Your Digital Avatar May One Day Get Sick Before You Do
Artificial intelligence is everywhere, just not in the way you think it is.
These networks, loosely designed after the human brain, are interconnected computers that have the ability to "learn."
"There's the perception of AI in the glossy magazines," says Anders Kofod-Petersen, a professor of Artificial Intelligence at the Norwegian University of Science and Technology. "That's the sci-fi version. It resembles the small guy in the movie AI. It might be benevolent or it might be evil, but it's generally intelligent and conscious."
"And this is, of course, as far from the truth as you can possibly get."
What Exactly Is Artificial Intelligence, Anyway?
Let's start with how you got to this piece. You likely came to it through social media. Your Facebook account, Twitter feed, or perhaps a Google search. AI influences all of those things, machine learning helping to run the algorithms that decide what you see, when, and where. AI isn't the little humanoid figure; it's the system that controls the figure.
"AI is being confused with robotics," Eleonore Pauwels, Director of the Anticipatory Intelligence Lab with the Science and Technology Innovation Program at the Wilson Center, says. "What AI is right now is a data optimization system, a very powerful data optimization system."
The revolution in recent years hasn't come from the method scientists and other researchers use. The general ideas and philosophies have been around since the late 1960s. Instead, the big change has been the dramatic increase in computing power, primarily due to the development of neural networks. These networks, loosely designed after the human brain, are interconnected computers that have the ability to "learn." An AI, for example, can be taught to spot a picture of a cat by looking at hundreds of thousands of pictures that have been labeled "cat" and "learning" what a cat looks like. Or an AI can beat a human at Go, an achievement that just five years ago Kofod-Petersen thought wouldn't be accomplished for decades.
"It's very difficult to argue that something is intelligent if it can't learn, and these algorithms are getting pretty good at learning stuff. What they are not good at is learning how to learn."
Medicine is the field where this expertise in perception tasks might have the most influence. It's already having an impact as iPhones use AI to detect cancer, Apple watches alert the wearer to a heart problem, AI spots tuberculosis and the spread of breast cancer with a higher accuracy than human doctors, and more. Every few months, another study demonstrates more possibility. (The New Yorker published an article about medicine and AI last year, so you know it's a serious topic.)
But this is only the beginning. "I personally think genomics and precision medicine is where AI is going to be the biggest game-changer," Pauwels says. "It's going to completely change how we think about health, our genomes, and how we think about our relationship between our genotype and phenotype."
The Fundamental Breakthrough That Must Be Solved
To get there, however, researchers will need to make another breakthrough, and there's debate about how long that will take. Kofod-Petersen explains: "If we want to move from this narrow intelligence to this broader intelligence, that's a very difficult problem. It basically boils down to that we haven't got a clue about what intelligence actually is. We don't know what intelligence means in a biological sense. We think we might recognize it but we're not completely sure. There isn't a working definition. We kind of agree with the biologists that learning is an aspect of it. It's very difficult to argue that something is intelligent if it can't learn, and these algorithms are getting pretty good at learning stuff. What they are not good at is learning how to learn. They can learn specific tasks but we haven't approached how to teach them to learn to learn."
In other words, current AI is very, very good at identifying that a picture of a cat is, in fact, a cat – and getting better at doing so at an incredibly rapid pace – but the system only knows what a "cat" is because that's what a programmer told it a furry thing with whiskers and two pointy ears is called. If the programmer instead decided to label the training images as "dogs," the AI wouldn't say "no, that's a cat." Instead, it would simply call a furry thing with whiskers and two pointy ears a dog. AI systems lack the explicit inference that humans do effortlessly, almost without thinking.
Pauwels believes that the next step is for AI to transition from supervised to unsupervised learning. The latter means that the AI isn't answering questions that a programmer asks it ("Is this a cat?"). Instead, it's almost like it's looking at the data it has, coming up with its own questions and hypothesis, and answering them or putting them to the test. Combining this ability with the frankly insane processing power of the computer system could result in game-changing discoveries.
In the not-too-distant future, a doctor could run diagnostics on a digital avatar, watching which medical conditions present themselves before the person gets sick in real life.
One company in China plans to develop a way to create a digital avatar of an individual person, then simulate that person's health and medical information into the future. In the not-too-distant future, a doctor could run diagnostics on a digital avatar, watching which medical conditions presented themselves – cancer or a heart condition or anything, really – and help the real-life version prevent those conditions from beginning or treating them before they became a life-threatening issue.
That, obviously, would be an incredibly powerful technology, and it's just one of the many possibilities that unsupervised AI presents. It's also terrifying in the potential for misuse. Even the term "unsupervised AI" brings to mind a dystopian landscape where AI takes over and enslaves humanity. (Pick your favorite movie. There are dozens.) This is a concern, something for developers, programmers, and scientists to consider as they build the systems of the future.
The Ethical Problem That Deserves More Attention
But the more immediate concern about AI is much more mundane. We think of AI as an unbiased system. That's incorrect. Algorithms, after all, are designed by someone or a team, and those people have explicit or implicit biases. Intentionally, or more likely not, they introduce these biases into the very code that forms the basis for the AI. Current systems have a bias against people of color. Facebook tried to rectify the situation and failed. These are two small examples of a larger, potentially systemic problem.
It's vital and necessary for the people developing AI today to be aware of these issues. And, yes, avoid sending us to the brink of a James Cameron movie. But AI is too powerful a tool to ignore. Today, it's identifying cats and on the verge of detecting cancer. In not too many tomorrows, it will be on the forefront of medical innovation. If we are careful, aware, and smart, it will help simulate results, create designer drugs, and revolutionize individualize medicine. "AI is the only way to get there," Pauwels says.
Six Questions about the Kids' COVID Vaccine, Answered by an Infectious Disease Doctor
I enthusiastically support the vaccination against COVID for children aged 5-11 years old. As an infectious disease doctor who took care of hundreds of COVID-19 patients over the past 20 months, I have seen the immediate and long-term consequences of COVID-19 on patients – and on their families. As a father of two daughters, I have lived through the fear and anxiety of protecting my kids at all cost from the scourges of the pandemic and worried constantly about bringing the virus home from work.
It is imperative that we vaccinate as many children in the community as possible. There are several reasons why. First children do get sick from COVID-19. Over the course of the pandemic in the U.S, more than 2 million children aged 5-11 have become infected, more than 8000 have been hospitalized, and more than 100 have died, making COVID one of the top 10 causes of pediatric deaths in this age group over the past year. Children are also susceptible to chronic consequences of COVID such as long COVID and multisystem inflammatory syndrome in children (MIS-C). Most studies demonstrate that 10-30% of children will develop chronic symptoms following COVID-19. These include complaints of brain fog, fatigue, trouble breathing, fever, headache, muscle and joint pains, abdominal pain, mood swings and even psychiatric disorders. Symptoms typically last from 4-8 weeks in children, with some reporting symptoms that persist for many months.
Second, children are increasingly recognized as vectors who can bring infection into the house, potentially transmitting infection to vulnerable household members. Finally, we have all seen the mayhem that results when one child in the classroom becomes infected with COVID and the other students get sent home to quarantine – across the U.S., more than 2000 schools have been affected this way.
We now have an extraordinarily effective vaccine with more than 90 percent efficacy at preventing symptomatic infection. Vaccinating children will boost our countrywide vaccination rate which is trailing many countries after an early start. Nevertheless, there are still many questions and concerns that parents have as the vaccine gets rolled out. I will address six of them here.
"Novel Vaccine Technology"
Even though this is a relatively new vaccine, the technology is not new. Scientists had worked on mRNA vaccines for decades prior to the COVID mRNA vaccine breakthrough. Furthermore, experience with the Pfizer COVID vaccine is rapidly growing. By now it has been more than a year and a half since the Pfizer trials began in March 2020, and more than 7 billion doses have already been administered globally, including in 13.7 million adolescents in the U.S. alone.
"Will This Vaccine Alter My Child's DNA?"
No. This is not how mRNA works. DNA is present in the cell's nucleus. The mRNA only stays in the outside cytoplasm, gets destroyed and never enters the inner sanctum of the nucleus. Furthermore, for the mRNA to be ever integrated into DNA, it requires a special enzyme called reverse transcriptase which humans don't have. Proteins (that look like the spike proteins on SARS-CoV-2) are made directly from this mRNA message without involvement of our DNA at any time. Pieces of spike proteins get displayed on the outside of our cells and our body makes protective antibodies that then protects us handily against the future real virus if it were ever to enter our (or our children's) bodies. Our children's DNA or genes can never be affected by an mRNA vaccine.
"Lack of Info on Long-Term Side Effects"
Unlike medications that are taken daily or periodically and can build up over time, the mRNA in the Pfizer vaccine is evanescent. It literally is just the messenger (that is what the "m" in mRNA stands for) and the messenger quickly disappears. mRNA is extremely fragile and easily inactivated – that's why we need to encase it in a special fatty bubble and store the vaccines at extremely cold temperatures. Our cells break down and destroy the mRNA within a few days after receiving the instructions to make the virus spike proteins. The presence of these fragments of the virus (note this is not "live" virus) prompts our immune system to generate protective antibodies to the real thing. Our bodies break down mRNA all the time in normal cellular processes – this is nothing new.
What the transience of the delivery system means is that most of the effects of the mRNA vaccines are expected to be more immediate (sore arm, redness at the site, fever, chills etc.), with no long-term side effects anticipated. A severe allergic response has been reported to occur in some generally within the first 15 minutes, is very rare, and everyone gets observed for that as part of standard vaccine administration. Even with the very uncommon complication of myocarditis (inflammation of the heart muscle) and pericarditis (inflammation of the lining of the heart) seen primarily in young men under the age of 30 following mRNA vaccines, these typically happen within days to 2 weeks and many return to work or school in days. In the 70-year history of pediatric (and adult vaccines), dangerous complications happen in the first two months. There have been millions of adolescents as young as 12 years and thousands in the initial trial of children aged 5-11 who have already received the vaccine and are well beyond the two-month period of observation. There is no biological reason to believe that younger children will have a different long-term side effect profile compared to adolescents or adults.
"Small Sample Size in Kids and the Trial Design"
Although the Pfizer trial in children aged 5-11 was relatively small, it was big enough to give us statistical confidence in assessing safety and efficacy outcomes. Scientists spend a lot of time determining the right sample size of a study during the design phase. On one hand, you want to conduct the study efficiently so that resources are used in a cost-effective way and that you get a timely answer, especially in a fast-moving pandemic. On the other hand, you want to make sure you have enough sample size so that you can answer the question confidently as to whether the intervention works and whether there are adverse effects. The more profound the effect size of the intervention (in this case the vaccine), the fewer the numbers of children needed in the trials.
Statistics help investigators determine whether the results seen would have appeared by chance or not. In this case, the effect was real and impressive. Over 3,000 children around the world have received the vaccines through the trials alone with no serious side effects detected. The first press release reported that the immune response in children aged 5-11 was similar (at one-third the vaccine dose) to the response in the comparator group aged 16-25 years old. Extrapolating clinical efficacy results from immune response measurements ("immunobridging" study) would already have been acceptable if this was the only data. This is a standard trial design for many pediatric vaccines. Vaccines are first tested in the lab, followed by animals then adults. Only when deemed safe in adults and various regulatory bodies have signed off, do the pediatric vaccine trials commence.
Because children's immune systems and bodies are in a constant state of development, the vaccines must be right-sized. Investigators typically conduct "age de-escalation" studies in various age groups. The lowest dose is first tried so see if that is effective, then the dose is increased gradually as needed. Immune response is the easiest, safest and most efficient way to test the efficacy of pediatric vaccines. This is a typical size and design of a childhood vaccine seeking regulatory approval. There is no reason to think that the clinical efficacy would be any different in children vs. adults for a given antibody response, given the experience already in the remainder of the population, including older children and adolescents. Although this was primarily designed as an "immunobridging" study, the initial immunologic response data was followed by real clinical outcomes in this population. Reporting on the outcomes of 2,268 children in the randomized controlled trial, the vaccine was 90.7% effective at preventing symptomatic infection.
"Fear of Myocarditis"
Myocarditis (inflammation of the heart muscle) and pericarditis (inflammation of the lining of the heart) have been associated with receipt of the mRNA vaccines, particularly among male adolescents and young adults, typically within a few days after receiving the second dose. But this is very rare. For every million vaccine recipients, you would expect 41 cases in males, and 4 cases in females aged 12-29 years-old. The risk in older age groups is substantially lower. It is important to recognize that the risk of myocarditis associated with COVID is substantially higher. Patients present with new chest pain, shortness of breath, or palpitations after receiving an mRNA vaccine (more common after the second dose). But outcomes are good if associated with the vaccine. Most respond well to treatment and resolve symptoms within a week. There have been no deaths associated with vaccine-associated myocarditis.
In contrast, COVID-associated myocarditis has been associated with more severe cases as well as other complications including chronic symptoms of long COVID. The risk of myocarditis is likely related to vaccine dose, so the fact that one-third the dose of the vaccine will be used in the 5-11 year-olds is expected to correspond to a lower risk of myocarditis. At the lower dose given to younger kids, there has been a lower incidence of adverse effects reported compared to older children and adults who received the full dose. In addition, baseline rates of myocarditis not associated with vaccination are much lower in children ages 5-11 years than in older children, so the same may hold true for vaccine-associated myocarditis cases. This is because myocarditis is associated with sex hormones (particularly testosterone) that surge during puberty. In support of this, the incidence of vaccine-associated myocarditis is lower in 12–15-year-old boys, compared to those who were older than 16 years old. There were no cases of myocarditis reported in the experience to date of 5–11-year-old children in the trials, although the trial was too small to pick up on such a rare effect.
"Optimal Dose Spacing Interval: Longer Than 3 Weeks?"
There is a biologic basis for increasing the interval between vaccine doses in general. Priming the immune system with the first shot and then waiting gives the second shot a better chance of prompting a secondary immune reaction that results in a more durable response (with more T cell driven immune memory). One study from the U.K. showed that the antibody response in people over 80 was more than 3 times higher if they delayed the second dose to after 12 weeks for the Pfizer vaccine instead of the 3 weeks studied in trials. In a study of 503 British health care workers, there were twice as many neutralizing antibodies produced in a longer interval group (6-14 weeks) versus a shorter interval group (3-4 weeks) between doses. However, the safety and efficacy with longer intervals has not been evaluated in the pediatric or other COVID vaccine trials.
In the U.S., the C.D.C. reported that 88 percent of counties are at a "high" or "substantial" level of community transmission. Also, Europe is already experiencing a winter surge of infections that may predict more U.S. winter cases as international travel reopens. During a time of high community virus burden with a highly transmissible Delta variant, relying on one dose of vaccine for several more weeks until the second may leave many more susceptible to infection while waiting. One study from England showed that one dose of the Pfizer vaccine was only 33% protective against symptomatic Delta infection in contrast to 50% for the Alpha variant in adults. There has been no corollary information in children but we would expect less protection in general from one vaccine dose vs. two. This is a particularly important issue with the upcoming holiday season when an increased number of families will travel. Some countries such as the U.K. and Norway have proceeded with only offering older than 12 year-olds one dose of vaccine rather than two, but this was before the current European surge which may change the risk-benefit calculus. There are no plans to only offer one vaccine dose in the U.S. at this time. However a lower dose of the vaccine will likely be studied in the future for adolescents aged 12-15.
For parents worried about the potential risk of adverse effects of two doses of vaccines in their children, it is reasonable to wait 6-12 weeks for the second shot but it all depends on your risk-benefit calculus. There is biological plausibility to pursue this strategy. Although there is no pediatric-specific data to draw from, a longer interval may lengthen immune memory and potentially decrease the risk of myocarditis, particularly in boys. There may only be partial benefit in eliciting protective antibodies after one vaccine dose but only 2-4% of children are hospitalized with COVID once infected, with risk of severe illness increasing if they have comorbidities.
There are also some data indicating that 40% of children have already been exposed to infection naturally and may not need further protection after one shot. However, this percentage is likely a large overestimation given the way the data was collected. Using antibody tests to ascertain previous infection in children may be problematic for several reasons: uncertainty regarding duration of protection, variability in symptoms in children with most having very mild symptoms, and the lack of standardization of antibody tests in general. Overall, if the child has medical comorbidities such as diabetes, parents are planning to travel with their children, if local epidemiology shows increasing cases, and if there are elderly or immunocompromised individuals in the household, I would vaccinate children with two doses as per the original recommended schedule.
Bottom line: Given the time of the year and circulating Delta, I would probably stick with the recommended 3-week interval between doses for now for most children. But if parents choose a longer interval between the first and second dose for their children, I wouldn't worry too much about it. Better to be vaccinated - even if slowly, over time -- than not at all.
Food Poisoning Sickens Millions a Year. Now, a Surprising Weapon Is Helping Protect Against Contamination.
Every year, one in seven people in America comes down with a foodborne illness, typically caused by a bacterial pathogen, including E.Coli, listeria, salmonella, or campylobacter. That adds up to 48 million people, of which 120,000 are hospitalized and 3000 die, according to the Centers for Disease Control. And the variety of foods that can be contaminated with bacterial pathogens is growing too. In the 20th century, E.Coli and listeria lurked primarily within meat. Now they find their way into lettuce, spinach, and other leafy greens, causing periodic consumer scares and product recalls. Onions are the most recent suspected culprit of a nationwide salmonella outbreak.
Some of these incidents are almost inevitable because of how Mother Nature works, explains Divya Jaroni, associate professor of animal and food sciences at Oklahoma State University. These common foodborne pathogens come from the cattle's intestines when the animals shed them in their manure—and then they get washed into rivers and lakes, especially in heavy rains. When this water is later used to irrigate produce farms, the bugs end up on salad greens. Plus, many small farms do both—herd cattle and grow produce.
"Unfortunately for us, these pathogens are part of the microflora of the cows' intestinal tract," Jaroni says. "Some farmers may have an acre or two of cattle pastures, and an acre of a produce farm nearby, so it's easy for this water to contaminate the crops."
Food producers and packagers fight bacteria by potent chemicals, with chlorine being the go-to disinfectant. Cattle carcasses, for example, are typically washed by chlorine solutions as the animals' intestines are removed. Leafy greens are bathed in water with added chlorine solutions. However, because the same "bath" can be used for multiple veggie batches and chlorine evaporates over time, the later rounds may not kill all of the bacteria, sparing some. The natural and organic producers avoid chlorine, substituting it with lactic acid, a more holistic sanitizer, but even with all these efforts, some pathogens survive, sickening consumers and causing food recalls. As we farm more animals and grow more produce, while also striving to use fewer chemicals and more organic growing methods, it will be harder to control bacteria's spread.
"It took us a long time to convince the FDA phages were safe and efficient alternatives. But we had worked with them to gather all the data they needed, and the FDA was very supportive in the end."
Luckily, bacteria have their own killers. Called bacteriophages, or phages for short, they are viruses that prey on bacteria only. Under the electron microscope, they look like fantasy spaceships, with oblong bodies, spider-like legs and long tails. Much smaller than a bacterium, phages pierce the microbes' cells with their tails, sneak in and begin multiplying inside, eventually bursting the microbes open—and then proceed to infect more of them. The best part is that these phages are harmless to humans. Moreover, recent research finds that millions of phages dwell on us and in us—in our nose, throat, skin and gut, protecting us from bacterial infections as part of our healthy microbiome. A recent study suggested that we absorb about 30 billion phages into our bodies on a daily basis. Now, ingeniously, they are starting to be deployed as anti-microbial agents in the food industry.
A Maryland-based phage research company called Intralytix is doing just that. Founded by Alexander Sulakvelidze, a microbiologist and epidemiologist who came to the United States from Tbilisi, the capital of Georgia, Intralytix makes and sells five different FDA-approved phage cocktails that work against some of the most notorious food pathogens: ListShield for Listeria, SalmoFresh for Salmonella, ShigaShield for Shigella, another foodborne bug, and EcoShield for E.coli, including the infamous strain that caused the Jack in the Box outbreak in 1993 that killed four children and sickened 732 people across four states. Earlier this month, the FDA granted its approval to yet another Intralytix phage for managing Campylobacter contamination, named CampyShield. "We call it safety by nature," Sulakvelidze says.
Intralytix grows phages inside massive 1500-liter fermenters, feeding them bacterial "fodder."
Photo credit: Living Radiant Photography
Phage preparations are relatively straightforward to make. In nature, phages thrive in any body of water where bacteria live too, including rivers, lakes and bays. "I can dip a bucket into the Chesapeake Bay, and it will be full of all kinds of phages," Sulakvelidze says. "Sewage is another great place to look for specific phages of interest, because it's teeming with all sorts of bacteria—and therefore the viruses that prey on them." In lab settings, Intralytix grows phages inside massive 1500-liter fermenters, feeding them bacterial "fodder." Once phages multiply enough, they are harvested, dispensed into containers and shipped to food producers who have adopted this disinfecting practice into their preparation process. Typically, it's done by computer-controlled sprayer systems that disperse mist-like phage preparations onto the food.
Unlike chemicals like chlorine or antibiotics, which kill a wide spectrum of bacteria, phages are more specialized, each feeding on specific microbial species. A phage that targets salmonella will not prey on listeria and vice versa. So food producers may sometimes use a combo of different phage preparations. Intralytix is continuously researching and testing new phages. With a contract from the National Institutes of Health, Intralytix is expanding its automated high-throughput robot that tests which phages work best against which bacteria, speeding up the development of the new phage cocktails.
Phages have other "talents." In her recent study, Jaroni found that phages have the ability to destroy bacterial biofilms—colonies of microorganisms that tend to grow on surfaces of the food processing equipment, surrounding themselves with protective coating that even very harsh chemicals can't crack. "Phages are very clever," Jaroni says. "They produce enzymes that target the biofilms, and once they break through, they can reach the bacteria."
Convincing the FDA that phages were safe to use on food products was no easy feat, Sulakvelidze says. In his home country of Georgia, phages have been used as antimicrobial remedies for over a century, but the FDA was leery of using viruses as food safety agents. "It took us a long time to convince the FDA phages were safe and efficient alternatives," Sulakvelidze says. "But we had worked with them to gather all the data they needed, and the FDA was very supportive in the end." The agency had granted Intralytix its first approval in 2006, and over the past 10 years, the company's sales increased by over 15-fold. "We currently sell to about 40 companies and are in discussions with several other large food producers," Sulakvelidze says. One indicator that the industry now understands and appreciates the science of phages was that his company was ranked as Top Food Safety Provider in 2021 by Food and Beverage Technology Review, he adds. Notably, phage sprays are kosher, halal and organic-certified.
Intralytix's phage cocktails to safeguard food from bacteria are approved for consumers in addition to food producers, but currently the company sells to food producers only. Selling retail requires different packaging like easy-to-use spray bottles and different marketing that would inform people about phages' antimicrobial qualities. But ultimately, giving people the ability to remove pathogens from their food with probiotic phage sprays is the goal, Sulakvelidze says.
It's not the company's only goal. Now Intralytix is going a step further, investigating phages' probiotic and therapeutic abilities. Because phages are highly specialized in the bacteria they target, they can be used to treat infections caused by specific pathogens while leaving the beneficial species of our microbiome intact. In an ongoing clinical trial with Mount Sinai, Intralytix is now investigating a potential phage treatment against a certain type of E. coli for patients with Crohn's disease, and is about to start another clinical trial for treating bacterial dysentery.
"Now that we have proved that phages are safe and effective against foodborne bacteria," Sulakvelidze says, "we are going to demonstrate their potential in therapeutic applications."
Lina Zeldovich has written about science, medicine and technology for Popular Science, Smithsonian, National Geographic, Scientific American, Reader’s Digest, the New York Times and other major national and international publications. A Columbia J-School alumna, she has won several awards for her stories, including the ASJA Crisis Coverage Award for Covid reporting, and has been a contributing editor at Nautilus Magazine. In 2021, Zeldovich released her first book, The Other Dark Matter, published by the University of Chicago Press, about the science and business of turning waste into wealth and health. You can find her on http://linazeldovich.com/ and @linazeldovich.