Search This Blog

Saturday, September 30, 2023

How colleges brazenly get around Supreme Court’s affirmative action ruling

 Now that race-based affirmative action in college admissions has been overturned in a landmark Supreme Court decision, colleges, and universities are scrambling to diversify their student bodies without running afoul of civil rights law.

Several top-ranked schools are rolling out a slew of new essay prompts that fish for demographic information with leading questions — and some are going so far as to directly ask about prospective students’ race.

Johns Hopkins University in Baltimore asks students to “tell us about an aspect of your identity (e.g. race, gender, sexuality, religion, community, etc.) or a life experience that has shaped you as an individual…”

Meanwhile, Rice University in Houston asks applicants: “What perspectives shaped by your background, experiences, upbringing, and/or racial identity inspire you to join our community of change agents at Rice?”

And every single Ivy League school has added an application question about students’ backgrounds, according to college admission expert and Ivy Coach managing partner Brian Taylor.

It’s a clever loophole: ask about race … without expressly requiring students to write about their race.

And some schools aren’t even remotely subtle about their motivations.

Sarah Lawrence College in Bronxville, New York, even cites the Supreme Court’s decision in its essay prompt.

“In the syllabus of a 2023 majority decision of the Supreme Court written by Chief Justice John Roberts, the author notes: ‘Nothing prohibits universities from considering an applicant’s discussion of how race affected the applicant’s life, so long as that discussion is concretely tied to a quality of character or unique ability that the particular applicant can contribute to the university,’” the Sarah Lawrence application reads.

Rice University’s application essay question about students’ backgrounds is among the most brazen.
Getty Images

“Drawing upon examples from your life, a quality of your character, and/or a unique ability you possess, describe how you believe your goals for a college education might be impacted, influenced, or affected by the Court’s decision.”

These schools are pushing the envelope as far as possible — and the federal government seems to be egging them on.

The Biden Administration’s Department of Education is even weighing in, giving colleges tips on how to “enhance racial diversity” in higher education without running afoul of the Supreme Court ruling.

In a report released Thursday, the administration urged schools to increase targeted outreach to non-white communities and give “meaningful consideration in admissions to the adversity students have faced … including racial discrimination.”

New York’s Sarah Lawrence College explicitly cites the Supreme Court’s ruling in their admissions essay prompt.
AP

Taylor says these new essay prompts leave many students grasping at straws: “They’re often confused because they think to themselves, ‘If I’m not an underrepresented minority if I’m not a member of the LGBTQ community, how do I answer this question?’”

Schools are right that diversity is important. We don’t want colleges filled only with ultra-privileged students who could afford the best SAT tutors — or whose families forked over a massive donation.

However, implicitly asking about race makes students feel pressured to write about their ethnicity rather than their character to help their admissions prospects.

“A number of students are disappointed that they feel they have to write about their race in their essay prompt. And they’re correct to think so,” Taylor said, referring to increasing their admissions odds. “They need to let it be known that they are Black or Latino or Native American, and they need to let it be known how that shaped who they are.”

Bunmi Omisore, a 19-year-old Duke freshman, told The Post she’s glad she was in the last class to apply before the ruling for this very reason.

“I wrote about things like my family, ‘The Bachelor’ and biking in my application essays,” Omisore said. “But if I were applying now, I think I would have to forfeit writing about some of those parts of my personality and opt for writing about things that I don’t really like thinking about, like my experiences with racism or my racial trauma.

“You’re going to be having a lot of minority students basically telling a single story, and it’s not fair because that takes away from the uniqueness of the applicant,” she added.

Duke University student Bunmi Omisore is worried that the Supreme Court’s ruling will inadvertently pressure students to write about their race in application essays.
Courtesy of Bunmi Omisore

Not only is this tactic a brazen abuse of a legal loophole — it also reduces students to their immutable characteristics and incentivizes them to performatively boil themselves down to their race. That’s the opposite of progress.

Abolishing legacy admissions — which disproportionately favor white applicants — and implementing socioeconomic affirmative action, which would boost disadvantaged students of all backgrounds, are two better ways to promote diversity

Colleges need to figure out how to do this without indulging in race essentialism.

https://nypost.com/2023/09/30/how-colleges-are-skirting-supreme-couts-affirmative-action-ruling/

Abortion may not be a game-changer for Democrats in 2024

 Since the Dobbs v. Jackson Women’s Health Organization Supreme Court decision was announced last year, there has been much discussion about the impact of the decision on the 2024 elections. Some have referred to the decision as a “political earthquake,” or asserted that the “political landscape has been altered significantly, with no reversal in sight.”

Indeed, Democrats believe that abortion is a winning issue for them. They have already started airing ads attacking Donald Trump and Florida Gov. Ron DeSantis for supporting abortion bans, and Joe Biden plans to make his 2024 campaign for president “the most overtly abortion rights platform” in history. 

There is no doubt that the abortion issue was influential in 2022. Studies suggested it was a motivating factor for women under 50 in the midterm elections that year. Many experts credit the issue for flipping the Michigan legislature from Republican to Democratic control and for mitigating Democratic losses generally in that election cycle. Also, when abortion was literally on the ballot, as it was in the six states that held abortion referenda, initiatives protecting abortion access prevailed in each state, including reliably red states like Kansas and Kentucky.

As a policy issue, Americans have long been divided on the legality of abortion. Even after Dobbs, which held that the states have the ability to make their own abortion laws, we have seen movement toward the extremes. There has been a shift by Democratic and independent voters to maintain or expand the protections provided under Roe v. Wade (1973), while a majority of Republicans have not softened their stances on abortion and support a complete ban.

Although national polls about attitudes toward abortion are interesting, they are often not instructive for examining the issue’s effect on 2024 elections since the Dobbs decision threw abortion back to the states. Since the decision was handed down, states have passed various types of abortion laws. States like Indiana and North Dakota have passed laws banning abortion entirely, states like South Carolina have a six-week ban and North Carolina passed a twelve-week ban with some exceptions.

Despite Democrats’ confidence that Americans are on their side concerning abortion, in a recent Meredith Poll, we found that North Carolinians were almost evenly divided between those supporting the new abortion law (47 percent) and those opposed to it (45 percent). We also found that only 14 percent of those surveyed identified abortion as the top issue in terms of their voting decision in 2024—well behind economic issues, which was the top issue for 40 percent of respondents. Finally, we asked about the new abortion law’s effect on motivating the respondents to vote, and there were no differences in the law’s impact on men and women voters or voters in different age groups. There were only slight differences between Democratic and Republican voters who saw abortion as a motivating issue.

The other problem with giving abortion too much significance in the 2024 elections is that it discounts the effects of polarization on the American voter. Two kinds of polarization are affective polarization, or feeling close to your chosen party and negatively about the opposing party, and policy polarization, or people of different parties choosing very different policy outcomes. Using data from the Meredith Poll from 2017-2023, we analyzed responses to questions about policy issues, including abortion, as well as their perceptions of people affiliated with the two major parties. We found that affective polarization was significantly stronger than policy polarization, even on a divisive issue like abortion.

The bottom line for 2024 is whether voters—even those who have strong policy preferences on the issue of abortion—will vote based on their position on abortion or for their political party. Our evidence suggests that Democrats may be overstating the influence of abortion, and party identification will be a better predictor of a person’s vote.

David McLennan is a professor of political science at Meredith College and director of the Meredith Poll. Whitney Ross Manzo is an associate professor of political science at Meredith and assistant director of the Meredith Poll.

https://thehill.com/opinion/congress-blog/4231921-abortion-may-not-be-a-game-changer-for-democrats-in-2024/

House Passes Bill Averting a Government Shutdown, Senate Approval Expected

 Just like that, Republican dissidents were tossed aside and a government shutdown will be avoided via a stopgap measure until mid-November.

Shutdown Likely Avoided

The Wall Street Journal reports Congress Set to Avert Government Shutdown

The House voted 335-91 for the funding measure, which includes $16 billion in disaster relief but omits aid for Ukraine. It also excludes border-security measures sought by Republicans. The margin exceeded the two-thirds majority needed to clear the bill through the House, which considered the legislation under special procedures requiring a supermajority of votes. All but one Democrat voted in favor of the measure, while nearly half of Republicans voted against it.

In reaching for a compromise, Republicans argued that the party had exhausted its options after dissident conservatives derailed an earlier plan, and said that the only choice now was to pass a bill extending funding at 2023’s $1.6 trillion annual rate through Nov. 17. That squares with major components of the approach being taken in the Senate, except that the Senate version includes an emergency $6 billion for Ukraine.

McCarthy dismissed opposition from conservatives who have threatened to oust him as speaker. “If somebody wants to make a motion against me, bring it. There has to be an adult in the room,” he said after the vote.

Republicans missed an opportunity to send the message representative Matt Gaetz wanted.

The same bill that did pass plus some border security provisions with minimal aid to Ukraine (not a full blockage), may have gotten through the Senate.

Instead, as noted McCarthy’s House Bill Goes Down in Embarrassing 198-232 Vote.

I commented ….

One and Only Winner

The only winner in this battle is the Democrats. The idea that 10-12 holdouts are going to swing this puppy is nonsensical.

You may or may not agree with what Matt Gaetz wants, but that is irrelevant. Gaetz does not have the votes to replace McCarthy and 10 cannot tell 212 what to do.

The last vote for House Speaker went 15 rounds and if McCarthy is ousted, the next vote might go 30 or 100 rounds. Along the way, the top Democrat will get more votes than anyone Gaetz wants.

Gaetz is gleeful over handing a talking point to Democrats. No good can possibly come from it.

Nothing Instead of Something

And here we are. Instead of getting something, Republicans get nothing.

McCarthy now taunts Gaetz: “If somebody wants to make a motion against me, bring it.”

https://mishtalk.com/economics/the-house-passes-bill-averting-a-government-shutdown-senate-approval-expected/

Did the government get a bad deal on the Covid-19 boosters?

 After Pfizer and Moderna hiked the prices of their Covid-19 vaccines this year, the federal government will now pay nearly three times more than it did previously for each dose. And it’s paying more than countries that did far less to support vaccine development.

Health and Human Services Secretary Xavier Becerra glossed over the dramatic increase in the amount the government is paying for vaccines in a public appearance last week, when he got vaccinated at a CVS pharmacy in D.C.

Becerra said then that the price that the federal government will pay for some vaccines “will be similar to what we paid in the past for the vaccines.”

However, the Centers for Disease Control and Prevention is paying $81.61 for the Moderna booster this year, and $85.10 for the Pfizer shot — which is around triple the amount the federal government paid for each shot last year. Moderna charged the government $26.36 per booster dose last year, and Pfizer charged $30.48.

The change is partly because the federal government ran out of money to continue to buy the vaccines in bulk to distribute across the country, so the vast majority of vaccine purchases are instead flowing through normal channels and will be paid for by a variety of different insurance plans this year, including Medicare and Medicaid.

The CDC is buying the vaccines for its $1 billion program to provide vaccines for uninsured people. The doses will be sent to state and local health departments and community health centers.

The CDC’s prices this year are significantly discounted from the price the companies are charging private insurers, however, which are $115 for the Pfizer vaccine and $128 for Moderna.

The United States is also set to pay a much higher price than is being considered in Europe for the same Moderna product.

European Union officials are in talks to potentially buy Moderna doses for around 25 euros apiece, the equivalent of around $26.30, according to a report in the Financial Times.

Though the price is not final, the potential amount is roughly one-third of what the United States is paying.

Senate health committee Chair Bernie Sanders (I-Vt.), who called Moderna’s CEO Stéphane Bancel to Capitol Hill over the company’s decision to hike its vaccine price in March, was not pleased by reports of how much lower the EU’s price could be.

“Europeans are now negotiating with Moderna for a new vaccine, and their price is going to be substantially less in Europe than it is in the United States. So that’s exactly the issue. We’re trying to get a hold of reasonable pricing,” Sanders said in a brief interview at the Capitol.

Sanders has held up the confirmation of President Biden’s nominee to lead the National Institutes of Health to push for the federal government to include contracting provisions mandating that the U.S. should pay similar prices to other countries if they pay to develop a drug. HHS did so in a recent contract for a drug to prevent Covid-19 with the company Regeneron.

“Taxpayers and U.S. patients are getting a raw deal, again. It’s time for the Biden administration to announce a comprehensive reasonable pricing clause. Drugs and vaccines that benefit from public R&D funding must offer a fair price in return,” Public Citizen Access to Medicines Director Peter Maybarduk said.

https://www.statnews.com/2023/09/28/covid-vaccines-higher-price/

She got off dialysis. Her doctor thinks others with acute kidney injury can too

  It was during some of the darkest days of the pandemic — November 2020 — when Melissa Lawson, then 36, was helicoptered from an emergency room in Santa Rosa, Calif., to a hospital here, her blood so thick with white blood cells doctors thought she might not survive the night.

She did, after undergoing a procedure to clear her blood. Diagnosed with acute myeloid leukemia, she got chemo, and less than a month later was in remission. But then her kidneys suddenly stopped working — she’s still not sure why — her lungs filled with fluid, and she was rushed to the ICU, where she coded twice. Her parents were called in (despite strict Covid protocols) to say their last goodbyes.

Lawson survived again, but was in the ICU for 51 days, alone. She battled infections. Her appendix was removed, her bowel resected. Worst of all, she told STAT, was the dialysis she needed three times a week for her acute kidney injury. The sessions left her freezing, nauseous and vomiting, with pounding headaches — and feeling anxious and depressed.

“It’s sad going into a big room where it feels like everyone’s dying. People are moaning and screaming,” said Lawson, a children’s hairdresser who goes by the nickname Liddy. She’d pull her eye mask down and face mask up, wrap herself in a blanket, and try to disappear.

She was told her kidneys would heal. But they didn’t; dialysis became a regular routine. She moved to UCSF Medical Center seeking better care and a place that would allow her parents to visit. There, she met Chi-yuan Hsu, UCSF’s chief of nephrology, who was looking to study patients who might be successfully weaned from dialysis. He believed many patients with acute kidney injury like Lawson stayed on dialysis for longer than they needed.

The results of a new study by Hsu, published Thursday in the Journal of the American Society of Nephrology, validate his suspicions. The study of nearly 8,000 patients, nearly 2,000 with acute kidney injury, found 40% of patients with acute kidney injury recovered their kidney function. But of these patients, just 18% were weaned from dialysis through having fewer sessions, and 9% by having shorter sessions.

More than 70% of these patients ended up eventually stopping dialysis without any weaning — “cold turkey” as Hsu puts it — suggesting they could have been having fewer, or shorter treatments earlier. This is important, he said, because dialysis not only impacts quality of life, as it did for Lawson, it can also lead to infection and heart damage, and possibly — this is still under debate — to additional kidney injury that could inhibit recovery and lead to a need for permanent dialysis.

The study confirms, Hsu and his co-authors said, that patients with acute kidney injury who may be able to stop dialysis (and who make up between 15-20% of new dialysis patients nationally) may be losing the option of timely withdrawal from dialysis because they are treated largely the same as those with end-stage kidney disease who have no hope of stopping dialysis unless they receive a kidney transplant.

It’s not a surprise that the dialysis industry caters to those with end-stage disease; until 2017, patients with acute kidney injury weren’t even categorized separately (or eligible for reimbursement from Medicare). Historically, most either recovered quickly after brief dialysis treatment or, until recent medical advances, died from whatever caused their kidney injuries.

“When the system was designed, these patients were not considered. Then the patients started surviving,” said Hsu. Despite that, he said, dialysis “units are staffed, designed, and structured for end-stage kidney disease.”

Patients with acute kidney injury don’t always get the more frequent lab tests and assessments needed to reveal kidney recovery. The study found only one quarter of such patients received timed urine collections to test their kidney function during their first month of dialysis, something the authors said should be done far more often.

When Lawson learned of Hsu’s interest in weaning patients off dialysis, she eagerly signed on. She went six days, safely, without treatment. For the first time, Lawson said, she felt relatively healthy, and had hope for her future.

Hsu was inspired to start his research on dialysis weaning by one of his previous patients, Marla Levy, who’d been on dialysis after three open-heart surgeries. Hsu examined her and thought her kidneys were healthy enough to work on their own and that she could stop dialysis.

“It was like being told I had won a hundred million dollars in the lottery. It was the best news of my life,” Levy said. The day she had her dialysis catheter taken out, she said in a writeup for UCSF, “is up there with getting married and giving birth to my children.”

Hsu thought Lawson, with her youth, her eagerness to heal, and extremely supportive family was an ideal candidate for weaning. “I think there are people we could take off if we invested the time,” said Hsu. “For these people, it’s an enormous improvement in life.”

But after being transferred from UCSF to a rehab hospital, she was put back on dialysis three times a week because she couldn’t be monitored as closely. It was a huge blow.

Lawson dreaded the side effects and the six hours it took to have and travel to each procedure. She’d return to her room at 10 p.m. to a cold meal on her dinner tray. It didn’t matter, she said, she was too nauseous to eat.

“It could have been a hot 5-star meal and I wouldn’t have been able to stomach it because of dialysis,” she said. She dropped to 87 pounds.

She eventually left the rehab hospital for home after five and half months of hospitalization, with a feeding tube in place and a schedule for dialysis three times a week near her home.

Lawson started making friends with other dialysis regulars. She baked cookies for them and went out for long breakfasts to discuss their interests: the Raiders, the 49ers, trains. “It was my crew,” she said, though they were decades older and mostly suffering from chronic conditions such as diabetes and hypertension. Lawson couldn’t have been a squarer peg.

Still in touch with Hsu, Lawson entered another trial in May of 2022 where her dialysis treatments dropped to two days a week. But she still felt terrible most days, threw up constantly, and felt sad and anxious with no energy for most activities. “I started to feel how am I going to do this forever?” she said, “How will I work and have a family?”

After an unsuccessful attempt at home dialysis last fall, doctors urged her to get a fistula — a kind of “supervein” that makes it easier for patients on long-term dialysis to endure the treatment. She dragged her feet. She didn’t want the fistula surgery, which can be disfiguring, and still thought, despite what the doctors said, that maybe she didn’t need dialysis at all.

At her pre-surgical consult for the fistula this past March, her surgeons were shocked. “Both doctors when they came in did a double take and said, ‘You need a fistula?’” she said. She looked far too young. “They thought I was the patient,” said Lawson’s mother, Debbie, who has attended most appointments and keeps careful track of her daughter’s medical journey neatly written in a 100-page red Oxford composition book. But a round of lab tests shortly before the surgery revealed a turnaround.

Against all expectations, Lawson had been steadily recovering. A few days later, just days after her birthday, Hsu called and told her to cancel the fistula surgery and stop dialysis.

Six months later, she hasn’t had a dialysis treatment. There’s no more nausea and no more depression. She feels strongly that her faith in God played a role in her survival. She is not out of the woods (a phrase she’s come to hate because doctors use it so much) — a bout of pneumonia, for instance, could knock her back into dialysis. But, for now, she’s off the machines and hasn’t felt so good, or so free, for years.

Ian McCoy, an assistant professor of nephrology at UCSF and the study’s lead author, emphasizes that patients should not go off dialysis without their doctor’s OK. It could be a death sentence. “You can fill up with fluid in your lungs. If your potassium and electrolytes are off, it can affect heart rhythms.” On the other hand, he said, staying on dialysis poses its own risks. Just how much? Doctors don’t know.

Acute kidney injury patients have been difficult to study in clinical trials. “These patients tend to be extraordinarily sick,” said Paul Palevsky, a professor of medicine at the University of Pittsburgh School of Medicine and immediate past president of the National Kidney Foundation who has helped conduct such trials.

Palevsky told STAT he has no doubt that some acute kidney patients are on dialysis longer than needed and could be better managed. Perhaps they should have less frequent or shorter dialysis or just stop altogether if their kidneys show signs of recovery. The problem is, he said, no data exists to show what’s best. “There is no right answer,” he said.

Acute kidney patients also differ widely in the pace of their recovery and need for dialysis, he said, adding, “Since there’s no uniform approach, what it requires is careful attention to the patient.”

That’s readily done by nephrologists like Hsu and McCoy, academics who work in a hospital where lab results can be obtained in a matter of hours. It’s much harder, they said, for nephrologists in private practice, who may be monitoring 100 or more patients at a time and not getting timely results.

Palevsky advises patients to be proactive like Lawson was. “Keep asking how your kidneys are doing and if there is evidence of recovery. If urine output is picking up, point that out,” he said.

Hsu and McCoy believe structural issues — from how dialysis is reimbursed to how chairs in dialysis centers are allocated — may be keeping patients in dialysis too long.

The for-profit dialysis industry has come under scrutiny over concerns that a focus on cost-cutting has affected quality of care. A study last year found that home dialysis is used much less frequently in the U.S. than in other countries, and questioned whether financial incentives played a role.

While the current paper did not analyze economics or suggest that financial motives keep patients on dialysis schedules heavier than they need, McCoy said it’s likely money plays some role. “Units are set up on a three-day-a-week schedule. You have [dialysis] one day a week, a chair is empty. Who’s going to take that?” he asked.

Nephrologists are paid less to see patients who are not on dialysis, despite the complexity of their cases, he noted. And more frequent lab tests can raise costs. Fistulas are promoted because dialysis centers can get penalized if too many of their patients have catheters, due to the risk of infection.

But providers may also want to keep patients on dialysis for the support they receive from social workers and nutritionists while in treatment. There may also be a comfort level for physicians dealing with patients who are very ill. “If someone’s on dialysis, you know they’re pretty safe,” said Hsu.

Hsu isn’t calling for a radical revision of dialysis care. But he does want to get word out that his fellow nephrologists should be on the lookout for patients like Lawson. That’s the message he’ll give at a talk at the annual American Society of Nephrology meeting in November.

“My message is don’t do anything different right now, just be more vigilant,” he said.

Hsu and McCoy agree more research is needed. Studies like the one released Thursday — looking closely at the records of dialysis patients — have been difficult because the large for-profit dialysis centers have not generally shared such data. This study was possible, they said, because of the cooperation and financial support of Satellite, a mid-sized nonprofit dialysis provider with 70 clinics based in San Jose, Calif.

Wael Hussein, a nephrologist and chief medical officer for research and development at Satellite and a co-author, said the study highlights a “gap between the expected level of care and what is being delivered” and “should be a stimulant for more research to develop solutions to improve care.”

He said part of the problem was dialysis centers not receiving timely information from hospitals about patients so they can better tailor treatment. “A lot of information you need to manage patients, you just don’t get,” he said. The Centers for Medicare and Medicaid Services should also allocate more funding for the care of acute kidney patients on dialysis, he said, since they require more resources, such as more frequent testing.

Lawson still has serious kidney disease and she’ll always be under surveillance for a return of her cancer. But she couldn’t be happier. Her bone marrow biopsies have been clear for three years, her feeding tube is out, her weight has climbed back over 100 pounds, and she’s back to playing with her dog Tucker. She recently went on a month-long road trip across the U.S. with her family and is training for a walking marathon.

She’s happy to share the details of her case in hopes it will help others, she said during an hours-long interview within sight of the 11th floor of the UCSF Medical Center where she spent so much time. She said providers need to work as hard as possible to take people off dialysis who no longer need it — and give them their lives back.

“They don’t realize how difficult it is,” she said as she rubbed the small raised reddish welt near her shoulder where her dialysis catheter used to be.

https://www.statnews.com/2023/09/28/dialysis-acute-kidney-injury-study/

AI-driven data analysis could exacerbate misaligned incentives in biomedical research

 This summer OpenAI released Code Interpreter, a plug-in for the popular ChatGPT tool that allows it to take in datasets, write and run Python code, and “create charts, edit files, perform math, etc.“ It aims to be nothing short of the ideal statistical collaborator or research software engineer, providing the necessary skill and speed to overcome the limitations of one’s research program at a fraction of the price.

It is a bad omen, then, that while statisticians are known for pestering researchers with difficult but important questions like “What are we even trying to learn?”, Code Interpreter responds to even half-baked requests with a cheerful “Sure, I’d be happy to.” There are risks to working with a collaborator that has both extraordinary efficiency and an unmatched desire to please.

Let me state up front that AI’s benefits for science could be immense, with potentially transformative implications for the life and physical sciences. The democratization of data analysis represented by tools like Code Interpreter may be no exception. Giving all scholars access to advanced methods will open the doors to innovative research that would otherwise have been filed away as unachievable. Yet just like its potential to accelerate the exceptional qualities of science, AI is at risk of accelerating its many flaws.

One risk is widely — and rightfully — discussed: inaccuracy. The improvement of AI-driven software development will dramatically increase the speed and complexity of scientific programming, decrease the training required to write advanced code, and provide a veneer of authority to the output. Given that very few parties are incentivized to spend time and money on careful code review, it will be almost impossible to assess the accuracy of plausible-looking code that runs cleanly.

But there is another problem that is getting far less attention: Rapid, complex, and technically sound data analysis is insufficient — and sometimes antithetical — to the generation of real knowledge.

Indeed, for many of the commonly cited issues on the spectrum of scientific misconduct — think HARKingp-hacking, and publication bias — the primary sources of friction for unaware (or unscrupulous) researchers are human constraints on speed and technical capacity. In the context of a discipline still grappling with those practices, AI tools that become more efficient at running complex scientific studies, more effective at writing them up, and more skilled at responding to users’ requests and feedback are liable to pollute the literature with compelling and elegantly presented false-positive results.

This misaligned optimization scheme is familiar to human scientists already; the Center for Open Science, to take one example, has spent a decade re-training scientific fields that overoptimized for productivity and prestige to begin rewarding rigor and reproducibility. These efforts have revealed scientific models that reorient the incentives placed on researchers. For example, registered reports — a form of scientific publication where manuscripts are submitted, reviewed, and accepted based solely on the proposed question and empirical approach — may provide a setting in which AI tools could advance biomedical knowledge rather than muddy it.

Yet new publication models cannot break academic science’s fixation on quantity. Without broader shifts in norms and incentives, the speed offered by AI tools could threaten the potential value of some of biomedical science’s most promising new offerings, like preprints and large public datasets. In this context, we may be forced to rely more and more heavily on meta-analyses (perhaps conducted by AI) but with less and less ability to factor expert judgment and methodological credibility into their evaluation.

Instead, it will be important to rethink how we reward the production of science. “Quality over quantity” is so trite as to be meaningless, but the continued importance of metrics like publication count and H-index — and even the persistent demand for paper mills — demonstrate that we have yet to fully embody its spirit.

There exist clues for a potential path forward. Some existing norms and policies encourage researchers to focus on a few high-quality research outputs at key career stages. For example, in economics, faculty candidacy relies largely on a single “job market paper,” and in biomedicine, the Howard Hughes Medical Institute requests that scientists highlight five key articles on applications. More recently, the National Institute for Neurological Disorders and Stroke rolled out a series of rigor-focused grants aiming to support education and implementation. Such efforts can collectively shift incentive structures toward slow, rigorous, and ambitious work; more funding and gatekeeping bodies should consider moving in this direction.

In addition to aligning human systems, efforts to optimize AI-driven research tools towards both technical capacity and knowledge generation will be vital. Ongoing AI alignment programs focused on safety may offer clues for building responsible virtual collaborators. For example, efforts to improve the transparency and reproducibility of AI output could produce relevant insights. Yet the case of AI in scientific research is unique enough — for instance, an identical response to an identical prompt could be either valid or invalid depending on past conversations that are unavailable to the AI — that it likely requires its own set of solutions.

To meet this moment, we will need to build and support research programs aiming to understand and improve both the tools and researchers’ relationship to them. There are countless fields of study that touch these topics, including, though certainly not limited to, human-computer interaction, sociology of science, AI safety, AI ethics, and metascience. Collaboration and conversation across these domains would provide strong insight into the most fruitful paths forward.

Scientific institutions can and should help facilitate this work. In addition to supporting efforts to unlock AI’s potential benefits for scientific productivity, government and philanthropic funders should invest in research focused on understanding how AI can be steered towards effective generation of reliable and trustworthy knowledge; as argued above, these goals can often be at odds in the context of human social systems.

A good example of this kind of institutional support is the National Institute of Standards and Technology’s Trustworthy and Responsible AI Resource Center, which will soon pilot several initiatives aimed at providing a space for researchers to elicit and study AI’s real-world effects on users in a controlled environment. NSF’s Artificial Intelligence Research Institutes — some with a focus on human-AI collaboration — represent another promising approach.

In general, it is understandable that the conversation around AI-driven research tools is an optimistic one. An eager, technically skilled, and highly efficient collaborator is a dream for any scientist. Matching one to every scientist could be a dream for society. But to reach that goal, we need to remember that, sometimes, a perfect collaborator is too good to be true

Jordan Dworkin is the program lead for metascience at the Federation of American Scientists, a nonprofit, nonpartisan policy research organization working to develop and implement innovative ideas in science and technology.

https://www.statnews.com/2023/09/29/ai-chatgpt-health-science-rigor-trust-biomedical-research/