Science Friday - Data-Collecting Smart TVs, Microbiome Cooking, Cannabis Pollution. Oct 4, 2019, Part 1

Episode Date: October 4, 2019

Today, it’s much easier to find smart TVs on the market. Companies like Vizio and Samsung create devices capable of internet connection and with built-in apps that let you quickly access your favorite streaming services. But that convenience comes with a hidden cost—one you pay for with your data.  Smart TVs have joined the list of internet connected devices looking to harvest your data. They can track what shows you watch, then use that data to deliver targeted ads, just like Facebook. Not worried about what media companies know about your binge watching habits? New research suggests that’s not everything smart TVs are doing. If you are the owner of just one of many “internet of things” devices in your home, those devices could be talking to each other, influencing what gets advertised to you on your phone, tablet, and TV screen. Dave Choffnes, associate professor of computer science at Northeastern University, and Nick Feamster, director of the Center for Data and Computing at the University of Chicago, join Ira to share what they each found when they looked into the spying habits of your smart devices. Cooking food changes it in fundamental ways. Cooked starches are easier to digest. Seared meats are less likely to give us foodborne pathogens. And overall, we get more energy out of cooked foods than raw. But scientists are still pursuing a pivotal question about cooking: How did its invention change our bodies and shape our evolution? Did it shrink our teeth and digestive tracts? Or did it increase our brain size? Researchers writing in Nature Microbiology reported a new chapter in our understanding of how cooking has changed us: The microbial communities in our guts change dramatically if our food is cooked or raw. And mice whose microbiomes were associated with raw foods seem to gain weight more easily—but their microbiomes also showed signs of damage from plant-generated antimicrobial chemicals. Harvard researcher Rachel Carmody explains the findings, and what our microbiomes might say about cooking food and evolution. Between water and electricity, Colorado’s legal cannabis industry already has a big environmental footprint. But what about Front Range air quality? Could the plant itself be contributing to air pollution? No, it’s not the pot smoke. Colorado Department of Public Health and Environment is conducting a study of terpenes, the organic compounds that make the cannabis plant smell so strong. Terpenes are classified as volatile organic compounds. Many consumer products release VOCs, like acetone in nail polish remover and butanal from barbecues and stoves. VOCs from terpenes are harmless until they combine with combustion gases to create ozone. That’s why the state is studying marijuana emissions—it’s about where it’s grown. Unlike other VOC-emitting crops, like lavender, cannabis is often cultivated in greenhouses in the industrial areas of cities, near highways and lots of cars. “Here in Colorado, as far as air quality concerns go, ozone is our largest pollutant of concern. We are not meeting the national ambient air quality standards for ozone,” said CDPHE’s lead researcher on this project, Kaitlin Urso. Denver’s ozone problem is especially bad. According to the American Lung Association, it has the nation’s 12th worst air quality. Usually, it’s the Environmental Protection Agency that studies emissions from new industries. Since marijuana is still a federally controlled substance, it can’t. With the feds on the sideline, Urso said it’s now up to the state to figure out, essentially, “how many pounds of VOCs are emitted into our atmosphere per pound of marijuana grown?”   Subscribe to this podcast. Plus, to stay updated on all things science, sign up for Science Friday's newsletters.

This is Science Friday. I'm Ira Flato. Egypt is wrapping up construction on its first utility-sized solar power plant. The 1.8 gigawatt plant will feature up to 7.2 million photovoltaic solar panels when it's done. It comes with a $4.1 million price tag. It's predicted to produce enough energy to power up to a million homes, and it's poised to be one of the world's largest solar facilities. The solar park will be so big it will be visible from space. Here to talk about that as well as other short subjects and sciences.

Amy Nordrum, news editor at the I-Triple-E Spectrum. It's good to have, Amy. Thanks, Ira. This is crazy, this solar plan. It's a massive project. It's been underway for years now, and it's finally just about to come online. I think about 80% of the solar panels and projects there are now complete, and it should be online later this year or early next year, producing power for Egypt.

And why Egypt in particular? Well, Egypt years ago started investing in this project because they were a fossil fuel-based place, and they were running into a lot of blackouts because of fuel shortages. And they have this massive solar resource. I mean, they have more solar radiation than almost any other country in the world.

But they hadn't really invested in solar in the way that you might think. So they wanted to diversify their electricity production, and they started this project. Most electricity-generating and transmission assets in Egypt are state-owned. but they actually brought in a lot of private developers to help with this project, which probably will make it more successful in the end.

So are they going to do the whole thing with the batteries and the panels, or are they just doing the first step? Right now it's just the first step. This was an important step. One of the engineers I talked with said it's done three main things for Egypt. It's lowered the cost of PV panels in the country. It's trained a huge workforce of people that worked on this project that didn't previously know how to install these panels.

And it's also convinced people that solar power has a lot of potential there after they ran into some hiccups with some concentrated solar power plants before, which are a different type. But there's no battery storage plan for the project yet. That might come later. You know, I've heard stories over the years about all of North African countries, Morocco, Egypt, you know, being the power supply for Europe, like, you know, building cables that go through the Mediterranean. Yeah, I mean, it could be a good idea.

I will say there's some tricky parts about installing all these panels in the desert. They have to think about things like dust and sand blowing over the panels. and so they have this whole system in place for cleaning them with these large adapted tractors that have huge brushes and are going to drive by and brush the panels off. So you do have to figure out some things because of the location, but there's a lot of potential there. More jobs for everybody. Your next story is about a study that challenges most of what we've been told about, red meat consumption. It broke earlier in the weekend.

It's had a lot of reaction, right? Absolutely. A research group known as Nutra Rex published actually five studies on Monday that really upended a lot of nutrition advice concerning red meat that consumers have heard in recent years. And they looked at past evidence. So this was like a systematic review of all the evidence studied to date on the effects of red meat consumption on things like cardiovascular disease and cancer and total mortality. And they applied a new criteria called grade. This is a system of evaluating the quality of evidence in previous studies. And through that system, they came to the conclusion and the

recommendation that really most people don't need to decrease their red meat consumption. and can, you know, eat as much red meat as they like. And that's really upset some people who've been saying all this common knowledge for years. Yeah, there's many nutrition recommendations that suggest lowering your red meat consumption for health as well as other reasons. But this new method called grade is really supposed to look at the most rigorous evidence. And so it ended up throwing out a lot of observational trials, things like case studies that aren't considered to be as strong of evidence.

And the researchers who did the studies say that it's trying to. to push nutrition science more the direction of randomized controlled trials, which are the highest quality evidence but haven't been able, haven't been done as much in the field so far. But you know there's been other opposition by environmentalists about eating red meat, not just about the nutritional value of it. Yeah, that's a very important point. This study only looked at the health effects, and certainly if you were toast to eat less red meat for other reasons, such as environmental impacts of meat production, you know, this recommendation doesn't apply to those concerns. That's still

definitely a factor, a lot of people's decisions. Next up, a new study that reveals the unexpected strength of animal silk, animal silk. That's right, the silk produced by spiders and worms actually gets stronger and tougher as it is exposed to colder and colder temperatures. And this is really weird because most fibers become more brittle and actually break when they're at extremely cold temperatures. But researchers in the UK and China chilled silk samples down to as cold as negative 320 degrees Fahrenheit using liquid-neutral. nitrogen and tested it for strength and stretchability and found that the colder it got, the stronger silk became. That's crazy.

It's really unexpected. They think it has to do with the composition of silk. The spinning mechanism used to produce silk creates these really long chains of super thin nanofibrils, as they call them. So these are chains of molecules that make up a thread of silk. And as the temperatures drop, these threads harden, forming a kind of maze that makes it difficult for fractures or breaks to get from one side to the other, and that ends up making silk stronger in the end.

So, but where are you going to find a use for something that's strong at 300 degrees while freezing? Let me tell you, they have some ideas. I mean, this is a group of biologists. It'll be up to others to kind of act on these. But some of the applications there are suggesting is like, if you need something flexible and strong in a really cold environment, like Antarctica or even outer space, so say you need to build, like, a net or a sail that could withstand some of those really cold temperatures

and remain strong and flexible, you know, silk might be your go-to material. I'm thinking space elevator. There you go. That's my idea. You know, how many years they were looking for something. So it was silk, it was there all along. Let's move on to a different kind of thread. One of the more galactic nature, something called galactic filaments.

Tell us about that. Yeah, these are really fascinating to learn about this week. I hadn't heard of them before, but they're the largest structures in the universe. These are filaments or threads of hydrogen gas, you can think of them as, that stretch from galaxy to galaxy, and in fact from clusters of galaxies to clusters of galaxies. They make up a thing called the cosmic web that astronomers have theorized and physicists have theorized existed since the 1990s. But this week, for the first time, researchers at the Reichen Institute in Japan published a study in science in which they confirm the existence of these filaments in this web. They made the first direct observations of these giant threads of gas and published it for the first time this week. Well, that's good.

Always waiting to find out more about what's out there. Yeah, absolutely. All right, Amy, thank you very much. Jamie Nogerman, German News editor at the I-Triple-E Spectrum. Now it's time to check in on the state of science. This is KERNO. St. Louis Public Radio News. Local science stories of national significance.

marijuana is a big business, especially in Colorado, since cannabis was legalized years ago, seven years ago there. The state has dealt with a lot of unexpected pot-related problems. There's the environmental toll that comes from the large amounts of electricity and water needed to cultivate the plants, and now a new concern about how cannabis cultivation might impact air quality. My next guest is here to clear the air and tell us more about cannabis and air. air pollution. Michael Sazick is a climate and environment reporter at Colorado Public Radio. Welcome, Michael. Hi, Aaron. Thanks for having me on. You're welcome. Okay. In your story, you talked to some

researchers from Colorado Department of Public Health and Environment who are concerned about cannabis and air pollution. What was their main concern? Yeah, so they're actually looking at how the plant itself might be contributing to ozone. So this isn't about smoking the plant. This isn't about that water and that electricity that these grow houses uses, this is actually about terpenes, which is an organic compound that's classified as a volatile organic compound, and how those gases might actually mix with combustion emissions, which come from cars, to mix and create ozone. Oh, so this is that kind of air pollution problem. Interesting.

Yes, and it's a big issue in Denver, yes. And cannabis isn't the only plant that releases turpines, right? Yeah, so any, think of plants that you can, you know, might have strong smells. So like lavender or pine trees, they can produce terpenes. And these terpenes are actually harmless all by themselves. But what the concern is is that when these turpines, which, again, are classified as volatile organic compounds, when those VOCs mixed with combustion emissions, it can create ozone. And in the presence of sunlight, that's the kind of the, kind of the.

perfect equation for ozone. So are they concerned about it in big cities like Denver or places like that or generally all around the state? So it's Denver is definitely the biggest issue in Colorado, but ozone can can spread. But it really is where the, so in Colorado and in Denver, that's where a lot of our marijuana and cannabis grow houses are. There are over 300 within the city itself. So, and they're usually industrial areas where they are in.

warehouses. So they're by highways and lots of cars. And so that's really where the concern is, is that mixture of these types of gases all right here in the city of Denver. And it can be big issues for cities across the country that are legalizing cannabis and might be experiencing that same issue. Now, I know that you also talked to a researcher at the University of North Carolina in Chapel Hill, who has done some work around air pollution and cannabis. What has he found so far? Yeah, so some of his research has come out. Some is still waiting to come out. He shared some of his findings with us early. And he found that VOC's tons of emissions per year could be 66 to 657 tons of VOC emissions per year coming from these grows in Colorado. And he acknowledges that that is a very broad number and that the research through the Colorado Department of Public Health and Environment and other research will help. help narrow that number down so we can really decide exactly how much tons of emissions are coming from these marijuana grows. One of the interesting sidebars to this story is that he had to grow it in his garage, right?

Because cannabis growing is not legal federally, right? Yeah, so he came to Colorado to do his research because it's not legal in North Carolina. And he thought that once he got here, that his research would be in the clear, but he actually teamed up with the National Center for Atmospheric Research, and their labs are federally funded. So when he came to Colorado and he planned on growing those plants in their labs, he found out he couldn't do that. So he had to improvise and he grew them in a garage in Boulder.

And he grew a couple different strains and tested the air quality kind of throughout their life cycle. And he found that different types of VOCs in different amounts would, you know, be released depending on the type of strain of marijuana cannabis and at which phase of its life it was in. Terrific. Thank you, Michael Elizabeth Zakis, a climate environmental reporter for Colorado Public Radio. Thank you for joining us today. Yes. Thanks for having me on. I really appreciate it. This is Science Friday. I'm Ira Flato. How many of you have a smart TV at home? Raise your hand. No. No. You can put it down now. You know, the brands like Vizio and Samsung that come with an internet connection, they have those built-in apps at the bottom of the screen that let you quickly load your favorite streaming services. Well, they are the few. future, they are affordable and convenient, but they also come with a hidden cost when you pay

with your data. By that, I mean that the smart TVs and streaming devices have joined ranks with websites and cell phone apps in harvesting and sharing your information. They track what shows you watch, and then they share that data with third parties to deliver targeted ads, just like notorious privacy rights Bogerman Facebook. Yeah, so this week we asked our listeners, how concerned are you about your internet connected devices snooping on you? And a lot of you sent in your answers via the Science Friday box pop app. And here's a sample of what you had to say. I have smart speakers from both Google and Amazon all over the house, and I'm not the least bit worried. So I do have a smart speaker, and lately I've been kind of suspicious of it.

I usually thought that was kind of funny when people thought they were being listened to. But lately, without me speaking, I can see the light turning on without me saying her name first and then just will automatically turn off. But it seems like it's listening somehow. Thanks to Jeff from Kingston, Laura from Maryland for sharing their thoughts with us. And if you'd like to share your thoughts with us right now, give us a call on number 844724-8255. or you can tweet us at SciFRI. What do you think about, how concerned are you,

about your internet-connected device snooping on you? But maybe you're not worried about what media companies know about your binge-watching habits. New research suggests that not all smart TVs are tracking. Here with me to talk about two new studies looking at TV tracking are my guests. Let me introduce them to you. Dave Chuffness, Associate Professor of Computer Science

at Northeastern University, and Nick Feimster, Professor of Computer Science and Director for the Center for Data and Computing at the University of Chicago. Welcome to Science Friday. Thanks for having me. Thanks. Great to talk to you, Ira. Nice.

Thank you. You're welcome. David, most people are aware that websites and cell phone apps are collecting your data. What data is your smart TV collecting? Well, it really depends on the smart TV. And, of course, it goes beyond smart TVs. But ultimately, what we know is, as we increasingly have internet-connected devices in our homes,

they're communicating with other companies over the internet, often communicating with servers run by Amazon, Google, number of other companies, and communicating with known advertisers and analytics companies who, at least on mobile apps and on websites, are collecting data about individuals for tracking and profiling. Now, let's say I don't have a smart TV, but I have an Apple TV or a Roku or a stick,

you know, an Amazon stick connected. Are they also collecting? data? For the most part, anything that's connected to the internet can be collecting data. And we see even by just turning on devices, not even interacting with them at all, they're communicating with other destinations. They're collecting data. Nick, in a recent study, you looked at how the apps on your smart TVs are sending information to third parties. Give me an idea of what kind of data they're collecting. Is it what channel you're watching, what commercials you're watching, What kind of stuff do they want to know?

It's a variety of things, Ira, and I think one of the things that we found surprising was that some of the channels that we looked at, we found them sending quite detailed information, including the video title, so exactly the title of the show or the video that you were watching. Of the 2000 channels that we looked at on Roku and Amazon, many of them sent back unique identifiers, something, called an ad ID. And many of them also sent back other more detailed information, including the serial number of the device, as well as your city, state, and zip code in some cases. And so what are the third parties doing with this when they get that information, take? We don't know everything that they're doing with it, and I think that's probably something that, you know, certainly we should be more concerned about. But one thing that we definitely know

that they're doing with it is advertising to us, basically sending us targeted ads. many of the most prevalent trackers. Generally, the organizations that collect data about us, we refer to them as trackers. And generally those are advertising companies. Many of the most prominent trackers are actually Google and Facebook. David, your paper looked at what the TV itself was tracking when you turned it on. Tell us what you found. One of the big challenges we have in this field is for a lot of these devices in your home, we can't actually see what is the data that they're collecting.

For instance, we can't even see that there's an ad ID like Nick just mentioned. But we can see who they're communicating with. So, for example, what we found is on TVs like the LGTV or Samsung TV, if you open an app, like Netflix, for example, it'll communicate with Netflix, as you'd expect. But even if you don't sign in and you don't have an account. Now, if you turn off your TV and turn it back on, that app automatically starts again and will communicate with Netflix again. And so, you know, what we see is that the apps are essentially sending information about individuals. Those other destinations can learn your IP address from that. They can figure out where you're located.

And they know what kind of TV you have. And combine that with a bunch of other information that they track from other sources like your mobile device. they can start building a profile of where you are, when you're home, and eventually even what you're watching. Let's talk about the other devices, because you also discovered that our smart TVs might be talking with other smart devices in our home, right? Yeah, so the thing that we're concerned about is increasingly,

it's not just the TVs and that they're watching what we're watching, but we have devices that have cameras on them, devices with microphones, motion sensors. So essentially the space that we traditionally considered private. The space between the walls of our home are now increasingly occupied by devices that are watching us from the inside. So we saw examples of cases, as one of your listeners pointed out, where the smart speakers are listening when they shouldn't be, when the wake words are not triggered. We see cases where video cameras are taking footage or images and sending them over the

internet, sometimes not even securely, and all of this is happening usually without users being aware. So they're on all the time, and even when we don't say the magic name that we evoke them. Right. Exactly. We actually did a small study, and we're expanding it, but we started by just trying to play a lot of dialogue, things that people say to each other at these devices. So we started with a Gilmore Girls episode because they say a lot of words per minute. So we'd have a lot of data to throw at it. And we saw some activations that were kind of, they made sense, and things sounded like Alexa.

but we also saw things that, you know, weren't terribly expected. Like, I need medical assistance, woke up the device that was listening for Alexa. And we think it's somewhere between medical and assistance that kind of sounds like Alexa. But it does kind of get you thinking what other kinds of things might they be listening for or accidentally wake up to. Let's go to the phones because lots of people are wanting to talk about this. Let's go to Houston with David. Hi, welcome to Science Friday. Go ahead, David.

Hi, are you there? Yeah, go ahead. Hi. Well, so I've been listening, and I like you guys, I listen all the time. I work at NASA, and for the smart TV thing, personally, I just don't even enable them to the network. I just leave them off the network and make them dumb TV. And part of that is what we did about three years ago when these things started coming out. We were concerned about the traffic that was coming out of these devices and, you know, who they were. talking to, so we set up a lab and instrumented all these devices and started kind of tracking what they were doing. And what we found was a lot of the traffic is encrypted, and so you don't really know what they're saying.

And we also found that they need to communicate back to their mother's ship in order to do the updates and take care of themselves. And so you had to have the connection open, but kind of the solution set was just to put them on a network that's not on the same network. that your data is. So that's kind of where we got you. You walled them off. Is that one answer, David, or Nick?

Yeah. I think a couple of things. He's right on the money that in many cases a lot of the traffic, and David mentioned this as well, a lot of the traffic that we'd like to know more about that's leaving our houses is encrypted, so it can be tough to learn about it. In our study, we managed to actually break the encryption

for two of these devices, the Roku and the Amazon Fire TV. So we were able to get a little bit more information about the types of things that the devices were sending back. But in general, that's a tricky problem. And it's true that one way you could approach this problem is say, you know, I'm just going to make everything dumb by not giving it an internet connection. That may be fine for some devices,

but for other devices, it just may not work if you don't give it any internet connectivity. And so now you're left with this question, And what connections do I allow which ones are actually necessary for the device to work versus which ones could you block and maybe avoid some of the risk of information exposure? There's an additional consideration there to Iro, which is that many households and many spaces are multi-user. So you might make one decision, but your wife or your kids might have different ideas about what's okay, what kind of data is all right and what should or shouldn't be connected to the network. So these starts of situations particularly get particularly complicated when we have situations with multiple users or people who don't have

the autonomy to make choices. Let's go to the phones. Let's go to Laura in Huntington, Long Island. Hi, Laura. Hi, here are you. Hi there. Go ahead. Okay, so what happened is there's a game, and I'm forgetting the name that kids play on Alexa where it's basically Alexa saying, I can guess anyone in the world that you're thinking about. And it's usually like, is this person the Disney princess or whatever? They narrow it down. And then Alexa will say, are you thinking about Elsa? And my daughter laughs and says, yes.

But they can also do it with like real people, right? So the kids always pick out babysitters or family members. And one day my daughter was playing and she thought of her dad. So through the questions, we got down to where Alexa said, says, does this male live in your house? And my daughter says yes. And then Alexis says, was this male in love with your mother? And she laughed and says, yes. And then it says, does this male inoculate all the children in his house?

And I look at her. My daughter knows vaccinate. She doesn't know inoculate. And I put my finger out like, don't answer that. We do inoculate, but I don't want, you know, that's ridiculous. So Alexa repeated the question. And again, said, does this man inoculate the children in his house? All the children in his house?

and I just said, Alexa, mind your own business. And Alexa says, you know, okay, I think you're thinking of your father. Is that right? My daughter left and says, yes. Wow. But it was pretty horrifying. Like, okay, one thing for marketing to collect information, like, does your dad need a new car? But, you know, obviously inoculation is such a hot topic right now, particularly in New York,

where there's a lot of students who didn't get to return to school in the fall. Okay, okay. Okay, Laura, we're running enough time. Let me get it's a great, great anecdote. What do you think of that? Related to Laura's observation and related to our study of third parties collecting data, we actually just completed a study of Alexa apps. They're called skills.

And one of the shocking things, Laura, is that not only was the Alexa trying to collect that information about you, but you may not even know who or what organization was doing that. because when you ask a question of Alexa, sometimes it'll install software from third parties to help get answers to those questions, often without asking you. So not only is somebody asking you about inoculation practices in your house,

but you probably don't even have any idea who that is. It's probably not Amazon. Hi, Mara Flater. This is Science Friday from WNYC Studios. Talking about your Internet of Things. In a few minutes we have. I want to talk to you, Nick, about a project out of Princeton called the Internet of Things Inspector. It's a piece of software that I actually installed last night to watch it.

It pings all your devices in your home. It keeps track of what they're sending. And I used it on my Roku, and on the website at Science Friday.com, you can see a 20-minute little piece of video I put up there for, It shows 20 minutes of usage, and how many things my Roku was talking to in just those 20 minutes was amazing, Nick. Yeah, you can see, Ira, this is not just academic paperwork, right? It's real problems in your home as well. And the credit for this tool, I should mention, goes to a postdoctoral researcher at Princeton, Danny Huang.

And we got together, and we basically realized that, you know, people are buying all kinds of things, and they just have no idea when they plug it into the internet, what's going on? Like, what is this device talking to? And, you know, it's one thing to write a paper, but, you know, we really want to provide users, consumers with better information so that when they buy something off the internet or off of the shelves at the store, they know what they're getting into and what they're letting into their home. And the idea was like, can we make this as close as possible to a one-click, give the user the best information they can in the easiest way possible.

And I'm excited that you were able to get the information, and I hope many of your listeners will try it out too. Well, it's great. It's called the Internet of Things Inspector, and you can just Google it and find it. Google it. And it really shocking. There's a graph of my Roco up there about how many different places it was going to.

David, is there any way to keep your smart home disconnected and still have your devices working? So this is actually a project that we're actively working on. So what we do is we interact in an automated way with devices and then just block one, then the next connection, and so on. Keep blocking connections and see if the device still works. So at the end of this study, we'll be able to produce something where you could, say, install a device in your home

that just blocks everything that's unnecessary for a device and only keep the connections you actually need. I started looking at the hub called Hubitat, which supposedly will keep things in your home connected to each other but does not go out to the internet. And I'm looking into that. It's pretty interesting. Are you familiar with that? Not precisely, but it really depends on what kind of device you're using.

For instance, some devices you cannot interact with them unless it has an internet connection. And this is actually one of the counterintuitive things for a lot of people is you're in your home with your smartphone trying to turn on your lights or whatever operation you're doing. And you think you make that operation and all that data, all that network traffic, all the commands stay in your home, when in many cases it actually leaves your home, goes off to some server in the cloud, and then comes back from the cloud to talk to your device. So that's another example, this counterintuitive behavior that you could see, for example, from IOT inspector, that, you know, invisibly there's a lot going on in your home that you may not expect. Yeah, Ira, another one that I would add is that the study on smart TVs had an interesting result where you could turn on, tracking protection on the Roku, and it basically barely did anything as far as blocking the trackers that the device communicated with.

There's some other interesting technology in this area. One of them is called Winston Privacy. If you've read 1984, you'll get the reference. These guys are basically trying to build a firewall that you could drop into your home, put it basically in between your home router and the access point, and do exactly what David is talking about. Well, Nick, we'll check back with you and David. Famster and David Chuffness, working on keeping you isolated from bad things on the internet.

Thank you both for taking time to be with us today. This is Science Friday. I'm Ira Flato. We've all heard the phrase, you are what you eat, and when it comes to the human gut microbiome, all those microorganisms, trillions of them that live in our digestive system and assist some of our most basic bodily functions, it seems more and more like that is really true. Study after study links the food we eat to the health and structure of our gut microbes. But what about the food we cook? Does it matter to a microbe if that sweet potato is raw or is it boiled?

The research published in Nature microbiology earlier this week found evidence that, in fact, it does. Mice-fed raw diets had vastly different microbiomes than mice-consuming. cooked food. And at the end of the day, that could lead to a whole host of new insights into human evolution and our own first forays into flame broiling our food. Here to explain why is Dr. Rachel Carmody, Assistant Professor of Human Evolutionary Biology at Harvard University and the lead author on this new research. Welcome to Science Friday. Thank you very much, Ira. It's nice to be here. Nice to have you. So how much does cooking actually change our food? Well, cooking profoundly changes our food, as does other forms of food processing.

But unlike non-thermal forms like grounding or pounding, cooking can both physically transform the food and it can chemically transform the food. And it does so in different ways for different kinds of foods. So, for example, for carbohydrates like starch, cooking can actually take starch, which in its raw native state exists in a really tightly bound, and it can kind of loosen those bonds and allow them to swell with water. And this makes it much easier for our enzymes, our amylases, to get in and break that starch down so that we can take advantage of the carbohydrate.

For proteins, it's doing something very similar, which proteins exist naturally and kind of tightly wound balls of yarn. You can kind of think of it that way. And cooking unwinds that yarn and allows our protein digestive enzymes, proteases, to come in and cleave off amino acids that our bodies can make use of. So it essentially allows us to digest a greater fraction of the food that we eat in the small intestine. And then what we can't digest in the small intestine passes into the colon where it becomes

the work of our gut microbial communities. The other thing that cooking essentially does, as those other forms of food processing, is you can think of it as externalizing part of the digestive process. And by starting the process of breaking down foods, our bodies basically have to do less work to complete the process. And so there's a metabolic cost of digestion that goes down when we've cooked our food compared to when we eat it raw. And these are things that we knew about the cooking process even before we conducted

this study. But interestingly, what we had found is, you know, if you take humans or you take mice and you feed them raw versus cooked food, you do see that. the cooked food gives more energy overall because it's easier to digest and because it's less costly to digest. But when you add up the effects on digestibility and the metabolic cost of digestion, they didn't actually make up 100% of the difference we were seeing just in terms of how much extra energy were these humans or were these mice getting. And so we really started to think there was another process happening. And around the time I was doing this work and realized, oh my gosh,

there's another mechanism that I can't explain. Work was coming out showing diet effects the gut microbial community, different microbial communities, feedback into energy metabolism in the sense that you could have two different communities that return different amounts of energy back to the host. And so that's really where we started with this project was to try to understand is this part of the picture. And so you actually found that the microbiome responds differently to cook versus raw food. Yeah, that's right.

And we tested two different kinds of foods. We tested meat and we tested tubers. So you can think of these as potatoes, sweet potatoes, underground storage organs of plants. And the reason we chose these two foods is I'm a human evolutionary biologist. These are representative foods of types that we thought would have sustained and provided the bulk of calories for ancestral humans for at least the last couple of million years. And so we were interested in kind of understanding how cooking would have improved.

of these particular foods and then what that relationship would have been with the gut microbial community. And so we fed, we started with an experiment where we fed mice, Ron cooked meat, and Ron cooked sweet potato. And we just surveyed to see what would happen to the microbial community in terms of its composition, meaning the different combinations of bugs that are present in that community, as well as its function. So what particular genes was that community transcribing?

What does that tell us about what it's seeing in its environment? as well as its behavior. And when we fed the raw versus cooked meat, we actually didn't see very many changes with the microbial community. And I can explain why in a bit. But when we fed the raw and cooked sweet potato, it was night and day. Within 24 hours, we saw completely different microbial communities,

and they were doing very different things. And the things they were doing differently were digesting starch and sugar, as well as producing a whole array of products for fighting off, compounds that we call xenobiotics. These are things that aren't nutrients, but things that can come in with the diet that the microbial community may treat as sort of foreign compounds. And so you can think of this as sort of a detoxification effect. And those effects were really seen within the first 24 hours, and we became very, very curious about why those particular effects. And if we think back to kind of what cooking is doing to food, we knew that in the case

of these plant foods that are rich in starch, that cooking would be changing the digestibility of starch. So one of our hypotheses was, okay, so you cook the sweet potato, more of it disappears in the small intestine because our bodies have an easier time breaking it down. Less of that starch makes it into the colon. And so this is fundamentally changing the nutrients that are entering that colonic environment where the majority of the gut microbiome resides. And like in any environment, if you change the nutrient flows in, it's going to be a lot of

going to change competition among species. And so we had a strong hypothesis to suggest that this is one of the mechanisms involved. But that additional kind of interesting tweak that we saw with gut microbial behavior in the sense that they were fighting off these xenobiotic or these foreign compounds gave us an idea about a potential second mechanism, which is that when plants grow in the ground, they actually produce a whole array of natural antimicrobial compounds. and they do this to protect themselves against predation. And if you eat these plant foods in a raw state,

some of those antimicrobial compounds could actually retain their activity and exert them within the gut and change the gut microbiome in that way, whereas if you eat them cooked, some of those compounds may have been broken down outside the body, so when you eat them, they don't really have the same kind of effect. So the raw foods are bringing in with them the ability to fight microbes in the gut, like they fight it in the soil. I think that's exactly right, although we didn't really realize this until we started doing this research, and we said, oh, we put two and two together and said, well, of course they would be doing this.

Why not? They're doing this in their native state. We're eating them in their native state. Some of these compounds may actually still be active. And so then we designed a series of experiments to kind of understand the role of digestibility and the role of these xenobiotic compounds in shaping the gut microbiome. And the big picture is that when we design diets, that were identical except for the digestibility of the starch fraction, we could actually recapitulate

in the gut, a gut microbial community that looked very similar to what we saw on the raw versus the cooked diets, purely by manipulating the digestibility of starch. And when we tested this across other kinds of plant foods, so not just sweet potato, we tested white potato and corn and pea and beet and carrot, what we found is that our ability to manipulate the gut microbial community with cooking was largely restricted to starch or retreat. foods that were improved to the greatest extent whose starch digestibility was improved to the greatest extent by cooking. Is that why you could not get a difference in the meat because it wasn't starch-related?

So the meat didn't show much of a difference? Yeah. So obviously meat does not have a lot of starch. We do know that meat, the digestibility of meat, is improved to some extent with cooking because, like I said, cooking kind of unwinds these proteins, these like balls of yarn. but meat is a high digestibility food to begin with. So the net effect of cooking on the improvement of digestibility is pretty small in meat. It's a pretty highly digestible food even when raw.

And we all know this, you know, if you just take a step back and look at cultures and what cultures are doing, there are many cultures out there that will readily eat meat raw or fish or other kind of plant, sorry, animal materials raw, because historically they've returned a decent amount of that energy, whereas with things that are very rich in starch, cultures worldwide tend to cook these items because we don't get much out of them unless we do. And so, you know, it makes sense that digestibility would be one of the drivers

for the reason we didn't see an effect in meat, and we saw this profound effect in the sweet potato. Is there any then recommendations for whether, you know, broadly recommendations, for whether we should cook our vegetables or starches or eat them raw? Well, the broad recommendation is that our bodies are not designed to eat this stuff raw. And we know this because there have been studies of long-term raw foodists who have been raw foods for at least three years, and they eat at least between 80% of their food raw, up to 100% of their food raw. And what we find is that over time these people have lower and lower body mass index,

their weight for height goes down and down. And you would think in a modern industrialized society, oh, that's great. This is a great weight loss tool. It is. But what you find is that among women who are consuming 100% raw food diet, they are so energy limited that they actually stop ovarian cycling. They have problems with reproduction.

And from an evolutionary perspective, what this suggests is that our bodies, through a long history of cooking our food have essentially become adapted and dependent on incorporating at least some fraction of cooked items into the diet. We don't seem to thrive very well on 100% raw food diet. That said, depending on people's goals, if they're trying to lose weight, it could be a good kind of short-term scenario.

But over the long-term, this is not the diet we were meant to eat. We were actually biologically committed to cooking some fraction of our diets at this point. Are you saying that that's an evolutionary change? That has happened? It is an evolutionary change that's happened. We've now seen it. So my colleague Richard Rangham has really been the person advancing this idea. But he kind of took it from first principles, and he said, okay, well, you know,

if we're human evolutionary biologists, let's think about the things that really make humans unique. in order to understand why humans are the way we are today. And, of course, one of the things that does make humans unique is that we're the only species to control fire. We're the only species, save for a few domesticated animals where we cook the food for them. We're really the only species that is consuming cooked food on a regular basis. And so he started there and said, well, I wonder if cooking has actually transformed our biology.

And we know that at least for the last two million years, humans have looked really different in the way we digest our food compared to our closest living relatives, chimpanzees, bonobos, and gorillas. We see changes in the kind of way that we chew our food. We see smaller chewing muscles, smaller molars. We see smaller teeth and a dental arcade, like our mouths get smaller in general. We also have a smaller gut. So you can kind of imagine this.

If you look at a gorilla, it's got this enormous stomach. its rib cage kind of flares out because it's containing a ton of intestines. Humans, on the other hand, have a waist. We've got small intestines that kind of fit over a much smaller pelvis. And we've got these tiny structures for digestion, and yet we've got really expensive bodies. We've got large brains that are metabolically expensive. We've got relatively large bodies compared to chimpanzees,

which are actually our closest living relatives. and we actually show a lot of adaptations, surprisingly, for high energy expenditure. We tend to think of ourselves as just sitting around on the couch all day nowadays. But actually, humans as a species expend a lot of energy every day compared to a chimpanzee. I have to interrupt for a second. I remind our listeners that this is Science Friday from WNIC Studios. You can finish Dr. Comedy. Oh, great, thanks.

So I was just saying, so this combination of features of small structures for digestion, but a high energy budget means that we must have, around two million years ago, started eating a diet that wasn't just more of the same foods we were eating. We were actually eating a fundamentally different diet. We were packing more calories into a smaller amount of space. And particularly these small structures for digestion means that these foods must have been easy to chew and easy to break down within the gut. And so the long-standing hypothesis was, well, okay, maybe at this time in evolution,

we just started eating a ton of animal foods. We were hunters and we went after meat, and this explained all these changes that we see in the human body. But we don't actually see many adaptations that are specific for the digestion of animal foods. And so my colleague Richard Rangam really proposed that cooking and other forms of non-thermal food processing may actually have enabled ancestral humans to improve the foods that were readily available. And the single largest source of readily available calories out on the landscape

at the time would have been these underground storage organs like sweet potato we can think of it as a sweet potato tubers and so that's really set the stage we know that the human body has undergone these evolutionary changes for a high quality diet we think some of it may have come in from increasing the amount of meat we ate but we actually think a lot of it would have come in from cultural adaptations like non-thermal food processing as well as cooking that would have enabled us to just get more out of the foods that we had ready access to. I just have about a minute left. Is there any, where do you go from here in about 30 seconds?

Where do you go from the next kind of research you'd like to do? Well, so I think what our study with the microbes actually tells us is that just as the human body has changed in response to cooking, these microbes would have actually changed in response to us cooking as well. And so this sets up a stage where we can look for signals of. human microbial co-evolution. And my lab is actually pursuing this right now. We're looking at what is functionally unique about the human microbiomes in the context

of cooking. And we're also trying to understand the fundamental nature of the host microbial relationship, the conditions under which we compete with one another versus the conditions under which we are cooperative. We have to leave it right there, Dr. Carmen. We've run out of time. But thank you for that great explanation. No worries.

Thank you so much for having me. for your interest in our work. Rachel Carmony, Assistant Professor of Human Evolution or Biology at Harvard. One last thing before we go, if you left your heart in San Francisco, so did we. So we're going back for it. We'll be putting on a special evening of science entertainment. On Saturday, November 16th at the Sydney Goldstein Theater, will take a close look at the tiny adorable face mites that are living in the pores of your skin.

Talk about building ethical, artificial intelligence, much more. We're going to have videos and live music, science conversations, and your questions. So join us. Here are the details. Tickets and info, sciencefriiday.com slash San Francisco. ScienceFriety.com slash San Francisco will be in San Francisco Saturday night, November 16th, ScienceFriety.com slash San Francisco. Charles Berkowitz is our director, senior producer Christopher Antealiata.

Producers are Alexa Lim, Christy Taylor, Katie Feather. We also have production help today from. Daniel Adel Hamid, she and Elefator also were working with us. We had technical engineering help today for Mitch Kim and Kevin Wolf, B.J. Leiterman, composed our theme music. And let us know on the Science Friday Vox Pop app. It's there. Wherever you get your apps, you can talk back to us and maybe get on the radio. I'm Ira Flato in New York.