The GOP loves Big Government in health care — if it’s blocking abortion or trans care.
As the Republican Party begins to define itself for the 2024 presidential campaign, its philosophy on health care is coming into focus: The government is not here to help you. But it should be allowed to get in your way.
Wednesday night’s debate, which featured eight of the leading not-Donald-Trump candidates for the Republican nomination, spent little time on health care except for an extended exchange on abortion, covered in depth by Vox’s Rachel Cohen.
But if you piece together the abortion debate with other moments in which health care briefly took the spotlight, the party’s position on federal power in the medical sphere begins to take shape. On abortion, on health care for transgender people, even on mental health care, the candidates were comfortable flexing governmental authority to dictate the terms of medical treatment.
But when it comes to using that same authority to protect people during a global pandemic or providing health coverage to people with low incomes, they don’t want the government getting involved.
How the GOP candidates want to use state power to restrict health care
Abortion — which Fox moderator Martha MacCallum cast as a “losing” political issue for Republicans ever since the Supreme Court overturned Roe v. Wade — led to a contentious exchange in which former South Carolina Gov. Nikki Haley urged her fellow candidates to be square with GOP voters: It will be difficult to pass a national abortion ban given the political realities of the US Senate.
Haley was largely an outlier, though North Dakota Gov. Doug Burgum also argued the issue would be rightfully left to the states under the 10th Amendment. Former Vice President Mike Pence and US Sen. Tim Scott of South Carolina made impassioned pleas for a 15-week national abortion ban. Scott argued that the decision on abortion policy should be taken away from the states — the elected representatives who are closest to the pregnant people affected and who conservatives have traditionally argued should make these decisions — because they cannot be trusted to set these standards on their own.
“We cannot let states like California, New York, and Illinois have abortions on-demand up until the day of birth. That is immoral. It is unethical. It is wrong,” Scott said to applause from the audience of Republican primary voters. (For the record, fact-checkers have repeatedly corrected Republicans for making similar claims in the past and only 1 percent of abortions occur after 21 weeks of gestation.)
“We must have a president of the United States who will advocate and fight for at the minimum a 15-week limit,” Scott continued.
The entire premise of the anti-abortion movement is to involve the state in a life-altering medical decision otherwise handled privately between a doctor and a patient. In spite of the evidence of the economic and health consequences of restricting abortion and in spite of the political losses the GOP has incurred since the Supreme Court negated the right to privacy enshrined in Roe, Scott and others reaffirmed a commitment to keeping the state in that position. The crowd roared.
Transgender rights and the myriad conservative laws passed in the past few years to restrict access to gender-affirming care were referenced only obliquely in the debate but carried the same message. Florida Gov. Ron DeSantis touted his record of “eliminat[ing] gender ideology from our K-12 schools.” North Dakota Gov. Doug Burgum defended his signing of a bill that banned certain students from participating in school sports. In his closing remarks, Scott pledged that “if God made you a man, you play sports against men.”
As Katelyn Burns wrote for Vox in 2021, medicine itself has undergone a transformation on the subject of gender diversity and gender-affirming care. Historically, people who identified as a different gender than they were assigned at birth were viewed as having a disorder that needed to be fixed. More recently, most doctors have come to believe that such patients should be handled more humanely and affirmatively; permitting them to make a social gender transition (changing their name and pronouns, using a different bathroom, etc.) is treated as a given. Medical interventions for young people are to be taken on a case-by-case basis.
But as the GOP candidates’ comments at the debate made clear, they have little interest in permitting even a social transition for transgender youth. And in statehouses across the US, the party has sought to restrict this kind of medical care as much as possible.
“Trans advocates have pointed out that these bills fit comfortably within the larger GOP plan to seize minority power in an effort to force their preferred gender dynamics,” Burns wrote. The state, not the individual nor their doctor, is dictating the terms of health care.
In one of the most striking tangents of the night, businessman Vivek Ramaswamy even advocated for reopening “mental health institutions” that have closed over the decades as the country sought to cut costs (starting in the Reagan administration) and tried — but has largely failed — to invest in more humane home- and community-based services. He framed putting more people with mental health problems into those institutions as one of the most important ways to address crime.
“Just over the same period that we have closed mental health institutions, we have seen a spike in violent crime. Do we have the spine to bring them back?” he said. “I think we should. As president, I will.”
I have interviewed people who were subjected to the cruel treatment that was often commonplace in the mental health asylums of old. The new mental health paradigm is focused much more on treating people where they are. While the US has struggled to erect a sufficient mental health system given the scale of the country’s needs, few people who work in the field would argue that a return to the institutions of that era is the recipe for fixing our mental health crisis.
But Ramaswamy, the biggest non-Trump sensation of the GOP primary so far, does. He even implicitly allowed that “drugging up people in those psychiatric institutions with Zoloft and Seroquel” would be part of the program (but not the only part; he also wants a return to the patriotism, faith, and family that he believes constitutes the American character).
It demonstrated an astonishing willingness to use the power of the state to interfere with the ability of certain people, particularly the historically marginalized, to achieve health and well-being — a consistent theme of the night.
Republicans still don’t want the government to help people get health care
What affirmative actions the candidates might take to improve people’s health was generally neglected in the first debate of the 2024 campaign. No candidate laid out a health care reform agenda, as Republicans are still in the health policy wilderness after their failure to repeal and replace the Affordable Care Act.
But a few select moments revealed that GOP candidates, while perfectly comfortable interfering with certain medical decisions, remain opposed to using that same government authority to provide assistance to people who need access to health care or to protect people whose health may be at risk in a public health emergency. They will still cling to the veneer of small government and individual liberty in certain contexts.
The most obvious such moment is when DeSantis took a line of questioning about the economy and touted his willingness to shirk the public health consensus on Covid-19.
“Why are we in this mess? A major reason is because how this federal government handled Covid-19 by locking down this economy. It was a mistake. It should have never happened,” he said. “In Florida, we led the country out of lockdown. We kept our state free and open. And I can tell you this: As your president, I will never let the deep state bureaucrats lock you down.”
DeSantis’s pandemic record is complicated and I covered it in depth earlier this year. In brief, yes, the state enjoyed an earlier economic recovery as pandemic restrictions were sunset, but it also had a middling performance in terms of health outcomes. DeSantis also pushed beyond ending lockdowns to embrace vaccine skepticism, including empowering a vaccine skeptic as his state surgeon general. Experts who saw the administration’s response up close believe DeSantis’s about-face was detrimental to the health of his state’s citizens — and believe it was more politically calculated than anything else.
But a governor who signed a six-week abortion ban and a ban on gender-affirming care for minors while placing new requirements on such care for adults, putting the state firmly in the middle of individual medical decisions, was proudly unwilling to use the same power to try to tame the spread of a virus that ultimately killed more than 1 million Americans including 80,000 Floridians.
One other brief moment, on another subject, reinforced the same ideas. As Haley tried to take Republicans to task for their role in increasing government spending in recent years, she mentioned the increase in Medicaid enrollment that followed the institution of an emergency policy that allowed people to stay on the program’s rolls.
“Our Republicans did this to us too. When they passed that $2.2 trillion Covid stimulus bill, they left us with 90 million people on Medicaid, 42 million people on food stamps” she said. “No one has told you how to fix it. I’ll tell you how to fix it. They need to stop the spending.”
Her meaning was clear: It was a problem that so many people enjoyed government benefits and didn’t need to worry about their health coverage, even during the pandemic. As it happens, the government is unwinding that emergency policy — and millions of Americans are losing their health insurance as a result.
Nobody asked the Republican candidates about what they might do to address the dramatic increase in the uninsured rate that will result. As Wednesday night’s debate made clear, they don’t think of that as the government’s job.
But if you are a certain kind of person who needs a certain kind of treatment, they are ready and willing to put the state between you and your health care.
Why breaks, water, and shade for a sea of US workers could trigger the next big labor battle.
It’s been a summer of record-breaking heat waves. In July, a heat dome smothered much of the South, while Phoenix sweltered under a 31-day streak of temperatures above 110 degrees.
Another wave of searing temperatures is currently sweeping the central US, with the heat index — a metric that combines relative humidity with air temperature to gauge what the temperature actually feels like — creeping above 120 degrees in many cities.
The Midwest may be used to acute winters, but not such searing summers: Some buildings have no air conditioning, and infrastructure such as asphalt roads may not have been built with extreme heat in mind.
As the mercury rises, many US jobs become magnitudes more difficult and dangerous — and not just on construction sites and farm fields, where the majority of heat-related deaths happen. Many workers toiling in places without sufficient cooling or ventilation face high risk, too, including those in factories and warehouses, restaurant employees, and delivery drivers and gig workers exposed to dangerous temperatures as they lug packages out in the sun. Even people who work in indoor environments with cooling mechanisms are at risk, given that faulty or weak HVAC systems are common — as teachers across the country are experiencing. The list of jobs where heat can be a health hazard is endless.
From 2017 to 2022, 121 workers died from heat, according to the federal Occupational Safety and Health Administration (OSHA), though the agency acknowledges that this is likely an undercount. The Bureau of Labor Statistics has recorded another almost 34,000 work-related heat injuries and illnesses that required time away from the job between 2011 and 2020. This summer, Eugene Gates, a 66-year-old letter carrier in Dallas, died on the job on a June day when the temperature reached 97 degrees, and a month later, Dario Mendoza, a 26-year-old farm worker, died after working in the fields amid the Phoenix area’s historic heat wave.
“Many people are saying that this summer feels worse than the summer before,” says Mayra Reiter, project director of occupational safety and health for Farmworker Justice, a worker protection nonprofit. “And it’s making people work in the fields that much harder — people are feeling unsafe.”
The recent spate of heat injuries and deaths among workers is just a preview of what’s to come as climate change continues to send summer temperatures soaring in many US cities. By 2050, some will see their average summer highs rise by as much as 6 degrees. According to a report from the Environmental Protection Agency, wildfire smoke — and even increased risk of diseases from insects such as ticks and mosquitoes, as the warmer weather expands their habitats — will add extra strain to outdoor workers. Indoor workers won’t be exempt from peril either, as buildings with poor insulation and ventilation become ovens during heat waves.
The far-reaching ripple effects of climate change will impact where we can work, when we can work, and how long we can work. Yet there’s no federal regulation setting a threshold for when heat becomes too dangerous for employees to work, or even what employers are required to provide their workers in those conditions — such as water, more frequent breaks, shade, or air conditioning.
For many workers, there’s simply no relief in sight.
How heat harms your health — and how it can make work deadly
There’s no single temperature at which physical exertion becomes risky for everyone. But researchers have found that at a wet bulb globe temperature — a reading that includes humidity, wind speed, whether there’s direct sun exposure, and more — of about 95 degrees Fahrenheit, even the most fit people will overheat. Exposure to that level of heat for as little as six hours could result in death.
Heat begins impacting how efficiently we can work, however, long before the temperature reaches 95 degrees. It varies by region, but in the US the optimal temperature for physical outdoor work is 57.7 degrees. Above that point, productivity starts to decline. It goes to show how quickly heat impacts our bodies.
As a worker overheats, they start feeling weak and dizzy. Their speech may slur. These first signs can lead to heat stroke, as well as rhabdomyolysis — in which muscle tissue breaks down — or even a heart attack. There are more long-term effects, too. A person’s heat tolerance can shrink after they suffer a heat stroke, and scientists are also looking into a possible link between working in heat and kidney disease that’s been observed among agricultural workers.
/cdn.vox-cdn.com/uploads/chorus_asset/file/24871110/GettyImages_1588298884.jpg)
Mario Tama/Getty Images
/cdn.vox-cdn.com/uploads/chorus_asset/file/24871126/GettyImages_1553720321.jpg)
Eddie Seal/Bloomberg via Getty Images
Temperatures also don’t have to be very high for strenuous work to become deadly. The lack of heat acclimatization all too frequently kills workers; the majority of workers who die from heat do so in the first few days at work. “A lot of workers will actually end up in heatstroke during the first week on the job,” says Brenda Jacklitsch, a health scientist at the National Institute for Occupational Safety and Health (NIOSH).
Extreme heat is also affecting productivity. According to a 2022 study by The Lancet, which is tracking the relationship between climate change and public health, about 470 billion hours of labor were lost in 2021 due to extreme heat. The US alone lost 2.5 billion hours, mostly in the construction, manufacturing, service, and agriculture sectors. “Workers slow down,” says Shouro Dasgupta, an environmental economist and a co-author of The Lancet report. “They either have to work fewer hours or they put less effort during the hours they can work.” Limited work hours will likely ripple down to all parts of the economy; it could affect the pace and availability of consumer goods and services, slow down delivery orders, and delay air travel.
Across the country, airport ground services workers, whether it’s cabin cleaners, baggage handlers, or others who have to be out on the tarmac, have reported facing vicious heat on the job. Rashele Bates, a 26-year-old who cleans plane cabins at the Charlotte, North Carolina airport and who is an active member of the Service Employees International Union (SEIU), told Vox she felt sick and nauseous on the job one day in late July after she and a coworker had been cleaning for about three hours without a proper break. Bates vomited, and her coworker passed out and went to the hospital.
“I’m sitting there drowning in sweat,” Bates recalls. The air conditioning isn’t always turned on in the plane while they work, Bates says, nor is there any air conditioning on the jet bridge connecting the plane to the terminal, where she and her coworker fell ill. Bates and a coalition of other airport workers sent a petition to the airline contractor she works for, Jetstream Ground Services, demanding water bottles and functioning air conditioning in each break room. The SEIU is also supporting legislation to improve labor conditions at airports. Jetstream Ground Services did not respond to a request for comment.
There’s no great mystery on how to address heat illness in the workplace. NIOSH, which operates under the Centers for Disease Control, published its first recommended standards on workplace heat safety back in 1972, and the recommendations haven’t changed much since.
“A lot of the work that’s done is not necessarily rocket science,” said Andrew Levinson, director of standards and guidance at OSHA. “It’s making sure people have water, rest, shade, and that they can ease into work that first week to make sure that they’re adequately acclimatized.”
Why don’t we have workplace heat protection laws?
About half of states could make their own heat standards for the workplace, but only California, Oregon, Washington, Colorado, and Minnesota have done so. In California, an 80-degree heat index triggers employer requirements for regular rest, shade, and water for outdoor workers. Oregon’s heat safety rules, which Reiter of Farmworker Justice said are the country’s strongest, also kick in at 80 degrees, and apply to indoor workers, too.
Texas, meanwhile, is going in the other direction. Governor Greg Abbott signed a bill in June that bans local ordinances going beyond state-level workplace safety rules, after Austin and Dallas had passed ordinances in 2010 and 2015, respectively, mandating more frequent water breaks for construction workers. On a state level, employers are not required to give any breaks to employees. Texas is the most lethal state for heat-related deaths at work — according to the Bureau of Labor Statistics, which also keeps track of heat deaths on the job, 42 people have died there since 2011.
A federal OSHA heat standard is still years away, and not because the agency isn’t aware of the problem. It began working on a proposed rule in 2021, but creating a new standard is notoriously hard because Congress requires the agency to follow a long, labyrinthine process. For example, it must do a cost-benefit analysis, calculating the cost not just to the agency but for all affected industries as well, and go through multiple rounds of public comment.
“That doesn’t mean that the agency is not out there, shouting from the rooftops on the importance of this,” says Levinson, who added that a lot of workers are calling on OSHA to develop heat standards quickly.
The agency also has to contend with political headwinds. The conservative stance on workplace safety rules is broadly that they shouldn’t exist. “Republican administrations don’t do OSHA rules,” says Jordan Barab, a former deputy assistant secretary of labor at the agency under the Obama administration who now publishes a workplace safety newsletter.
In the absence of a heat standard, OSHA has been deploying the “General Duty” clause — an all-purpose rule that employers must provide a workplace free from “recognized hazards.” There are four steps to the clause, though, and OSHA has to build its case for proving each one. “These cases are very legally vulnerable,” says Barab.
Reiter of Farmworker Justice says that OSHA could issue an emergency heat standard while it works on issuing a more permanent standard, but rules like these have been successfully challenged in court. Instead, Reiter’s group, along with a number of unions, are backing a federal bill, recently introduced in the Senate, that would give OSHA a year to issue an interim standard and make it harder to challenge. The bill has won support from a number of Democratic lawmakers.
The Coalition of Immokalee Workers (CIW), which advocates for farm workers in Florida and beyond, isn’t waiting on the federal government. In 2011, the group launched the Fair Food Program, which sets stronger labor standards, including heat standards, on farms. So far more than 20 crop growers, along with Burger King, McDonald’s, Whole Foods, Trader Joe’s, and other large food companies, have signed on.
Gerardo Reyes Chavez, an organizer with CIW and a former farmworker himself, says the difference between conditions on farms signed up with the program and those that aren’t are like night and day. “Before, you didn’t have any way to protect yourself, you were basically down to your luck. But the way I see it, human rights shouldn’t depend on luck.”
/cdn.vox-cdn.com/uploads/chorus_asset/file/24871151/GettyImages_1542215814.jpg)
Giorgio Viera/AFP via Getty Images
Other unions — including the United Steelworkers, the Teamsters, the SEIU, the ROC United (which advocates for restaurant workers), National Nurses United, and more — are taking up regulations directly with major employers: Recently, UPS drivers unionized with the Teamsters won a contract that, among other changes, requires new company vehicles to have air conditioning. Many existing delivery cars will also be retrofitted with AC, fans, and vents.
Some employers and industry representatives are less than thrilled about a looming OSHA heat standard. The American Farm Bureau Federation, a trade association representing farm producers — an industry with some of the highest numbers of heat-related worker injuries and deaths — has argued that the General Duty clause already protected workers and questioned whether new regulations could be implemented “without imposing new, onerous burdens on farmers and ranchers that will lead to economic losses.” It then implored the agency to focus on “the responsibility on the individual employee and the personal health choices that are made outside of the workplace,” a sentiment that has been echoed by other pro-business groups. The organization declined an interview request for this story.
It goes to show how heat isn’t necessarily what’s making workers sick, Barab says. “It’s really more employers’ refusal to implement measures to protect workers. That’s killing workers.”
It takes forever to get drugs on the market. AI could help speed up the process.
In the 1968 novel and film 2001: A Space Odyssey, the artificial intelligence system Hal (short for HAL 9000) kills its spaceship’s astronauts.
In reality, the nickname Hal refers to a different kind of killer: not of humans, but of bacteria.
In February 2020 — more than five decades after the science fiction film introduced the world to perhaps the first great AI villain — a team of researchers at the Massachusetts Institute of Technology used artificial intelligence to discover an antibiotic capable of killing E. coli, which hospitalizes thousands of people a year, as well as an antibiotic-resistant strain of another common bacterial infection, Acinetobacter baumannii. And taking a page from 2001, they named it halicin, after HAL 9000.
The discovery of halicin paints a picture of just how rapid AI-assisted drug discovery can be. Scientists trained their AI model by introducing it to approximately 2,500 molecules (1,700 of which were FDA-approved drugs, and 800 of which were natural products). Once the researchers trained the model to understand which molecules could kill E. coli, the team ran 6,000 compounds through the system, including existing drugs, failed drugs, natural products, and a variety of other compounds.
The system found halicin in a fraction of the time that traditional methods would take, said Bowen Lou, an assistant professor at the University of Connecticut’s School of Business who studies how AI is changing the pharmaceutical industry. “Not only can halicin kill many species of antibiotic-resistant bacteria, it is also structurally distinct from prior antibiotics,” he said in an email. “This discovery is groundbreaking because antibiotic-resistant ‘superbugs’ are a major public health issue that traditional methods have largely failed to address.”
“The idea that you can look at the structures of a small molecule and predict its properties is a very old idea. The way people thought of it is, if you can identify some structures within the molecule, some functional groups, and so on, you can sort of say, ‘What does it do?’” said Regina Barzilay, a distinguished professor of AI and health with MIT’s School of Engineering and co-author of a May 2023 study that identified another potential antibiotic candidate by building upon the methods used in the initial halicin study.
Prior to the use of AI, the challenge of discovering these structures and identifying a drug’s potential use was primarily one of speed, efficiency, and cost. Past analyses show that, between the early 1990s and the late 2000s, the typical drug discovery and development process took 12 years or more. In the case of halicin, the MIT team used AI that can test more than 100 million chemical compounds over the course of only a few days. “It became clear that molecular science is really a good place to apply machine learning and to use new technology,” Barzilay said.
With at least 700,000 deaths every year attributed to drug-resistant diseases — a number projected to grow to 10 million deaths annually by 2050 — the need for speed is great, especially given that the rate of drug advancements has stalled in recent decades. Since 1987, the year scientists identified the last successful antibiotic class used in treating patients, the world has entered what scientists call the “discovery void.”
Crucially, AI can analyze vast amounts of medical data, and, as the discovery of halicin suggests, it can meaningfully accelerate the drug discovery process. This new technology continues to spur significant advancements in the medical field and holds the potential to improve patient outcomes and facilitate more precise treatment methods. It could also lower costs, which would be vital for antibiotic development, given that at least some of the industry’s stagnation is due not to the inability to identify new drugs, but to a lack of market interest and incentive.
“The fact that 90 percent of drugs fail in the clinic tells us that there’s room for improvement. It’s a really complex system. This is exactly what machine learning is made for: really complex systems,” Chris Gibson, the co-founder and CEO of biotech company Recursion, told Vox of recent breakthroughs in the drug discovery space. “It doesn’t mean getting rid of the role people play in many ways, but it augments and turns our scientists into super scientists to have these tools to go faster and to explore more broadly.”
To be clear, the AI programs researchers use for drug discovery vary greatly from science fiction’s AI creations. These advances in pharmaceutical development do not mean robot doctors will run the medical field any time in the near future. But halicin and other recent breakthroughs represent the ability of AI to transform the pharmaceutical industry and bridge more than three decades of the antibiotic discovery void.
The history of AI in medicine
Language models and image generators like OpenAI’s ChatGPT, Google’s Bard, and Midjourney introduced many to the concept of AI when they launched widely in late 2022 and early 2023. But scientists have been using AI — of a sort — for decades.
In 1965, researchers at Stanford University attempted to use a computer program to identify chemical compounds. Considered the “first application of AI to a problem of scientific reasoning,” the DENDRAL project paved the way for future uses of the technology in the scientific community.
Almost a decade later, scientists at Stanford led further developments in medical AI when they created the computer system MYCIN, which helped health care workers diagnose bloodborne bacterial infections in patients. This rules-based system posed a series of questions on symptoms, medical history, test results, and various other factors and would generate a response reporting the likelihood of a particular diagnosis.
However, the rigidity of rules-based systems means they lack the precision necessary to thrive in the ever-changing medical field. (Rules-based systems do not learn new information unless someone manually changes the rules of the program.)
Still, in the late 1970s and the 1980s, several institutions and scientists developed knowledge-based AI systems to help diagnose illnesses, including INTERNIST-I, CASNET, and QMR.
But, depending on your definition of AI, these early technologies don’t even qualify as such. “Many parties who you might talk to in our industry will say that AI existed for many decades or at least for over a decade,” said Alex Zhavoronkov, the CEO of Hong Kong-based AI drug discovery company Insilico Medicine. “They will be right if you define any form of machine learning to be AI.” (An MIT article defines machine learning as a machine’s ability to “imitate intelligent human behavior.”)
Zhavoronkov has a narrower definition of AI drug discovery, saying it refers specifically to the application of deep learning and generative learning in the drug discovery space. Deep learning is a type of machine learning where artificial neural networks (similar but not exactly like the neurons in our brain) allow a machine to learn and advance independent of human intervention.
The “deep learning revolution” — a time when development and use of the technology exploded — took off around 2014, Zhavoronkov said.
Throughout the 2000s, pharmaceutical giants and plucky startups saw an opportunity to accelerate the drug development process. Between 2008 and 2015, many companies focused on AI drug discovery launched, including Evaxion, Exscientia, Recursion, Benevolent AI, and Insilico Medicine. The industry grew even further in the late 2010s when Big Pharma started backing some of these new startups.
“It is worth noting that earlier generations of IT only achieved limited success in drug discovery,” Lou said. “Recent advances in AI have brought about a significant shift in this landscape. AI, with its powerful algorithms and data-driven approaches, has the potential to revolutionize the process of discovering new drugs.”
Recent developments
According to a number of experts who spoke to Vox, the cataloging of biological and chemical information aided in recent drug breakthroughs.
In 2018, DeepMind, a Google-backed AI research laboratory, developed Alphafold, a network that can determine a protein’s structure from its building blocks. “In my opinion, the most fundamental game-changer [in medical AI] is DeepMind’s Alphafold, which has now predicted the structure for essentially all proteins known to us and fundamentally advanced our understanding of biology,” Swarat Chaudhuri, a professor of computer science at the University of Texas Austin, told Vox in an email. “The findings from Alphafold are already having a massive impact on drug and vaccine development.”
Scientists have also been itemizing compounds, or molecules, into chemical libraries, such as the widely used Enamine REAL Space, which contains 36 billion novel molecules. Drug development and pharmaceutical companies order molecules from Enamine Real and then evaluate whether that molecule has the desired effect on the protein being studied (the structure of which is known thanks to Alphafold and other similar software).
Knowing the structure of these proteins and having access to a molecule library are instrumental in determining the potential usefulness of a drug candidate. In the case of halicin, the researchers found a successful antibiotic contender in a library of only 6,000 compounds.
With billions of data points, the potential for new drug discoveries is massive, and new advances could continue to accelerate the process. On August 8, Recursion announced that, in partnership with Nvidia, it predicted how the 36 billion target molecules in Enamine Real’s library interact with approximately 80,000 pockets, or protein binding sites, across over 15,000 proteins. Recursion evaluated around 2.8 quadrillion drug-target pairs, the first step to identifying new drugs.
“Think of this as locks and keys,” said Gibson. The target molecules are the keys, and the protein pockets are the locks. “The idea of a drug in many cases is like a key. You find a very specific key that fits a very specific lock.” Recursion’s work means all molecules in that library (not just those with a similar structure to known useful compounds) can be considered and tested.
Recursion’s development makes it easier to know which keys will fit which locks, but its predictions are not perfect. “It becomes a data layer upon which we can do fast searches,” Gibson said. “Just like a Google search result isn’t always exactly what you’re looking for, but they summarize the top searches … and usually what you want is in one of the top five. It’s the same kind of thing here. We can take 2.8 quadrillion parameters and basically say, ‘If you want an inhibitor of this particular protein, here’s the molecules you might start with.’”
According to Gibson, Recursion’s lab currently conducts as many experiments every 15 minutes as he did in his entire five-year PhD program. “It’s taking almost like an artisanal and bespoke science of the old days and turning it into an industrialized science,” he said. “It’s almost like when making automobiles went from handmade automobiles — every one a little bit custom — to the assembly line.”
Biological tools, robotic automation, and improved computation, among other technological advances, have all played a role in advancing the field of drug discovery and development, Gibson said. AI is simply a critical contributor to these improvements.
What can’t AI do?
AI systems cannot yet accomplish every part of the drug development process alone, particularly in the late stages. Returning to Gibson’s automobile-manufacturing analogy, he says the drug discovery space is currently at the point the Ford Motor Company was when founder Henry Ford quipped, “Any customer can have a car painted any color that he wants so long as it’s black.”
“We’re in the early days where there’s less flexibility in some of these datasets, but they’re built in a standardized way that allows machine learning to take off,” Gibson said.
In June, Insilico Medicine began clinical trials for what Zhavoronkov told CNBC was “the first fully generative AI drug to reach human clinical trials.” The drug, INS018_055, aims to treat idiopathic pulmonary fibrosis, a chronic lung disease, and reportedly relies on both an AI-discovered target and an AI-generated design. Insilico even uses robots in their target discovery lab to develop their small-molecule drug candidates.
Drug candidates typically require at minimum six or seven years to pass through all the necessary human trials, said Zhavoronkov, and the first truly AI-generated drug candidates only started popping up about four years ago. “That is the reason why we haven’t seen AI-generated drugs on the market,” he said. “Many of those drugs that are true AI drugs, they were created just a couple years ago, so they didn’t have the time to get into the human clinical trials. We are I think the first one with a true generative AI drug.”
Still, according to Gibson, we are far from removing people entirely from the drug discovery process. “What’s important to know is that [machine learning and AI] is an incredible tool and, used well, it can help us with many steps in the process. The idea that somebody hits ‘enter’ on an AI algorithm and a drug just pops out, I believe that is a fallacy,” said Gibson. “I am confident no such technology exists today.”
According to Chaudhuri, questions of reliability are a major limiting factor in expanding the role of AI in the medical field. “To deploy AI systems in safety-critical domains, for example, real-time decision-making in health care, you need to trust them,” he said. “But how do you trust a system like GPT-4, which gives you reasonable-sounding answers one minute and complete lies in the next?”
And the truth is, it’s not clear yet when people should put their full trust in the machine’s decision-making processes, Barzilay said. No one needs a person to double-check that the products recommended to them by an Amazon algorithm align with their shopping needs, but medical decisions carry far more weight.
Even with current limitations, however, experts see a great deal of promise in medical AI. “Today, we are paying several billion dollars for each molecule that goes into a drug. It’s unsustainable,” Barzilay said. “There are still a lot of diseases for which we don’t have good drugs, or even for diseases for which we do have approved drugs but we have a whole bunch of side effects.” For example, she said, the breast cancer treatment drug Tamoxifen, while often necessary, comes with a host of harmful side effects, including brain problems.
“What [Tamoxifen] shows to us is we’re nowhere close to where we want to be because of how we develop drugs,” Barzilay said. She believes AI, however, can change the process for the better: “I really think that machine learning should be part of each one of these processes. And I hope and believe that in five or 10 years, drug discovery will be different.”