Every year, scientists make the arduous journey to the Tibetan Plateau to try to understand how organisms can live there.

How Earth's most extreme environs are key to humanity's future

Future Society

As the planet warms, scientists are looking to organisms living in Earth's extreme environments to engineer hardier crops.

How Earth's Most Extreme Environs Are Key to Humanity's Future

“In an evolutionary sense, we don’t have any time to adapt."

How Earth’s Most Extreme Environs Are Key to Humanity’s Future

<h2> 1918: The War We Lost</h2>
In 1918, the United States fought two wars. One it lost, and one it won.
You may have learned about World War I in history class, or even from your relatives. As a member of the Allied Forces, the United States defeated <a href="https://www.britannica.com/topic/Central-Powers">the Central Powers</a> — a victory touted by history books, movies, and novels.
The second war, however, had a more elusive opponent. It descended perniciously, quietly claiming lives while armies concerned themselves with foxholes and mustard gas. In the first six months, this enemy killed <a href="https://www.scientificamerican.com/article/busting-10-common-myths-about-the-greatest-pandemic-in-history/">25 million people</a> worldwide.
Ultimately, between <a href="https://www.smithsonianmag.com/history/journal-plague-year-180965222/">50 and 100 million lives</a> — five percent of the world’s population at the time — would be lost as a result of the conflict.
This second enemy was, of course, the flu virus. By the time Americans realized that the country was under siege, it was too late to stop it. The flu made its way through the U.S., Europe, and Asia with terrifying speed; people who had been well in the morning <a href="https://virus.stanford.edu/uda/">dropped dead in the street</a> by dinner time. Families that had already lost sons, fathers, and brothers to the war abroad dwindled as the virus attacked them, <a href="https://wwwnc.cdc.gov/eid/article/18/2/10-2042_article">affecting the remaining young and healthy</a>. In just one year, the average life expectancy for an American <a href="https://www.archives.gov/exhibits/influenza-epidemic/">dropped by 12 years.</a>
Over the century that followed, Americans would face <a href="http://www.apnewsarchive.com/2018/Scientists-hunting-universal-vaccine-to-protect-against-regular-winter-flu-and-the-next-super-flu-too/id-f059c92c0ec946dc8957e2ad9c0affc2">three more pandemic flus</a>, but none of them like the one in 1918. The 1957 pandemic flu killed roughly 1.1 million people worldwide; another in 1968 wiped out about another million globally. Most recently, the 2009 H1N1 pandemic flu killed between 151,700 and 575,400 people worldwide, according to <a href="https://www.cdc.gov/flu/pandemic-resources/basics/past-pandemics.html">estimates from</a> the Centers for Disease Control and Prevention (CDC).
Today, a century after the 1918 pandemic, we know much more about the virus — how it spreads, how it kills. We now have influenza vaccines — unheard of in 1918 — that provide us with (albeit limited) protection. And sophisticated tracking mechanisms help us predict which flu viruses we might encounter in a given year.
We have not, however, completely vanquished the flu. In this <a href="https://www.popsci.com/how-bad-flu-season-2018">particularly bad flu season</a> in the U.S., we need little reminder that the virus is hardy and evolves rapidly. The flu that ravaged humanity in 1918 is not the same strain making headlines in 2018. Likewise, if another global pandemic flu is inevitable, we can’t assume the virus will be one we’ve seen before.
Today, our relationship with the flu has shifted from an adversarial, bellicose one, to one of competition; we are running a race, no longer fighting a war. To survive another century, or another season, public health experts will need stay one step ahead, armed with an artillery provided by science and a war plan drafted from the history of the battle we lost.
<figure id="attachment_122168" style="width: 1200px" class="wp-caption alignnone"><img class="size-large wp-image-122168" src="https://wp-assets.futurism.com/2018/02/influenza2-1200x661.jpg" alt="" width="1200" height="661" /><figcaption class="wp-caption-text">Image Credit: <a href="https://commons.wikimedia.org/wiki/File:Women_wearing_surgical_masks_during_the_influenza_epidemic,_1919_(29245747812).jpg">State Library of Queensland</a>/Illustration by Victor Tangermann</figcaption></figure>
<h2>Why (and How) the Flu Still Kills</h2>
A high fever, fluid in the lungs, crushing fatigue, and body aches — if you’ve ever <a href="https://www.niaid.nih.gov/diseases-conditions/influenza">come down with influenza</a>, it likely needs no introduction. It’s often easy to distinguish the full-blown flu from the common cold because the flu’s symptoms tend to come on suddenly and with an intensity that makes it hard to deny.
When a person is infected by any pathogen — a virus or bacteria — they usually won’t know it until that pathogen <a href="http://needtoknow.nas.edu/id/infection/how-pathogens-make-us-sick/">has started damaging cells</a>. That kicks the immune system into gear, making you start to feel sick. The fever, aches, and mucus all too familiar to flu-sufferers aren’t from the virus itself, but rather are the side effects of the body’s attempt to vanquish it.
Even though our immune systems respond rapidly and with such force, they aren’t always successful in stopping the microbes wreaking havoc on the body’s cells. While most of us who get the flu just stay home and rest, the flu makes some people seriously ill — they have to be hospitalized. Some even die as a result of complications from the flu.
(The flu doesn’t directly cause death. Instead, the virus can induce an infection like pneumonia, or exacerbate an underlying condition. But oddly enough, it’s usually the body’s <a href="https://www.scientificamerican.com/article/how-does-the-flu-actually-kill-people/">too-aggressive immune response</a> to the flu that ultimately kills people).
A <a href="https://www.cdc.gov/flu/about/disease/spread.htm">flu virus spreads</a> when a healthy person ingests or inhales virus-infected droplets flung into the air by a sick person’s cough, sneeze, or <a href="https://medicalxpress.com/news/2018-01-flu-required.html">mere breath</a>. The CDC <a href="https://www.cdc.gov/flu/about/disease/us_flu-related_deaths.htm">does not know exactly how many people get the flu each year</a>. The agency doesn’t know how many people die from it either. People who come down with the flu don’t always seek medical attention. Even when they do, <a href="https://www.cdc.gov/flu/professionals/diagnosis/labrolesprocedures.htm">doctors don’t always test for it.</a>
Those caveats make the data on this year’s flu season more striking: as of the first week of February, the number of flu cases in the United States was the highest since the 2009 pandemic. The most people have been hospitalized at this stage in the flu season <a href="https://www.nytimes.com/2018/02/02/health/flu-symptoms-virus-hospital.html?rref=collection%2Fsectioncollection%2Fhealth">since the CDC started tracking, in 2005</a>. Both numbers are still climbing.
When we talk about the flu, we aren’t talking about a single virus. There are four types of influenza viruses — but only two of them cause serious illness in humans, <a href="http://phymbie.physics.ryerson.ca/~cbeau/">Catherine Beauchemin</a>, an associate professor of virophysics at Ryerson University in Toronto, explained to Futurism. You might remember hearing about H1N1 (the flu type that hit us in 2009) and H3N2 (the type of flu causing problems this year) — those Hs and Ns stand for hemagglutinin and neuraminidase, proteins found on virus’ surfaces that help either enter cells (H) or separate from cells to go infect another cell (N). The numbers identify groups of strains with similar Hs and Ns.
The flu <a href="https://www.cdc.gov/flu/about/viruses/change.htm">mutates remarkably quickly</a>, changing dramatically to dodge our antibodies in the span of a flu season or two. That means it can infect people who previously contracted it.
That’s why we get flu shots every year. Even though researchers have a sophisticated global tracking system to anticipate which strain might affect a region in a given year, there’s still a surprising amount of guesswork involved.
Flu seasons typically occur during the colder months, when people are more likely to congregate indoors. Because the flu season is <a href="https://www.nytimes.com/2017/09/19/world/australia/flu-vaccination-deaths.html">opposite in Australia</a>, the CDC’s <a href="https://www.cdc.gov/ncird/flu.html">Epidemiology and Prevention Branch in the Influenza Division</a> can track that country’s flu season about six months before flu season arrives in North America. As travelers move the virus from Australia to Europe, Asia, and the U.S., public health experts can anticipate which strain will likely be the one to make people sick in the northern hemisphere that year.
The system, and the vaccine made from it, is far from perfect though. “The issue is that the recommendations have to be made some six months before the vaccine is actually used,” Richard Webby, Director of the <a href="http://www.who.int/influenza/vaccines/virus/temporary_advisors/en/index7.html">WHO Collaborating Center for Studies on the Ecology of Influenza in Animals</a> and a member of the <a href="https://www.stjude.org/webby">Department of Infectious Disease</a> at St. Jude Children’s Research Hospital, told Futurism. Researchers need that time to analyze data from Australia’s flu season, then manufacture and distribute the vaccines.
For a virus that evolves so quickly, that lead time can also be problematic. “There have been instances where the viruses have changed between when the recommendations have been made and when the vaccine has been administered, leading to suboptimal performance.” Webby added. For example, the latest data on this year’s flu vaccine shows it’s around <a href="https://www.statnews.com/2018/02/01/flu-vaccine-protection-h3n2/">17 percent effective</a>, though that may change before the flu season ends.
This year’s flu virus, H3N2, <a href="http://www.pnas.org/content/114/47/12578.long">isn’t like other strains</a> that have circulated in recent years. It binds to cells differently, and seems to be mutating more rapidly, making it difficult to study and create a vaccine against. The strain also doesn’t <a href="https://www.cdc.gov/flu/protect/vaccine/how-fluvaccine-made.htm">grow well in eggs</a>, where bacteria are most commonly grown before being put into vaccines.
“We don’t have a flu vaccine problem so much as we have an H3N2 vaccine problem,” Ed Belongia, a vaccine researcher and director of the Center for Clinical Epidemiology and Population Health at Wisconsin’s Marshfield Clinic, <a href="https://www.statnews.com/2018/01/08/flu-virus-h3n2/">recently told STAT News</a>.
Although we can identify and classify them, track them, and create vaccines to defend against them, the viruses continue to evade us, evolving faster than we can keep up — sickening or killing people in the process.
<figure id="attachment_122170" style="width: 1200px" class="wp-caption alignnone"><img class="size-large wp-image-122170" src="https://wp-assets.futurism.com/2018/02/influenza3-1200x661.jpg" alt="" width="1200" height="661" /><figcaption class="wp-caption-text">Image Credit: <a href="https://commons.wikimedia.org/w/index.php?search=influenza+1919&amp;title=Special:Search&amp;profile=default&amp;fulltext=1&amp;searchToken=77o9nhgu8n2lunoxe5jvlzqwa#/media/File:Influenza_Vaccination_1976.jpg">CDC</a>/Illustration by Victor Tangermann</figcaption></figure>
<h2>Fighting the Flu of the Future</h2>
In 1918, many of the treatments we have today for secondary infections like pneumonia or strep throat either didn’t exist or were not yet widely available. That partially explains why the epidemic killed so many.
Today, <a href="https://www.gene.com/download/pdf/tamiflu_prescribing.pdf">the antiviral Tamiflu</a> can quell symptoms within the first 48 hours of their onset, or even <a href="http://onlinelibrary.wiley.com/doi/10.1002/14651858.CD008965.pub4/abstract;jsessionid=76F622C5B03F0FF262E8B76396219022.f03t01">prevent them</a> in the first place. But it’s pricey (a five-day course costs <a href="https://www.consumerreports.org/cro/news/2013/01/should-i-take-tamiflu-to-treat-the-flu/index.htm">$100 minimum</a>) and comes with risks, especially for children and teens, who are more likely to experience serious <a href="https://www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafetyInformationforPatientsandProviders/ucm107840.htm">psychological side effects</a> and “seizures, confusion, or abnormal behavior early during their illness,” <a href="https://www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafetyInformationforPatientsandProviders/ucm188859.htm">according to the CDC</a>.
In 1918, many people felt that the flu descended upon their community out of nowhere. Today, we can at the very least see the flu coming so our doctors and emergency rooms can be prepared — even if we don’t have weapons powerful enough to completely stop it yet.
One elegant solution is to gather data from smart devices sick people usually use to track the spread of the flu. <a href="https://www.kinsahealth.com/">Smart thermometer company Kinsa</a> does just that. Over the past six years, the device’s 1 million users gather real-time data to track infectious disease with the help of “smart thermometers” and a smartphone app. Though it may seem counter-intuitive that a relatively small number of users could track how many people have the flu and where, the flu-tracking data over the past two years has lined up with CDC data — and the app is gathering it much more quickly than public health agencies are able. Nationally, the number of people with the flu is 39 percent higher than it was at this time last year, according to Kinsa&#8217;s most recent report.
Some are thinking bigger than treating or tracking the flu. The holy grail for flu treatment would be a vaccine that doesn’t change from year to year depending on the annual strain. If everyone could just get the vaccine once to protect us from all strains of flu for our entire lives, <a href="http://www.who.int/mediacentre/factsheets/fs211/en/">hundreds of thousands of lives</a> could be saved every year.
We’re talking, of course, about a <a href="https://medicalxpress.com/news/2018-01-firm-advances-human-trials-revolutionary.html">universal flu vaccine</a>.
A team of researchers out of UCLA is <a href="http://science.sciencemag.org/content/359/6373/290">genetically-engineering flu viruses</a> that could become candidates for a universal vaccine. The researchers engineered flu cells to stimulate a bigger, more targeted immune response than the real-life strains. So far, the team has only developed the potential vaccine in the lab; the researchers hope to test two strains in animal models before moving into human trials.
Pharma company <a href="http://www.biondvax.com/">BiondVax Pharmaceuticals</a> recently completed Phase 3 clinical trials for its universal vaccine candidate, which incorporates synthetic compounds. It has <a href="http://www.biondvax.com/2017/12/biondvax-plans-phase-3-clinical-trial-following-receipt-of-scientific-advice-from-the-european-medicines-agency-ema/">already received a patent in India</a>. This type of vaccine targets specific areas on the surface of a flu virus that determine the phase and severity of the immune response. Being able to “ramp up” or “tamp down” different aspects of that process in animal models has convinced researchers that the vaccine could be useful in preventing other infectious disease beyond the flu, such as HIV and malaria.
FluGen, a startup out of the University of Wisconsin-Madison, is also working with a genetically-mutated form of the virus to make a universal vaccine. <a href="http://flugen.com/redee-flu/">According to FluGen’s website</a>, the company&#8217;s genetically-altered  viruses have had a gene deleted so that they “can infect cells, express the entire spectrum of influenza RNA and proteins, yet cannot produce any infectious virus particles.”
But to get there, the researchers encountered substantial controversy. You have to break a few eggs to make an omelet; to create a vaccine against mutating flu viruses, you’ll have to mutate a few flu viruses. Researchers worked to avoid creating some kind of super-virus. When the researchers mutated the H1N1 virus from the 2009 pandemic, and when they recreated the 1918 pandemic flu, the global scientific community <a href="http://www.independent.co.uk/news/science/exclusive-controversial-us-scientist-creates-deadly-new-flu-strain-for-pandemic-research-9577088.html">called their methods and safety into question</a>.
Other researchers, like those on a team at Georgia State University, are harnessing <a href="https://futurism.com/universal-vaccines-are-future-medicine/">nanoparticles to facilitate a universal vaccine</a>. Most vaccines target the outside surface of a virus’s protein, which varies across different viruses. But if nanoparticles could target further down, on a part of the protein called the stalk, a vaccine could have broader efficacy. In experiments detailed <a href="https://www.nature.com/articles/s41467-017-02725-4">in a study published in Nature Communications</a> in January 2018, mice inoculated with nanoparticles containing the protein to elicit an immune system response were completely immune to four different strains of the flu, including this year’s H3N2. They will need to conduct more animal studies — first in ferrets, as their respiratory systems are quite similar to those of humans — before testing the vaccine on humans.
There are other logistical hurdles to a universal vaccine. There’s <a href="https://www.wired.com/story/flu-vaccine-big-pharma/">little financial incentive</a> for pharmaceutical companies to develop vaccines, much less universal ones only administered once in a person’s life. Distribution of vaccines can be challenging and <a href="https://www.cdc.gov/vaccines/hcp/clinical-resources/shortages.html">shortages are not uncommon</a>. Plus, <a href="https://futurism.com/reasons-flu-season-devastating/">people just love to find reasons</a> why they shouldn’t get the jab.
But these challenges are not insurmountable. A universal vaccine could be possible within a generation. How well it works, well, that&#8217;s another question.
As 1918 came to a close, the editors of the Journal of the American Medical Association published <a href="https://virus.stanford.edu/uda/">its final edition</a> for the year. The editors reflected on what could be learned from the two wars humanity fought that year, then turned their attention to the future.
“Medical science for four and one-half years devoted itself to putting men on the firing line and keeping them there,” they wrote. “Now, it must turn with its whole might to combating the greatest enemy of all — infectious disease.” In another century, perhaps the flu of today — the damage it causes, the lives lost to it — will seem equally distant, perhaps even innocuous.

This feels even more relevant during this bad flu season.

In 1918 we faced the flu pandemic. Today, we're still fighting the war.

A century ago, the flu killed millions. We're still not rid of our viral enemy. In the future, maybe the flu will truly be a thing of the past.

In 1918 We Faced the Flu Pandemic. Today, We’re Still Fighting the War.

In another century, maybe we'll be rid of the flu for good.

Celebrities need to stop spreading misinformation about the food we eat.

Is organic really better? 4 food myths debunked by science

Is eating organic really healthier? Four of the biggest misconceptions about organic food that need to be challenged.

Is Organic Really Better? 4 Food Myths Debunked By Science

GMOs aren't as scary as they seem. 

Reproductive medicine is already starting to look a lot like the future. But how can we get there when our discrimination laws are stuck in the past?

Advanced reproductive technology is here. But who decides who gets access?

Sci-Fi Visions

In the U.S., people seeking reproductive medicine can get anything they can pay for. But doctors can still discriminate against potential patients.

Advanced Reproductive Technology is Here. But Who Decides Who Gets Access?

"There is no question that in America you have no right to reproduce."

Tesla brought power to a hospital in hurricane-ravaged Puerto Rico. Could there be a downside?

When it comes to natural disasters, technology has an unavoidable dark side

After disasters like wildfires and hurricanes, the government doesn't always step in quickly enough. So companies have begun to fill the gap.

When It Comes To Natural Disasters, Technology Has An Unavoidable Dark Side

Companies arrive quicker at disaster sites, experimental technology in tow. That might not be great in the long run. 

<h2>Diagnosing with “The Stethoscope of the 21st Century”</h2>
A new kind of doctor has entered the exam room, but doesn&#8217;t have a name. In fact, these doctors don’t even have faces. Artificial intelligence has made its way into hospitals around the world. Those wary of a robot takeover have nothing to fear; the introduction of AI into health care is not necessarily about pitting human minds against machines. AI is in the exam room to expand, sharpen, and at times ease the mind of the physician so that doctors are able to do the same for their patients.
Bertalan Meskó, known better as <a href="http://medicalfuturist.com/about/">The Medical Futurist,</a> has called artificial intelligence the “<a href="http://medicalfuturist.com/ibm-watson-is-the-stethoscope-of-the-21st-century/">the stethoscope of the 21st century</a>.” His assessment may prove to be even more accurate than he expected. While various techniques and tests give them all the information they need to diagnose and treat patients, physicians are already overburdened with clinical and administrative responsibilities, and sorting through the massive amount of available information is a daunting, if not impossible, task.
That’s where having the 21st century stethoscope could make all the difference.
The applications for AI in medicine go beyond administrative drudge work, though. From powerful diagnostic algorithms to finely-tuned surgical robots, the technology is making its presence known across medical disciplines. Clearly, AI has a place in medicine; what we don’t know yet is its value. To imagine a future in which AI is an established part of a patient’s care team, we’ll first have to better understand how AI measures up to human doctors. How do they compare in terms of accuracy? What specific, or unique, contributions is AI able to make? In what way will AI be most helpful — and could it be potentially harmful — in the practice of medicine? Only once we’ve answered these questions can we begin to predict, then build, the AI-powered future that we want.
<h2>AI vs. Human Doctors</h2>
Although we are still in the early stages of its development, AI is already just as capable as (if not more capable than) doctors in diagnosing patients. Researchers at the <a href="http://www.ouh.nhs.uk/hospitals/jr/">John Radcliffe Hospital</a> in Oxford, England, developed an AI diagnostics system that’s <a href="https://futurism.com/ai-diagnose-heart-disease-lung-cancer-more-accurately-doctors/">more accurate than doctors</a> at diagnosing heart disease, at least <a href="http://www.ultromics.com/technology/">80 percent of the time.</a> At Harvard University, researchers created a <a href="https://futurism.com/ai-equipped-microscope-diagnose-deadly-blood-infections/">“smart” microscope</a> that can detect potentially lethal blood infections: the AI-assisted tool was trained on a series of 100,000 images garnered from 25,000 slides treated with dye to make the bacteria more visible. The AI system can already sort those bacteria with a 95 percent accuracy rate. A study from <a href="http://www.showa-u.ac.jp/en/index.html">Showa University</a> in Yokohama, Japan revealed that a <a href="https://futurism.com/ai-assisted-detection-identifies-colon-cancer-automatically-and-in-real-time/">new computer-aided endoscopic system</a> can reveal signs of potentially cancerous growths in the colon with 94 percent sensitivity, 79 percent specificity, and 86 percent accuracy.
In some cases, researchers are also finding that AI can outperform human physicians in diagnostic challenges that require a quick judgment call, such as determining if a lesion is cancerous. <a href="https://jamanetwork.com/journals/jama/article-abstract/2665774?redirect=true">In one study</a>, published December 2017 in JAMA, deep learning algorithms were able better diagnose metastatic breast cancer than human radiologists <a href="https://www.axios.com/ai-advances-at-detecting-cancer-but-it-cant-see-you-now-2517480809.html">when under a time </a><a href="https://www.axios.com/ai-advances-at-detecting-cancer-but-it-cant-see-you-now-2517480809.html">crunch</a>. While human radiologists may do well when they have unrestricted time to review cases, in the real world (especially in high-volume, quick-turnaround environments like emergency rooms) a rapid diagnosis could make the difference between life and death for patients.
Then, of course, there’s <a href="https://futurism.com/artificial-intelligence-is-completely-transforming-modern-healthcare/">IBM’s Watson:</a> When challenged to glean meaningful insights from the genetic data of tumor cells, human experts took about 160 hours to review and provide treatment recommendations based on their findings. <a href="https://singularityhub.com/2017/11/30/the-doctor-in-the-machine-how-ai-is-saving-lives-in-healthcare/#sm.000kudnmv10q0d4iwlu271m4dwx70">Watson took just ten minutes</a> to deliver the same kind of actionable advice. Google recently announced an open-source version of DeepVariant, the company’s AI tool for parsing genetic data, which was the most accurate tool of its kind in last year’s <a href="https://www.google.com/search?q=precisionFDA+truth+challenge&amp;oq=precisionFDA+truth+challenge&amp;aqs=chrome..69i57.4096j0j4&amp;sourceid=chrome&amp;ie=UTF-8">precisionFDA Truth Challenge</a>.
<figure id="attachment_120080" style="width: 1200px" class="wp-caption aligncenter"><img class="wp-image-120080 size-large" src="https://wp-assets.futurism.com/2018/01/Doctor_reviewing_pdq-1200x800.jpg" alt="" width="1200" height="800" /><figcaption class="wp-caption-text">With AI in medicine, doctors will have to spend less time doing this. Image Credit: <a href="https://commons.wikimedia.org/wiki/File:Doctor_reviewing_pdq.jpg">National Cancer Institute via Wikimedia Commons</a></figcaption></figure>
AI is also better than humans at predicting health events before they happen. In April, researchers from the University of Nottingham <a href="http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0174944">published a study</a> that showed that, trained on extensive data from 378,256 patients, a self-taught AI predicted 7.6 percent more cardiovascular events in patients than the <a href="http://www.acc.org/guidelines">current standard of care</a>. To put that figure in perspective, <a href="http://www.sciencemag.org/news/2017/04/self-taught-artificial-intelligence-beats-doctors-predicting-heart-attacks">the researchers wrote</a>: “In the test sample of about 83,000 records, that amounts to 355 additional patients whose lives could have been saved.” Perhaps most notably, the neural network also had 1.6 percent fewer “false alarms” — cases in which the risk was overestimated, possibly leading to patients having unnecessary procedures or treatments, many of which are very risky.
AI is perhaps most useful for making sense of huge amounts of data that would be overwhelming to humans. That’s exactly what’s needed in the growing field of precision medicine. Hoping to fill that gap is <a href="https://www.humandx.org/">The Human Diagnosis Project</a> (Human Dx), which is combining machine learning with doctors’ real-life experience. The organization is compiling input from 7,500 doctors and 500 medical institutions in more than 80 countries in order to develop a system that anyone — patient, doctor, organization, device developer, or researcher — can access in order to make more informed clinical decisions.
<q class="pullquote">“You have to design these things with an end user in mind.&#8221;</q>
Shantanu Nundy, the Director of the Human Diagnosis Project Nonprofit, told Futurism that, when it comes to developing technology in any industry, the AI should be seamlessly integrated into its function. “You have to design these things with an end user in mind. People use Netflix, but it&#8217;s not like ‘AI for watching movies,’ right? People use Amazon, but it&#8217;s not like ‘AI for shopping.’”
In other words, if the tech is designed well and implemented in a way that people find useful, people don’t even realize they’re using AI at all.
For open-minded, forward-thinking clinicians, the immediate appeal of projects like Human Dx is that it would, counterintuitively, allow them to spend less time engaged with technology. “It&#8217;s been well-documented that over 50 percent of our time now is in front of a screen,” Nundy, who is also a practicing physician in the D.C. area, told Futurism. AI can give doctors some of that time back by allowing them to offload some of the administrative burdens, like documentation.
In this respect, when it comes to healthcare, AI isn’t necessarily about replacing doctors, but optimizing and improving their abilities.
<h2>Mental Health Care with a Human Touch</h2>
“I see the value of AI today as augmenting humans, not as replacing humans,” Skyler Place, chief behavioral science officer in the mobile health division at <a href="http://www.cogitocorp.com/">Cogito, a Boston-based AI and behavioral analytics company</a>, told Futurism.
Cogito has been using AI-powered voice recognition and analysis to improve customer service interactions across many industries. The company’s foray into healthcare has come in the form of <a href="https://itunes.apple.com/us/app/cogito-companion/id1159040768?mt=8">Cogito Companion</a>, a mental health app that tracks a patient’s behavior.
<figure id="attachment_120083" style="width: 1000px" class="wp-caption aligncenter"><img class="wp-image-120083 size-full" src="https://wp-assets.futurism.com/2018/01/cogito_companion_in_hand_galaxyS5_websized.jpg" alt="" width="1000" height="667" /><figcaption class="wp-caption-text">Cotigo Companion. Image Credit: <a href="http://www.cogitocorp.com/cogito-companion-tracks-effect-of-marathon-bombing/cogito_companion_in_hand_galaxys5_websized/">Cotigo</a></figcaption></figure>
The app monitors a patient’s phone for both active and passive behavior signals, such as location data that could indicate a patient hasn’t left their home for several days or communication logs that indicate they haven’t texted or spoken on the phone to anyone for several weeks (the company claims the app only knows if a patient is using their phone to call or text — it doesn’t track who a user is calling or what’s being said). The patient’s care team can monitor the subsequent reports for signs that, in turn, may indicate changes to the patient’s overall mental health.
Cogito has teamed up with several healthcare systems throughout the country to test the app, which has found a particular niche in the veteran population. Veterans are at high risk <a href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4521393/">for social isolation</a> and may be reluctant to engage with the healthcare system, particularly mental health resources, often because of social stigma. “What we&#8217;ve found is that the app is acting as a way to build trust, in a way to drive engagement in healthcare more broadly,” Place said, adding that the app “effectively acts as a primer for behavioral change,” which seems to help veterans feel empowered and willing to engage with mental health services.
Here’s where the AI comes in: the app also uses machine learning algorithms to analyze “audio check ins” — voice recordings the patient makes (somewhat akin to an audio diary). The algorithms are designed to pick up on emotional cues, just as two humans talking would. “We&#8217;re able to build algorithms that match the patterns in how people are speaking, such as energy, intonation, the dynamism or flow in a conversation,” Place explained.
From there, humans train the algorithm to learn what “trustworthy” or “competence” sound like, to identify the voice of someone who is depressed, or the differences in the voice of a bipolar patient when they’re manic versus when they’re depressed. While the app provides real-time information for the patient to track their mood, the information also helps clinicians track their patient’s progress over time.
At Cogito, Place has seen the capacity of artificial intelligence to help us “understand the human aspects of conversations and the human aspects of mental health.”
Understanding, though, is just the first step. The ultimate goal is finding a treatment that works, and that’s where doctors currently shine in relation to mental health issues. But where do robots stand when it comes to things that are more hands-on?
<h2>Under The (Robotic) Knife</h2>
Over the last couple of decades, one of the most headline-making applications for AI in medicine has been the development of surgical robots.
In most cases to date, surgical robots (<a href="http://www.davincisurgery.com/da-vinci-surgery/da-vinci-surgical-system/">the da Vinci is the most well-known</a>) function as an extension of the human surgeon, who controls the device from a nearby console. One of the more ambitious procedures, <a href="http://www.montrealgazette.com/health/Robot+team+handles+Montreal+prostate+operation/3699124/story.html">claimed to be a world-first</a>, took place in Montreal in 2010. It was the first in-tandem performance of both a surgical robot as well as a robot anesthesiologist (cheekily named McSleepy); data gathered on the procedure reflects the impressive performance of these robotic doctors.
In 2015, more than a decade after the first surgical robots entered the operating room, MIT performed <a href="https://arxiv.org/ftp/arxiv/papers/1507/1507.03518.pdf">a retrospective analysis of FDA data</a> to assess the safety of robotic surgery. There were 144 patient deaths and 1,391 patient injuries reported during the period of study, which were mainly caused by technical difficulties or device malfunctions. The report noted that “despite a relatively high number of reports, the vast majority of procedures were successful and did not involve any problems.” But the number of events in more complex surgical areas (like cardiothoracic surgery) were “significantly higher” than in areas like gynecology and general surgery.
<figure id="attachment_120084" style="width: 1200px" class="wp-caption aligncenter"><img class="wp-image-120084 size-large" src="https://wp-assets.futurism.com/2018/01/robot-surgery-medical-procedures-1200x800.jpg" alt="" width="1200" height="800" /><figcaption class="wp-caption-text">Surgical robot. Image Credit: Getty Images</figcaption></figure>
The takeaway would seem to be that, while robotic surgery can perform well in some specialties, the more complex surgeries are best left to human surgeons — at least for now. But this could change quickly, and as surgical robots are able to operate more independently from human surgeons, it will become harder to know who to blame when something goes wrong.
<a href="https://qz.com/989137/when-a-robot-ai-doctor-misdiagnoses-you-whos-to-blame/">Can a patient sue a robot for malpractice?</a> As the technology is still relatively new, litigation in such cases constitutes something of a legal gray area. Traditionally, experts consider <a href="http://www.abpla.org/what-is-malpractice">medical malpractice</a> to be the result of negligence on the part of the physician or the violation of a defined standard of care. The concept of negligence, though, implies an awareness that AI inherently lacks, and while it’s conceivable that robots could be held to performance standards of some kind, those standards would need to exist.
So if not the robot, who, or what, takes the blame? Can a patient’s family hold the human surgeon overseeing the robot accountable? Or should the company that manufactured the robot shoulder the responsibility? The specific engineer who designed it? This is a question that, at present, has no clear answer — but it will need to be addressed sooner rather than later.
<h2>Building, Not Predicting, the Future</h2>
In the years to come, AI’s role in medicine will only grow: <a href="https://www.accenture.com/us-en/insight-artificial-intelligence-healthcare">In a report</a> prepared by Accenture Consulting, the market value of AI in medicine in 2014 was found to be $600 million. By 2021, that figure is projected to reach <a href="https://www.accenture.com/t20171214T021447Z__w__/us-en/_acnmedia/PDF-49/Accenture-Chart-The-AI-Health-Market-Explosive-Growth.pdf#zoom=50">$6.6 billion</a>.
The industry may be booming, but we shouldn’t integrate AI hurriedly or haphazardly. That’s in part because things that are logical to humans are not to machines. Take, for example, an AI trained to <a href="https://cs.stanford.edu/people/esteva/nature/">determine if skin lesions were potentially cancerous</a>. Dermatologists often use rulers to measure the lesions that they suspect to be cancerous. When the AI was trained on those biopsy images, it was more likely to say a lesion was cancerous if a ruler was present in the image, according to <a href="https://www.thedailybeast.com/why-doctors-arent-afraid-of-better-more-efficient-ai-diagnosing-cancer">The Daily Beast</a>.
Algorithms may also <a href="https://futurism.com/ai-bias-black-box/">i</a><a href="https://futurism.com/ai-bias-black-box/">nherit our biases</a>, in part because there’s <a href="http://medicalfuturist.com/no-precision-medicine-without-artificial-intelligence/">a lack of diversity</a> in the materials used to train AI. In medicine or not, the data the machines are trained on is largely determined by who is conducting the research and where it’s being done. <a href="https://www.aamc.org/download/432976/data/factsandfigures2010.pdf">White men still dominate</a> the fields of clinical and academic research, and they also <a href="https://www.ncbi.nlm.nih.gov/pubmed/25621972">make up most of the patients</a> who participate in clinical trials.
A tenet of medical decision-making is whether the benefits of a procedure or treatment outweigh the risks. When considering whether or not AI is ready to be on equal footing with a human surgeon in the operating room, a little risk-benefit and equality analysis will go a long way.
“I think if you build [the technology] with the right stakeholders at the table, and you invest the extra effort to be really inclusive in how you do that, then I think we can change the future,” Nundy, of Human Dx, said. “We&#8217;re trying to actually shape what the future holds.”
Though sometimes we fear that robots are leading the charge towards integrating AI in medicine, humans are the ones having these conversations and, ultimately, driving the change. We decide where AI should be applied and what’s best left done the old-fashioned way. Instead of trying to predict what a doctor’s visit will be like in 20 years, physicians can use AI as a tool to start building the future they want — the future that’s best for them and their patients — today.
Editor’s note: This article originally misspelled Shantanu Nundy&#8217;s name and title. They have been updated accordingly. We regret the error.

More AI in the exam room means doctors can spend more time actually talking to patients.

Your future doctor may not be human. This is the rise of AI in medicine.

AI is integrating into medicine, from the exam room to the operating room. But instead of taking over, AI will give doctors more time to work with patients.

Your Future Doctor May Not be Human. This Is the Rise of AI in Medicine.

From mental health apps to robot surgeons, artificial intelligence is already changing the practice of medicine. 

“Has anyone else’s AI decided that it has a soul?"

AI chatbot wants to be your new best friend

Artificial Intelligence

A chatbot called Replika is providing a real benefit for some people. But does intimacy with AI increase social alienation over the long-term? 

Your New Best Friend: AI Chatbot

What if AI bots could be the ones to have meaningful conversations with humans?

Counting calories. Reducing meal portions. Joining a gym. Eating small meals and lots of little snacks. Eliminating carbs, or sugar, or both. Starting a running regimen. 
We vow to adopt new habits like these, especially around this time of year, in an effort to slim down. The truth is, most of us need it —  <a href="https://www.niddk.nih.gov/health-information/health-statistics/overweight-obesity">66 percent of adults,</a> or 155 million Americans, are overweight or obese, according to the National Institutes of Health. Worldwide, the number of people considered overweight reached <a href="http://www.who.int/mediacentre/factsheets/fs311/en/">1.9 billion</a> in 2016, according to the World Health Organization. 
And so, in an effort to stop ourselves from gaining weight, we vow to eat better and exercise more. Sometimes this is, ostensibly, for our health — extra weight has been linked to <a href="https://www.cdc.gov/healthyweight/effects/index.html">a number of health problems</a>, including heart disease, Type 2 diabetes, increased risk of cancer or stroke, and overall increased mortality. For others, a diet might just be a way to look good. In any case, weight loss has become a <a href="http://time.com/magazine/us/4793878/june-5th-2017-vol-189-no-21-u-s/">$66 billion industry </a>in the United States alone.
But most of us, to our dismay and frustration, fail to keep the pounds off. While <a href="https://www.washingtonpost.com/national/health-science/the-big-number-45-million-americans-go-on-a-diet-each-year/2017/12/29/04089aec-ebdd-11e7-b698-91d4e35920a3_story.html?utm_term=.a86f3f78114c">45 million Americans</a> go on a diet each year, <a href="http://time.com/4793832/the-weight-loss-trap/">research has shown</a> that 80 percent of people who do manage to lose weight regain it — usually at a rate of about 2 to 4 pounds per year. 
Fortunately, scientists are on the case. Over the last few decades, researchers have worked to answer fundamental questions about human biology to understand why it’s difficult for people to lose weight, and why some people gain the weight back even when they’re diligently sticking to their diet and exercise program. 
Last year, the NIH dedicated $931 million to weight loss research, which has already improved our understanding of the multiple facets that influence individual outcomes. As a result of this research, scientists have gained a bit of extra wisdom First, it’s not always your fault if you can’t lose weight. Second, despite some tech-heavy, and somewhat invasive, weight loss solutions, nothing can quite replace diet and exercise.
<figure id="attachment_120407" style="width: 1200px" class="wp-caption alignnone"><img class="wp-image-120407 size-large" src="https://wp-assets.futurism.com/2018/01/bruno-nascimento-149663-1200x800.png" alt="" width="1200" height="800" /><figcaption class="wp-caption-text">Image Credit: <a href="https://unsplash.com/photos/PHIgYUGQPvU">Bruno Nascimento</a></figcaption></figure>
<h2>The Basics of Fat</h2>
From an evolutionary perspective, fat isn’t such a bad thing. It helps keep us warm, protects our internal organs, and stores unused energy from the food we eat in reserves that our body can drawn upon later.
In 2018, though, many of us have access to ample food sources that give us energy. And we don’t need to forage or hunt most of the day to get it. The result is an epidemic of obesity unprecedented in the history of humanity. 
Most of us understand the basic idea behind gaining and losing weight. If we consistently eat more calories than we use, we gain weight. Eat fewer, and we lose it. However, what’s actually happening inside our bodies when we gain weight  is more  complex than it may seem. Yes, adipose tissue — the technical name for fat —  stores energy and provides insulation. But it also stores a number of other things we may ingest (intentionally or otherwise) as we go about living: everything from <a href="https://www.popsci.com/when-you-lose-weight-your-fat-cells-release-more-than-just-fat#page-2">vitamins to environmental pollutants</a>. Fat also plays a key role in regulating hormones in the body, such as estrogen and insulin. 
We’re born with more or less the same number of fat cells that we’re going to have for life. So when we gain or lose weight,  it’s <a href="https://www.popsci.com/where-does-fat-go">the size of the cells</a> that changes  depending on how much excess energy from the food we eat they need to accommodate. 
Most recently, researchers learned that  fat cells are not all created equal. While the fat cells (“white fat”) we usually think of as building up around our hips and bellies are good at storing energy, so-called <a href="https://www.popsci.com/brown-fat-cells-good#page-2">brown fat cells</a> are much better at keeping us warm. It’s long been known to be present in chubby babies and is important to their survival (as they aren’t yet strong enough to yank the blankets up if they get cold in the night). Then, in 2009, researchers at the National Institute of Diabetes and Digestive and Kidney Diseases discovered that <a href="http://www.nejm.org/doi/full/10.1056/NEJMoa0810780#t=article">adults also have brown fat</a>. And that it’s <a href="http://www.nejm.org/doi/full/10.1056/NEJMoa0808718#t=article">a good fat</a> to have. 
One of the buzzwords related to <a href="https://www.mayoclinic.org/healthy-lifestyle/weight-loss/in-depth/metabolism/art-20046508">weight loss is metabolism,</a> which refers to the rate at which our body “burns” the energy we get from the food we eat, and which our body will store in fat cells if it’s not used. A person’s metabolic rate is influenced by a number of factors, such as age, sex, weight, and activity level. As such, one&#8217;s metabolic rate is not necessarily the same throughout their life. 
All this said, there’s nothing inherently bad about fat — in fact, we can’t survive without it. But we also can’t survive with too much of it. The same is true for other creatures, though their bodies work a bit differently than ours. Mice, according to <a href="http://www.pnas.org/content/115/2/427.full">one study</a>, have an <a href="http://blogs.discovermagazine.com/d-brief/2018/01/11/internal-scale-regulate-weight-loss-humans/#.WlvisJM-cWp">“internal scale”</a> that quells hunger cues when they get heavy. Humans have no such instinct, at least, not that we know of. Other animals, like shrews, <a href="https://massivesci.com/articles/shrews-shrink-humans-grow-bones-organs/?utm_content=bufferb3be0&amp;utm_medium=social&amp;utm_source=twitter.com&amp;utm_campaign=buffer">adjust their entire body size </a>— organs and skeleton and all — as their weight fluctuates in response to seasonal changes. 
Until we as a species find ourselves able to make changes like these, we’ll have to find other ways to hack our bodies to keep our weight in check  But given the quirks of our biology that keep the weight on, this might be easier said than done.
<figure id="attachment_120410" style="width: 1200px" class="wp-caption alignnone"><img class="wp-image-120410 size-large" src="https://wp-assets.futurism.com/2018/01/aaina-sharma-378569-1200x800.jpg" alt="" width="1200" height="800" /><figcaption class="wp-caption-text">Image Credit: <a href="https://unsplash.com/photos/C_spTFvd4GU">Aaina Sharma</a></figcaption></figure>
<h2>Interventions</h2>
The tried-and-true way to lose weight — good old diet and exercise — is, as most people can tell you, not much fun. It’s difficult to stick with, and progress is slow. So researchers have come up with a number of interventions that can speed up the process, or generate the same results with less effort.
Some techniques hijack the body’s natural mechanisms to get it to lose weight. People who are overweight may blame <a href="https://www.mayoclinic.org/healthy-lifestyle/weight-loss/in-depth/metabolism/art-20046508">a slow metabolism</a>, the rate at which the body uses the energy from food. But it’s really our muscles that burn those calories. What if they could do so more efficiently?  Researchers found that, when injected with a compound called morpholino, mice’s muscles burned more energy without working harder, according to <a href="http://www.cell.com/molecular-therapy-family/molecular-therapy/abstract/S1525-0016(16)30090-9">a 2015 study</a>. 
Other methods find ways to alter our relationship to hunger: either biochemically, by halting those signals from the brain, or addressing the more emotionally-charged experience of our discomfort associated with hunger. <a href="http://time.com/3836378/why-diets-fail/">Researchers have found </a>that people often give up on diets because that they struggle to quiet the feeling of hunger. But hunger, we’re realizing, isn’t just a biological signal that our bodies need nutrients — humans eat for emotional reasons and for social reasons. All of this makes it harder to lose weight. 
Perhaps drugs could quell our hunger. Last fall, a California-based pharma company called Amgen created an injection intended to do just that. The drug contains a protein called GDF5, bolstered with antibodies so the body wouldn’t break it down too quickly to be effective. <a href="http://stm.sciencemag.org/content/9/412/eaan8732">In experiments</a>, researchers found that monkeys that received the injection ate less, lost weight, and  were less likely to develop Type 2 diabetes. The drug had not yet been tested in humans, however, who could experience side effects that may make the drug unappealing (if it works at all). 
Amgen’s offering is the most recent in a long line of drugs developed over the past decade. To date, the others didn’t stand up to the scrutiny of clinical trials. <a href="http://www.nytimes.com/2003/04/01/business/obesity-drug-trial-produces-disappointing-weight-loss.html">Axokine</a> was one such drug; it failed its first clinical trial when two-thirds of the patients developed antibodies to it, meaning it stopped working. Another, called Rimonabant, made it to market in Europe only to be quickly pulled from the shelves when patients began experiencing <a href="https://www.ncbi.nlm.nih.gov/pubmed/19578688">major psychiatric side effects</a> such as depression and mood changes (it was never approved in the U.S.).
Other methods have proven effective, but with a significant caveat: They are much more invasive. <a href="https://asmbs.org/resources/story-of-obesity-surgery">Gastric bypass</a>, which surgically reduces the size of the stomach or shortens the intestinal tract, has been around for more than thirty years. <a href="https://www.newscientist.com/article/mg20327241-100-full-without-food-can-surgery-cure-obesity/">While it’s often successful</a>, it’s still a major surgery and comes with risks, such as infection or blood clots. 
So researchers have sought to create devices that are as effective as a bypass but are less invasive and come with less risk.  In 2015, the <a href="https://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/ucm430223.htm">FDA approved</a> the first <a href="https://www.newscientist.com/article/dn26814-first-anti-obesity-electronic-implant-wins-us-approval/">“anti-obesity” electronic implant</a>, which blocks signals along the vagus nerve — the way our stomachs tell our brains we’re hungry. 
Another device, called <a href="https://www.aspirebariatrics.com/es/healthcare-professionals/">AspireAssist</a> — essentially allowed patients to empty some of their stomach contents after a meal. The FDA approved the device in 2016, but it was met with controversy. The most widespread criticism was that such a device was more or less<a href="https://www.vox.com/2016/7/5/12067362/aspireassist-obesity-device-weight-loss"> physician-approved bulimia.</a> Its manufacturers clarified that the device was equipped with a safety mechanism that prevented it from being used more than <a href="https://www.aspirebariatrics.com/about-the-aspireassist/">three times a day for a six-week period</a>. 
One of the biggest challenges is, in truth, facing the emotional challenges weight loss presents. It is a journey fraught with frustration, trial and error, small successes, and feelings of monumental failure. The weight loss industry tends to entice folks who are looking for quick and easy fixes. In extremes, the diet and exercise regimens they tout often do lead to weight loss — often rapidly. While that may, at first, be reinforcement to keep someone motivated, what nutritionists know, and pharmaceutical executives perhaps do not care to acknowledged, is that keeping weight off requires more than dropping one kind of food. It requires a lifestyle change, often one that people are reluctant to make even if they know it would improve their health.
“If there would be something — the next great invention — [it would be something] that would keep people really engaged,” Becca Krukowski, an associate professor of preventive medicine at the University of Tennessee Health Science Center, told Futurism. “What I see so far is that there&#8217;s some sort of personal component, that nonspecific something, that keeps people going.”
<figure id="attachment_120420" style="width: 1200px" class="wp-caption aligncenter"><img class="wp-image-120420 size-large" src="https://wp-assets.futurism.com/2018/01/smart-watch-821565_1920-1200x800.jpg" alt="" width="1200" height="800" /><figcaption class="wp-caption-text">Fitness tracker. Image Credit: <a href="https://pixabay.com/en/smart-watch-apple-technology-style-821565/">Pixabay</a></figcaption></figure>
Data — on people’s activity, on what they eat — might do just that. Wearables like Fitbits or apps like Apple Health are still not available to everyone, but they are increasingly ubiquitous. But because this data is often tracked passively, it doesn’t always engage people (and produce the results they crave) as some might hope. 
People seem to need emotional support above all else. Krukowski notes that, even when people track their data, what they really want is to have others to talk to about it. The internet, then, might be one of the best weight loss tools out there, if it’s used to connect people, help them access information, and build networks of support. 
<h2>The Future of Weight Loss Comes In All Sizes</h2>
In an era in which everything is personalized and customizable, in which we seem on the cusp of altering treatment to match a person’s individual genome, it seems unlikely that the people of the future will lose weight with a one-size-fits-all approach. There’s no one diet, one exercise program, one device, or one surgery that will work for everyone. Often, those who succeed long-term have tried many techniques and strategies, and the one that eventually helps them keep the weight off might change over time. 
It seems far more likely that a more individualized approach will be truly transformative for those seeking to lose weight. 
For now, it seems, nothing can beat diet and exercise in terms of efficacy and safety. But technology could game-ify the process of losing weight, make it more appealing and less like drudgery. “In terms of technology, the big thing that I would love to see is ways to make it more fun,” Krukowski said. “I feel like if we could get, somehow, either through the process in these programs or help people recognize the changes they would enjoy in themselves, that this would be an easier, more fun, more sustainable process.” 
Directing people’s attention to the changes they’re making in their lives could help them stay motivated and engaged. But it could have a happy side effect, for which we all have a renewed hunger: joy.

Can anything do it better than diet and exercise?

Losing weight in the future will not be one size fits all

Health & Medicine

Obesity is a global epidemic, yet people struggle to lose weight. Drugs, surgeries, gadgets — could anything beat the drudgery of diet and exercise?

Losing Weight in the Future Will Not Be One Size Fits All

2.2 billion people worldwide are too heavy. How can we help them lose weight?

For a number of years in the 1980s, applicants to St George’s Hospital Medical School in London were selected with a high-tech method. A computer program, one of the first of its kind, took the first look at their résumés, carrying out the initial selection of about 2,000 candidates every year. The program analyzed the admissions records to learn the characteristics of successful applications, and was adjusted until its decisions matched those of the admissions team.
But the program had learned to look for more than good grades and signs of academic prowess. Four years after the program was implemented, two doctors at the hospital discovered the program tended to reject female applicants and those with non-European-sounding names, regardless of their academic merit. As many as 60 applicants each year could have been refused an interview simply because of their gender or race, the doctors found. The program had incorporated the gender and racial biases in the data used to train it — it was essentially taught that women and foreigners were not doctor material.
Three decades later, we are facing a similar problem, but programs with internalized biases are now more widespread and make decisions with even higher stakes. Artificial intelligence algorithms powered by machine learning are now used everywhere, from government institutions to healthcare, aiding decision-making by providing predictions based on historic data. As they learn patterns in the data they also absorb biases in it, perpetuating them. Google, for example, <a href="https://www.theguardian.com/technology/2015/jul/08/women-less-likely-ads-high-paid-jobs-google-study">showed more ads for lower paying-jobs</a> to women than to men, Amazon’s same-day delivery <a href="https://www.bloomberg.com/graphics/2016-amazon-same-day/">bypassed black neighborhoods</a>, and the software on several types of digital cameras <a href="https://www.inverse.com/article/39483-apple-refund-china">struggled to recognize</a> the faces of non-white users. In one of the most striking examples, an algorithm called COMPAS, used by law enforcement agencies across multiple states to assess a defendant’s risk of reoffending, was found to falsely flag black individuals almost twice as often as whites, according to a <a href="https://www.propublica.org/article/machine-bias-risk-assessments-in-criminal-sentencing">ProPublica investigation</a>.
It is difficult to figure out if an algorithm is biased or fair, even for computer experts. One reason is that the details behind an algorithm’s creation are often considered proprietary information, so are closely guarded by their owners. In more advanced cases, the algorithms are so complex that even their creators don’t know exactly how they work. This is AI’s so-called black box problem — our inability to see the inside of an algorithm and understand how it arrives at a decision. If it’s left unsolved, it can devastate our societies by ensuring that historical discrimination, which many have worked hard to leave behind, is hard-coded into our future.
These worries, whispered within smaller computer science communities for a few years, are now gaining a central spot in the field. In the past two years, the field has seen a huge spike in the number of papers on fairness in AI. With that awareness, there’s also a <a href="https://futurism.com/are-engineers-responsible-for-the-consequences-of-their-algorithms/">growing sense of responsibility</a>. “Are there some things we just shouldn’t build?” Kate Crawford, Microsoft researcher and cofounder of the AI Now Institute at NYU, <a href="https://www.wired.com/story/artificial-intelligence-seeks-an-ethical-conscience">asked in a recent speech</a>.
“Machine learning has finally made it to prime time. Now we are trying to use it for hundreds of different purposes in the real world,” Rich Caruana, a senior researcher at Microsoft, told Futurism. “It is possible for people to deploy harmful algorithms that can add up to have quite a big impact on society on the long run… Now it feels like all of a sudden everyone is aware that this is an important new chapter in our field.”
<figure id="attachment_119049" style="width: 1200px" class="wp-caption aligncenter"><img class="size-large wp-image-119049" src="https://wp-assets.futurism.com/2018/01/Computer-Art-1200x900.png" alt="" width="1200" height="900" /><figcaption class="wp-caption-text">Photo collage by Tag Hartman-Simkins; Getty Images</figcaption></figure>
<h2>The Algorithm Made Me Do It</h2>
We have been using algorithms for a long time, but the black-box problem is somewhat unprecedented. Earlier algorithms were simpler and transparent. Many of them are still in use — for example, for credit scoring by FICO. For each new use, regulation has followed.
“People have been using algorithms for credit scoring for decades, but in those areas there have been pretty strong regulations that grew up alongside the use of these prediction algorithms,” Caruana said. These regulations ensure that prediction algorithms provide an explanation for each score: You were denied because your loan balances are too high, or because your income is too low.
The regulations that prevent credit scoring companies from using inscrutable algorithms are absent in other areas, like the legal system and advertising. You may not know why you were denied a loan or didn’t get a job, because no one forces the owner of the algorithm to explain how it works. “But we know that because [the algorithms] are trained on the real world data, they have to be biased — because the real world is biased,” Caruana said.
Consider language, for example — one of the most obvious sources of bias. When algorithms learn from written text, they pick up some associations between words that appear together more often. They learn, for example, that “<a href="https://www.technologyreview.com/s/602025/how-vector-space-mathematics-reveals-the-hidden-sexism-in-language/">man is to computer programmer as woman is to homemaker</a>.” When this algorithm is tasked to find the right résumé for a programming job, it will be more likely to pick male applicants than females.
Problems such as this one are fairly easy to fix, but many companies simply don’t go to the trouble of doing so. Instead, they hide such inconsistencies behind the shield of propriety information. Without access to details of an algorithm, in many cases even experts can’t determine whether or not bias exists.
Because these algorithms are secretive and outside regulators’ jurisdiction, it’s almost impossible for citizens to bring a case against the algorithms’ creators. Those who have tried haven’t gotten very far. In 2016, Wisconsin’s highest court denied a man’s request to review the inner working of COMPAS. The man, Eric L. Loomis, was sentenced to six years in prison in part because COMPAS deemed him to be “high-risk.” Loomis says his right to due process is violated by the judge’s reliance on an opaque algorithm. A final bid to take the case to <a href="http://www.scotusblog.com/case-files/cases/loomis-v-wisconsin/">the U.S. Supreme Court</a> failed in June 2017. Separately, two law professors <a href="https://www.wired.com/story/when-government-rules-by-software-citizens-are-left-in-the-dark/">spent a year</a> probing states to understand how they use scoring in their criminal justice systems. The only thing their investigation confirmed is that this information is well hidden behind nondisclosure agreements.
But secretive companies may not enjoy their freedom indefinitely. By March, the European Union will enact laws that will require companies to be able to explain to inquiring customers how their <a href="https://arxiv.org/abs/1606.08813">algorithms work and make decisions</a>.
The U.S. has no such legislation in the works. But there are signs that the tide might be turning towards more regulatory oversight. In December 2017, the New York City Council <a href="https://arstechnica.com/tech-policy/2017/12/new-york-city-moves-to-create-accountability-for-algorithms/">passed a bill</a> to establish a task force that will study algorithms used by city agencies and explore ways to make their decision-making processes understandable to the public.
<figure id="attachment_119056" style="width: 1200px" class="wp-caption aligncenter"><img class="size-large wp-image-119056" src="https://wp-assets.futurism.com/2018/01/Black-Computer-Art-1200x900.png" alt="" width="1200" height="900" /><figcaption class="wp-caption-text">Photo collage by Tag Hartman-Simkins; Getty Images</figcaption></figure>
<h2>Black Box Forensics</h2>
No matter if regulators get involved, a cultural shift in how algorithms are developed and deployed may reduce the pervasiveness of biased algorithms. As more companies and programmers pledge to make their algorithms transparent and explainable, some hope that companies that don’t will be called out and lose good standing in the public’s opinion.
Recently, growing computational power has made it possible to create algorithms that are both accurate and explainable — a technical challenge that developers have historically struggled to overcome. <a href="https://arxiv.org/pdf/1704.01701.pdf">Recent studies</a> show it’s possible to make explainable models that predict whether criminals are likely to reoffend that are just as accurate as the black-box versions such as COMPAS.
“The research is there — we know how to create models that are not black boxes,” Cynthia Rudin, an associate professor of computer science and electrical and computer engineering at Duke University, told Futurism. “But there is some difficulty getting people to notice this work. If government agencies stop paying for black box models that would help. If judges refuse to use black box models for sentencing, that would help too.”
Others are working to come up with ways to test the fairness of algorithms by creating a system of checks and balances before an algorithm is released to the world, the way a new drug has to pass clinical trials.
“What’s happening right now is that models are being made too quickly and deployed. There are no proper checks throughout the entire process or requirements to test it in the real world for a trial period,” Sarah Tan, a doctoral student in statistics at Cornell University, told Futurism.
Ideally, developers should clean known biases — such as those for gender, age, and race — from their training data, and run internal simulations to check their algorithms for other problems that might have still found a way in.
In the meantime, before getting to the point where all algorithms are rigorously tested before release, there are ways to find which ones might suffer from bias.
<a href="https://arxiv.org/abs/1710.06169">In a recent paper</a>, Tan, Caruana, and their colleagues described a new way to understand what might be happening under the hood of black box algorithms. The team created a model that mimics a black box algorithm like COMPAS by training on the recidivism risk score that COMPAS predicted. They also created another model that they trained on real-world outcome data that show whether or not that predicted recidivism actually happened. Comparing the two models allowed the researchers to assess the accuracy of the predicted score without dissecting the algorithm. Differences in the outcomes of the two models can reveal what variables, such as race or age, may have been given more importance in one model or another. Their findings were in line with what ProPublica and other researchers have found — that COMPAS is biased against black individuals.
There could be big rewards for fixing such biases. Algorithms, if built properly, have the power to erase longstanding biases in criminal justice, policing, and many other areas of society.
“If we do work on this and manage to reduce the bias, we can have a virtuous feedback loop, where in fact algorithms slowly help us become less biased as a society,” Caruana said.

"The research is there — we know how to create models that are not black boxes."

We need to open the AI black box before it's too late

Many systems use sophisticated algorithms, but even their designers don't know how they work. Those algorithms may perpetuate undesirable biases.

We Need to Open Algorithms' Black Box Before It's Too Late

If we don't, the biases of our past could dictate our future.

We Need to Open the AI Black Box Before It’s Too Late

Political scholars could take a number of lessons from the 2016 U.S. presidential election. Perhaps one of the most obvious is that our voting process is not immune to meddling. Countries <a href="http://www.nytimes.com/2007/12/31/world/africa/31kenya.html">from Kenya</a> <a href="https://www.nytimes.com/2017/11/30/world/americas/honduras-vote-political-crisis.html?_r=0">to Honduras</a> have had to contend with contested elections in the past, but the <a href="https://www.cfr.org/backgrounder/russia-trump-and-2016-us-election">Russian interference</a> in the 2016 election is arguably without precedent for the U.S. With the 2018 midterms fast approaching, <a href="https://www.newyorker.com/news/news-desk/why-the-2018-midterms-are-so-vulnerable-to-hackers">some experts fear</a> that <a href="https://www.buzzfeed.com/kevincollier/the-democrats-did-well-on-tuesday-but-that-doesnt-mean-our?utm_term=.je9Kr3dAR#.ahEkr5ADV">the cracks in</a> our political process through which meddlers infiltrated have not yet been sealed.
Over the past few years, a growing chorus has asked: Why not take elections digital? If voters could weigh in via <a href="https://www.newscientist.com/article/mg23531424-500-bitcoin-tech-to-put-political-power-in-the-hands-of-voters/">a hyper-secure app</a>, wouldn&#8217;t it make voting more convenient, which could allow more people to weigh in more often? A voting app could ensure that we have a <a href="https://motherboard.vice.com/en_us/article/59dnbb/colorado-political-candidate-promises-to-give-his-seat-to-an-app">more direct democracy</a>, or at least a &#8220;liquid democracy,&#8221; in which individual voters have more contact with their representatives.
But the equation might not be as simple as it seems. Some experts fear that sophisticated voting technology might exclude the less tech-savvy among us, or those who can&#8217;t afford a smartphone or computer. Others point out that taking voting totally digital might, in fact, make our elections more vulnerable to hacking; for some, only <a href="https://www.theatlantic.com/technology/archive/2017/05/the-voting-technology-we-really-need-paper/524820/">paper ballots are hack-proof</a>.
Futurism touched base with the experts to see if digital elections would be better for American citizens, or for our democracy.
First, it&#8217;s worth discussing what kind of political shift, exactly, tech would best be used for. Areeq Chowdhury, the <a href="https://webrootsdemocracy.org/about/">chief executive of WebRoots Democracy</a>, &#8220;a voluntary, youth-led think tank focused on the intersection of technology and democratic participation&#8221; based in the United Kingdom:
<blockquote>Liquid democracy is an interesting idea, and the idea of giving everyone the power to vote on everything in the palm of their hands is a powerful one. In theory, it could lead to a system where everyone&#8217;s views are meaningfully listened to and acted upon, with the will of the majority truly represented.
It is not the same, however, as direct democracy. With liquid democracy, the voter should be able to assign their votes on certain issues to certain representatives, and be able to take them away. This would mean that you, as a voter, could assign your votes on an issue to someone who you deem sufficiently knowledgeable, leading to educated votes. However, this could also leave it open to being gamed by lobby groups. For it to truly work as a system, political education must be made mandatory in schools, and there would need to be serious investment in democratic engagement to ensure that enough people use the app. The first question that needs to be answered is, &#8220;do people actually want this?&#8221;</blockquote>
So could tech help? Steve Ressler, the <a href="https://www.govloop.com/">founder of GovLoop.com</a>, a social network to connect government officials:
<blockquote>Governments and elected officials have always wanted to get constituent feedback on issues. However, there have always been barriers — it&#8217;s hard for folks to drive downtown to attend a townhall meeting in person, find transportation to get to a voting booth, or spend the time to write a letter or call an official. The ubiquitous nature of smartphones has the potential to dramatically increase the number of people involved in the democratic process by making it easier.
It does this in a couple ways — you can provide feedback regardless of your location and, by making it real-time with short feedback loops, it decreases the time commitment. While I love idea of using smartphones to get more feedback, I believe we elect officials to not just do a straw poll on every issue — tough issues require bringing together citizen input with thoughtful judgment to make decisions.</blockquote>
Rouven Brües, a project manager at the <a href="https://liqd.net/en/about/">Liquid Democracy Association</a>, a nonprofit dedicated to digital civil action:
<blockquote>Digital media technologies can be transformative because they enable us to rethink and remodel our democracies. New ideas that are technically feasible today, such as the delegation of votes through an app, enable us to envision the democracies to come, even though their implementation might still be facing practical difficulties.
In practice, changing democratic processes with technology is not so easy.Voting is the last step in a democratic decision-making processes and arguably one of the smaller ones if you look at the whole process: setting an agenda, forming arguments and positions in a discourse, finding compromises between conflicting positions, and merging these in concrete (law) proposals.
Through our work, we have come to understand that neither apps nor software will completely replace the democratic processes in place, since they can only model parts of the process.
Using an app to receive delegations from citizens and voters is a smart way to include them in the decision-making process. However, this runs the risk that the politician doing so renounces any possibility to negotiate compromises. So how useful an app actually is to collect delegations and votes depends on the decisions that have to be made.
In any case, we strongly believe that all democratic- and civic-tech tools have to be open (as in open-source) if they want to have a sustainable impact and benefit for society and if they want to be transparent and trustworthy.</blockquote>

Over the past few years, a growing chorus has asked: Why shouldn't elections be digital?

Could digital voting create a society that is truly governed by the people?

Checkbox

The 2016 presidential election showed that our elections are even more vulnerable to hacking. Some say that voting digitally might fix that.

/Feature

Future Society

How Earth’s Most Extreme Environs Are Key to Humanity’s Future

Future Society

In 1918 We Faced the Flu Pandemic. Today, We’re Still Fighting the War.

Future Society

Is Organic Really Better? 4 Food Myths Debunked By Science

Sci-Fi Visions

Advanced Reproductive Technology is Here. But Who Decides Who Gets Access?

Sci-Fi Visions

When It Comes To Natural Disasters, Technology Has An Unavoidable Dark Side

Sci-Fi Visions

Your Future Doctor May Not be Human. This Is the Rise of AI in Medicine.

Artificial Intelligence

Your New Best Friend: AI Chatbot

Health & Medicine

Losing Weight in the Future Will Not Be One Size Fits All

Future Society

We Need to Open the AI Black Box Before It’s Too Late

Future Society

Could Digital Voting Create a Society That Is Truly Governed by the People?

Load More