceâ€ was still considered a tool of agribusiness and convenience food conglomerates.
The second thing that happened when I returned to MIT after that summer at GF was I decided I would teach myself how to cook. I bought utensils and a few cookbooks (The Joy of Cooking, The French Chef Cookbook), and set a few rules:
- Something new every week
- You kill it, you still eat it
- No cop-out meals
The â€œsomething newâ€ rule would eventually result in my developing some good kitchen skills as I added techniques to my repertoire. Julia Child was an excellent teacher in absentia. It helped that I had some lab experience; friends have said that I donâ€™t so much carve a chicken as dissect it.
For serious cooks, everything changed in 1984, when Harold McGee published On Food and Cooking: The Science and Lore of the Kitchen, still the most thorough examination of the science behind what happens in the kitchen. Not only did he explain why recipes worked or failed, but he also debunked a lot of common kitchen myths.Â Read McGee and you learn will how recipes work, as well as how to fix them when they fail.
In 1988 two physical chemists, HervÃ© This and Nicholas Kurti coined the term â€œmolecular gastronomy,â€ which they applied to the science of mechanisms that occurred during culinary transformations: heating, cooling, mixing, etc.Â The term has stuck, even though most chefs donâ€™t like it. It was coined to allow the chemists to get access to a series of scientific conferences.Â This also come up with the technique for cooking a perfect poached egg: Immerse it in 65 Â°C water for 45 minutes – the white sets, but the yolk remains liquid.
While a culinary revolution was starting in Europe, I left academic medical research and moved to the private biotech sector. I did a lot of work with tissue culture – growing specific cells as test substrates for drugs under development.Â I was also responsible for tending small-scale fermentation reactors – growth chambers where specially engineered cell lines expressed proteins that we purified to use as drugs. I had to develop scrupulous sterile handling techniques so as not to contaminate the cell lines.
I worked with a lot of specialized equipment for regulating temperatures and growth rates: water baths accurate to a tenth of a degree, balances accurate to a hundredth of a gram – all second nature to the work at hand.
Shortly after This and Kurti started their work, Ferran AdriÃ¡, a young chef at el Bulli – a tiny restaurant on the Costa Brava in Spain – started creating amazing food based on new techniques.Â Instead of sauces, he used foams, concentrated flavor essences into which air had been whipped.Â He also encapsulated liquids into edible spheres that would burst open in your mouth.
You can see a combination of the techniques in this dish, pea jelly, banana, and lime ice cream. Thereâ€™s ice cream, various gels, unusual textures.
This is a chocolate air, with crisp raspberry (frozen solid in liquid nitrogen and then shattered) and eucalyptus water ice.
This is instant chocolate cake, created by Ferranâ€™s dessert chef, his brother Albert. Cake batter is added to a whipped cream dispenser, charged with nitrous oxide, foamed into a paper cup, and cooked in a microwave for thirty seconds.Â To be more specific, this is my instant chocolate cake, made with the same technique.
These are olive spheres. Olives are purÃ©ed and strained, then mixed with a solution of calcium chloride and calcium lactate. Using a specially designed spoon to create the shape, the liquid is dunked into a solution of sodium alginate, a gel. When the alginate reacts with the lactate, it forms a skin around the liquid, creating a sphere that bursts open when eaten. The technique is called reverse spherification.
Iâ€™m passing around a sample of regular spherification, in which a liquid is mixed with the alginate and calcium lactate, then dropped into a calcium chloride bath. The drops solidify and retain their spherical shape. Youâ€™re eating apple juice â€œcaviar,â€ but this technique can be used with any flavored liquid, either sweet or savory.Â Spherification was actually invented in 1945 as a way to produce artificial chocolate-covered cherries.
AdriÃ¡ made it very clear why he cooked the way he did:
Provide unexpected contrasts of flavor, temperature, and texture. Nothing is what it seems. The idea is to provoke, surprise, and delight the diner.
To further that end, he closed his restaurant for half of every year in order to develop the next yearâ€™s menu, which would repeat no dish from any previous year.Â elBulli became the best restaurant in the world, a title it will hold until it closes for good at the end of this summer.
One of Adriaâ€™s contemporaries is Heston Blumenthal, chef at The Fat Duck, in Bray (outside of London), England.Â Blumenthal is self-trained, but heâ€™s a research fanatic with connections to people in the food science industry. He also has a lab kitchen across the street from his restaurant that runs year round.
The dish in this photo is a red cabbage gazpacho with pommerey grain mustard ice cream, a play on classic peasant borscht. This was an early appearance of a savory ice cream, which has become much more common now.
When this dish is presented, diners are told they have orange and beet jellies, and they should start with the orange. The orange-colored jelly is flavored with golden beets; the purple-colored jelly is flavored with blood oranges. Blumenthal plays with how your visual perception effects your taste.
This isnâ€™t a photo from the Fat Duck, itâ€™s from my dining room table. I served this dish as a pre-dessert course at She Who Must Be Obeyedâ€™s birthday dinner in May. Our guests were quite baffled by the disconnect between eye and tongue.
This is hot and iced tea. Drink from one side of the glass and the tea is hot, drink from the other side and the tea is cold. The drink is served in a double-walled glass to prevent your hand from picking up any temperature cues, the contents are fluid gels at two different temperatures, made with tea and gellan. A divider is placed in the glass, the gels are poured in simultaneously, and then the divider is removed. The slightly different densities of the gels keep them separated for a minute or so.
In his television special â€œHestonâ€™s Victorian Feast,â€ Blumenthal elaborated on this technique to create the fictional â€œDrink Meâ€ potion consumed by Alice, which tasted Â – in order – like toffee, hot buttered toast, custard, cherry tart, and turkey. He created infusions of each flavor, then incorporated progressively lager amounts of gellan to create a five-layered drink that delivered the flavors in order, making Carrollâ€™s fantasy a reality.
In a classic case of parallel evolution, the Clearly Canadian beverage company launched a drink called Orbitz in 1997. This is an unopened bottle that was handed to me as a free sample fourteen years ago, it has sat on a shelf in my kitchen undisturbed for all of that time. Notice how the little spheres (banana flavored) remain in suspension; theyâ€™re made from xanthan gum and are suspended in the same gellan solution that Blumentahl uses. The gellanâ€™s density is adjusted with sugar, which makes the solution clear and produces a neutral buoyancy state with the spheres.
Despite the lava light packaging, the drink never achieved escape velocity. Nobody wanted to drink something that was pineapple-banana-cherry-coconut flavored, at least not when it wasnâ€™t being served at The Fat Duck.
Even the toy industry got in on the act. Doctor Dreadfulâ€™s Food Lab was a kit that let you make creepy candies out of various gel-based powders. The bugs you can see molded here were gelatin,
… but you could also make â€œblobs,â€ â€œeyes,â€ and â€œwormsâ€ with flavored spherification components. This toy will be back on the market this fall, probably because of the rising popularity ofâ€¦
â€¦Popinâ€™ Cookinâ€™, a Japanese food toy. Using the same principles as Dr. Dreadful, but without the monster-themed equipment, you can produceÂ realistic looking â€œsushi.â€ The â€œuniâ€ (roe) are spheres made from the standard method, the rest of the bits are all gelatins with varying textures.
All of the things Iâ€™ve shown so far are manipulations of hydrocolloids, which are nothing more than solids dispersed in liquids. All of these solids youâ€™ve seen listed as ingredients in various foods are nothing more than texture and thickening agents with different properties.
Much to my surprise, it all came back to Jell-O.Â Iâ€™m also pretty sure we could use agar instead of formaldehyde to improve the viscosity of heated gels.
Another chef who utilizes modern techniques is Wylie Dufrense at his WD-50 restaurant in New York. This is aÂ deconstructed eggs benedict, with dehydrated Canadian bacon, almost solidified egg yolk cooked in a water bath, and hollandaise sauce that has been thickened with gellan, frozen into cubes, coated with English muffin crumbs, then deep fried. They become crispy on the outside and gooey on the inside.
Dufrense also pioneered the use of transglutaminase, an enzyme that bonds proteins together. He started calling it â€œmeat glueâ€ because of how it behaved.Â I was talking to Raj Patel, an author who writes about the world food supply, and he told me â€œI remember when â€˜meat glueâ€™ was another term for â€˜riceâ€™.â€
Activa is the brand name given to the enzyme by its manufacturer, Ajinomoto. The same company is the worldâ€™s largest supplier of MSG, and their scientists are credited with discovering â€œumami,â€ the fifth taste.
One of Dufreneâ€™s most popular dishes is called â€œbrickâ€ chicken. A chicken is completely deboned, leaving the skin intact. The meat side is dusted with transglutaminase, formed into a brick shape with the skin on the outside, and then sautÃ©ed. The finished dish has all crispy skin on the outside, and moist chicken on the inside.
These are photos from my preparation, I knew my chicken dissection skills would come in handy some day.
This is bacon-wrapped skirt steak, a recipe developed by Alex Tabot and Aki Kamozawa at Ideas in Food. It’s made by gluing two thin skirt steaks together to make a thicker steak, then bonding a solid layer of bacon strips to the outside. The whole thing is rolled in plastic wrap, then cooked in a constant-temperature water bath for about an hour. It gets a quick turn in a hot pan to crisp up the bacon before serving.
Also notice the barely-poached egg on the asparagus.
Donâ€™t confuse that dish with a bacon explosion, which is a bacon pot holder wrapped around a core of bacon-filled sausage. This bad boy has to sit in a barbecue smoker for four or five hours before itâ€™s ready.Â This photo is of the explosion I brought to last yearâ€™s Readercon, to prove to Scott Edelman that the dish wasnâ€™t a myth.
I would be remiss if I didnâ€™t mention one last chef, Grant Achatz of Alinea restaurant in Chicago. He was a sous chef at The French Laundry – the best restaurant in America – when he took a two-week working vacation at elBulli under Ferran AdriÃ¡. He would eventually openÂ Alinea, and is now considered the leading modernist chef in the US.
This dish is one of his classics, hot potato cold potato. The bowl contains vichyssoise – cold potato soup. Suspended on the pin is a fried potato topped with a slice of truffle, a chive sprig, and cubes of butter and parmesan cheese. Pull out the pin, let the garnishes fall into the cold soup, and slurp it down.
This is one of his dessert courses, dry caramel, salt.
The sample being passed around is the dry caramel alone. Itâ€™s made with the standard ingredients: sugar, butter and cream (with glucose syrup to create a smother texture), but before it completely cools, itâ€™s mixed with tapioca maltodextrin, a drying agent. It redistributes the moisture evenly across the mixture, creating this dry powder that reconstitutes in your mouth.
A few years ago Nathan Myrvhold, former director of Microsoft Research, asked on the eGullet forums (a known foodie hangout) if anyone could direct him to a source of cooking times and temperatures for sous vide cooking. Receiving no answer, he began researching the subject himself.
Sous vide cooking involves two techniques: The first is vacuum sealing food in plastic bags with various seasonings, aromatics, or other liquids. The second is cooking those sealed bags in a constant-temperature water bath for a set time until the food is done.Â The term â€œsous videâ€ is French for â€œunder vacuum,â€ which describes how the bags are sealed.
Much to my surprise, the constant temperatures needed for this kind of cooking are best produced by the same immersion circulators I used to work with tin the lab, and are made by the same company: Polysciences.
Donâ€™t confuse this technique with the boil-in-bags from the freezer section. Those contain fully-cooked food that you defrost in boiling water, sous vide is cooking raw ingredients to get them to the desired temperature.
You can find the Polysciences circulator in most restaurants, but the rest of us who canâ€™t afford the $1,000 price have come up with a number of what chef David Chang calls â€œghettoâ€ solutions.Â The simplest is just a pot on a stove with a well-regulated burner. It can be done, but it requires constant attention.Â The second is modifying a large rice cooker. Add a thermocouple (temperature probe) to the interior, have it send data to a power controller, and it will cycle the cookerâ€™s heating element to maintain constant temperature. You can also add a circulating pump to prevent hot spots from forming.Â The third solution is to build your own immersion circulator from commonly available parts. This is my circulator, built from teacup heaters, an aquarium pump, a thermocouple, and a temperature controller. Thatâ€™s a leg of lamb thatâ€™s been cooking for 24 hours at 57 Â°C (135 Â°F). When it came out of the bag I crisped the outside over a hot charcoal fire. I wound up with perfect medium rare lamb all the way through the meat, with a crispy crust.
Itâ€™s a good thing I didnâ€™t know about sous vide while I was still working in the lab; I would have been fired for contaminating the equipment with my after-hours cooking experiments.
In April of this year, Myrvhold finally published the results of his simple question to the forum: a six-volume book (actually five books and a waterproof kitchen manual with the recipes) called Modernist Cuisine (MC), which is the term that is slowly replacing â€œmolecular gastronomyâ€ as a description of the food weâ€™ve seen so far.
One of Myrvholdâ€™s collaborators is Chris Young, who worked at The Fat Duckâ€™s food lab. Being a Microsoft millionaire (and looking to rehabilitate his patent-troll bad guy persona), Myrvhold spared no expense in equipping the best food research lab money could buy.Â Thereâ€™s a lot of science in the book, but one volume is nothing but recipes. This is the hamburger recipe, which takes at least two solid days to make. The buns are baked from scratch, the sauce is a custom-blended mayonnaise, the lettuce is smoke-infused, the tomato is compressed under vacuum, the cheese is reformulated to melt perfectly, the burger is ground so that all of the fibers are aligned in one direction, and the ketchup is made with mushrooms.
Blumenthal came up with a similar recipe for his â€œIn Search of Perfectionâ€ TV series. I will never make either hamburger. I have my limits.
In order to show you what happens during different cooking processes, the researchers at MC cut appliances in half and photographed them while being used to cook. Hereâ€™s a charcoal grill coking hamburgers. The photo shows the most important flavor component of a grilled burger: the pyrolized fat from the flame-ups.
Hereâ€™s pad thai being cooked in a wok. The rig would catch fire every few seconds as hot oil spilled out onto the open flame.
I have cooked one MC recipe so far, the Thai beef short rib with sweet and sour glaze. It took three days to make and involved cooking the beef sous vide, browning it with a blowtorch, and creating fried beef jerky and dehydrated garlic chip garnishes. This was the main course for She Who Must Be Obeyedâ€™s birthday dinner.
This is another of the recipes, Astronaut Ramen. It looks like a styrofoam cup with a lid, but when you pour boiling water into it, the lid and cupÂ – which are made from freeze-dried gelatin – dissolve and become the base for the soup and freeze dried ingredients inside.Â I really want to make this someday, but I donâ€™t have access to a freeze drier. If you know of a lab I can use after hours, talk to me later.
That ramen is a far cry from what we thought astronaut food would look like. Here are Dave Bowman and Frank Poole enjoying â€œdinnerâ€ on board the Discovery. Everything was a solid paste except the coffee.Â It still beats food pills.
And so, finally, let’s talk about food pills.
As a freshman, I was given a physics problem to solve, which was really an exercise in estimation and calculation:Â Â How many candy bars did my professor need to eat in order to escape the Earthâ€™s orbit with a single jump?
We know the escape energy for a 1 kilogram mass, and we know the mass of the professor (300 pounds is roughly 136 kilos). We assumed 100% conversion of energy from one form to another, ignoring the inevitable heat radiated off during the process.
- Escape energy for 1 kg = 3.1 x 107 J
- 136 kg = 4.22 x 109 J
When we had to solve this problem, we also had to convert kilocalories (the energy unit of a bog-standard Hershey bar) to Joules, but Iâ€™ve reworked the calculation with a Mars bar (which, in deep-fried form, is part of my Scottish heritageâ€™s national cuisine) which helpfully lists its energy content in Joules.
- Mars bar = 9.67 x 105 J
How many bars do you think it would take?
- 4.22 x 109 J / 9.67 x 105 J = 4,360 Mars bars
Itâ€™s almost four and a half thousand, which would take quite a while to consume.
Food pills violate the laws of physics. The kinds of nutrients we consume arenâ€™t very energy dense. These calculations also ignore our need for vitamins, minerals, and other trace nutrients.
- 2,000 calories/day
- Carbohydrates & proteins = 4 calories/gram
- Fats = 9 calories/gram
- 2,000 calories fat = 450 capsules = 1/2 pound pills/day
Eat a bacon cheeseburger instead, itâ€™s one giant pill on a bun.
Which leads to an important point: people like to eat – and some people like to cook – real food. For cooks itâ€™s an act of creation, and sometimes art; for eaters, itâ€™s all about the enjoyment. Jim Kelly had it right: people get excited about real food.
â€œThe future has already arrived.Â Itâ€™s just not evenly distributed yet.â€
â€” William Gibson
William Gibson (Readercon 8 GoH) also had it right: The future of food is no different than the future of any other technology. As we have seen, it has already arrived – via Adria, Blumenthal, Achatz, and others – but itâ€™s still waiting for its even distribution.