The Efficacy of Personalized Drugs Designed for Only One Patient Cannot Be Tested by Randomized Double-Blind Clinical Trials (RCTs)

We know that there are times when therapies work for some patients, but not for others. But clinical trials often do not account for such differences. If the effects of the new drug are not widespread enough among the general population, the trial will be deemed a failure, and the F.D.A. will not allow the drug to be taken even by the patients who would have benefitted from it. Maybe the solution is liberty. Allow physicians liberty on what therapies to suggest, and patients liberty on what therapies to try. This especially makes sense when the disease is dire and no effective therapy is yet widely known.

Many predict that we are moving toward personalized medicine. We need less regulation and more liberty so personalized medicine can progress, and more patients can be more quickly cured of more diseases. We need a sense of urgency in requesting liberty.

(p. D3) A new drug, created to treat just one patient, has pushed the bounds of personalized medicine and has raised unexplored regulatory and ethical questions, scientists reported on Wednesday [Oct. 9, 2019].

The drug, described in The New England Journal of Medicine, is believed to be the first “custom” treatment for a genetic disease. It is called milasen, named after the only patient who will ever take it: Mila (mee-lah) Makovec, who lives with her mother, Julia Vitarello, in Longmont, Colo.

. . .

Ms. Vitarello . . . set up Mila’s Miracle Foundation and was appealing for donations on GoFundMe. So, she began fund-raising in earnest, eventually raising $3 million for a variety of research efforts.

Dr. Yu’s team oversaw development of the drug, tested it in rodents, and consulted with the Food and Drug Administration. In January 2018, the agency granted permission to give the drug to Mila. She got her first dose on Jan. 31, 2018.

With continued treatments, the number of seizures has diminished so much that the girl has between none and six a day, and they last less than a minute.

Milasen is believed to be the first drug developed for a single patient (CAR-T cancer therapies, while individualized, are not drugs). But the path forward is not clear, Dr. Yu and his colleagues acknowledged.

. . .

. . . how might a custom drug’s efficacy might be evaluated, and how should regulators weigh the urgency of the patient’s situation and the number of patients who could ultimately be treated.

For the full story see:

Gina Kolata. “Drug Designed for One Raises Many Questions.” The New York Times (Tuesday, October 15, 2019 [sic]): D3.

(Note: ellipses, and bracketed date, added.)

(Note: the online version of the story has the date Oct. 9, 2019 [sic], and has the title “Scientists Designed a Drug for Just One Patient. Her Name Is Mila.” Where the more detailed online version differs from the print version, the passages quoted above follow the print [sic] version.)

The academic article co-authored by Dr. Yu that reports on the personalized drug milasen is:

Kim, Jinkuk, Chunguang Hu, Christelle Moufawad El Achkar, Lauren E. Black, Julie Douville, Austin Larson, Mary K. Pendergast, Sara F. Goldkind, Eunjung A. Lee, Ashley Kuniholm, Aubrie Soucy, Jai Vaze, Nandkishore R. Belur, Kristina Fredriksen, Iva Stojkovska, Alla Tsytsykova, Myriam Armant, Renata L. DiDonato, Jaejoon Choi, Laura Cornelissen, Luis M. Pereira, Erika F. Augustine, Casie A. Genetti, Kira Dies, Brenda Barton, Lucinda Williams, Benjamin D. Goodlett, Bobbie L. Riley, Amy Pasternak, Emily R. Berry, Kelly A. Pflock, Stephen Chu, Chantal Reed, Kimberly Tyndall, Pankaj B. Agrawal, Alan H. Beggs, P. Ellen Grant, David K. Urion, Richard O. Snyder, Susan E. Waisbren, Annapurna Poduri, Peter J. Park, Al Patterson, Alessandra Biffi, Joseph R. Mazzulli, Olaf Bodamer, Charles B. Berde, and Timothy W. Yu. “Patient-Customized Oligonucleotide Therapy for a Rare Genetic Disease.” New England Journal of Medicine 381, no. 17 (Oct. 9, 2019): 1644-52.

An accompanying editorial commenting on the regulatory challenges raised by personalized drugs like milasen is:

Woodcock, Janet, and Peter Marks. “Drug Regulation in the Era of Individualized Therapies.” New England Journal of Medicine 381, no. 17 (Oct. 9, 2019): 1678-80.

Rigid Guidelines Don’t Allow for Individualizing Treatment and Discount the Doctor’s Clinical Judgment

The criticism of the Clovers sepsis clinical trial is that the the treatment and placebo arms of the trial each require rigid adherence to a protocol, and such adherence rules out personalizing individual treatment based on individual differences among patients and doctors’ clinical judgment based on past experiences. That criticism seems plausible and also seems to apply, not just to the Clovers sepsis clinical trial, but to all< randomized double-blind clinical trials.

(p. D1) A large government trial comparing treatments for a life-threatening condition called sepsis is putting participants at risk of organ failure and even death, critics charge, and should be immediately shut down.

A detailed analysis of the trial design prepared by senior investigators at the National Institutes of Health Clinical Center in Bethesda, Md., concluded that the study “places seriously ill patients at risk without the possibility of gaining information that can provide benefits either to the subjects or to future patients.”

In a letter to the federal Office for Human Research Protection, representatives of Public Citizen’s Health Research Group compared the study, called Clovers, to “an experiment that would be conducted on laboratory animals.”

“The human subjects of the Clovers trial, as designed and currently conducted, are unwitting guinea pigs in a physiology experiment,” Dr. Michael Carome and Dr. Sidney M. Wolfe wrote in their letter.

Begun in March, Clovers is funded by the N.I.H. — despite the criticism of its own investigators — and aims to enroll 2,320 pa-(p. D3)tients at 44 hospitals around the country.

. . .

At issue is whether patients participating in Clovers are being given treatment that deviates from usual care — so much so that lives may be endangered by the research.  . . .

When patients experience septic shock, current guidelines call for raising blood pressure by administering fluids within the first three hours of care, and then administering vasopressors within the first six hours if patients do not respond to fluids.

Vasopressors can be administered early on, during or after the infusion of fluids; a new treatment guideline for hospitals says the drugs should be started within the first hour if patients aren’t responding to intravenous fluids.

Many physicians have been critical of rigid guidelines like this one because they don’t allow for individualizing treatment and appear to discount the doctor’s clinical judgment.

Both fluids in large amounts and vasopressors can cause serious complications, but when a patient’s condition continues to deteriorate, doctors use both interventions, adjusting them depending on the severity of illness.

They generally start with fluids, which in small amounts are considered less toxic than vasopressors.

But participants in Clovers are randomly assigned to a “liberal fluids” group who receive large infusions of fluids in a very short time but limits the use of vasopressors, or to a “restrictive fluids” group in which fluids are minimized and drug treatment begun earlier.

For the full story see:

Roni Caryn Rabin. “Critics Demand Halt of a Sepsis Trial.” The New York Times (Tuesday, September 25, 2018 [sic]): D1 & D3.

(Note: ellipses added.)

(Note: the online version of the story has the date Sept. 24, 2018 [sic], and has the title “Trial by Fire: Critics Demand That a Huge Sepsis Study Be Stopped.”)

“Most Published Research Findings Are False”

(p. C1) How much of biomedical research is actually wrong? John Ioannidis, an epidemiologist and health-policy researcher at Stanford, was among the first to sound the alarm with a 2005 article in the journal PLOS Medicine. He showed that small sample sizes and bias in study design were chronic problems in the field and served to grossly overestimate positive results. His dramatic bottom line was that “most published research findings are false.”

The problem is especially acute in laboratory studies with animals, in which scientists often use just a few animals and fail to select them randomly. Such errors inevitably introduce bias. Large-scale human studies, of the sort used in drug testing, are less likely to be compromised in this way, but they have their own failings: It’s tempting for scientists (like everyone else) (p. C2) to see what they want to see in their findings, and data may be cherry-picked or massaged to arrive at a desired conclusion.

A paper published in February [2017] in the journal PLOS One by Estelle Dumas-Mallet and colleagues at the University of Bordeaux tracked 156 biomedical studies that had been the subject of stories in major English-language newspapers. Follow-up studies, they showed, overturned half of those initial positive results (though such disconfirmation rarely got follow-up news coverage). The studies dealt with a wide range of issues, including the biology of attention-deficit hyperactivity disorder, new breast-cancer susceptibility genes, a reported link between pesticide exposure and Parkinson’s disease, and the role of a virus in autism.

Reviews by pharmaceutical companies have delivered equally grim numbers. In 2011, scientists at Bayer published a paper in the journal Nature Reviews Drug Discovery showing that they could replicate only 25% of the findings of various studies. The following year, C. Glenn Begley, the head of cancer research at Amgen, reported in the journal Nature that he and his colleagues could reproduce only six of 53 seemingly promising studies, even after enlisting help from some of the original scientists.

With millions of dollars on the line, industry scientists overseeing clinical trials with human subjects have a stronger incentive to follow high standards. Such studies are often designed in cooperation with the U.S. Food and Drug Administration, which ultimately reviews the findings. Still, most clinical trials produce disappointing results, often because the lab studies on which they are based were themselves flawed.

For the full essay see:

Harris, Richard. “Dismal Science In the Search for Cures.” The Wall Street Journal (Saturday, April 8, 2017 [sic]): C1-C2.

(Note: bracketed year added.)

(Note: the online version of the essay was updated April 7, 2017 [sic], and has the title “The Breakdown in Biomedical Research.”)

The essay quoted above is adapted from Mr. Harris’s book:

Harris, Richard. Rigor Mortis: How Sloppy Science Creates Worthless Cures, Crushes Hope, and Wastes Billions. New York: Basic Books, 2017.

The 2005 paper by Ioannidis mentioned above is:

Ioannidis, John P. A. “Why Most Published Research Findings Are False.” PLoS Medicine 2, no. 8 (2005): 696-701.

Even With Quick Cheap Malaria Lab Tests, Dog Noses Can Still “Be Very Useful”

(p. D4) A small pilot study has shown that dogs can accurately identify socks worn overnight by children infected with malaria parasites — even when the children had cases so mild that they were not feverish.

The study, a collaboration between British and Gambian scientists and the British charity Medical Detection Dogs, was released last week at the annual convention of the American Society of Tropical Medicine and Hygiene.

. . .

The new study, its authors said, does not mean that dogs will replace laboratories. Inexpensive rapid tests for malaria have been available for over a decade; more than 200 million people in dozens of countries are infected each year.

But for sorting through crowds, malaria-sniffing dogs could potentially be very useful.

Some countries and regions that have eliminated the disease share heavily trafficked borders with others that have not. For example, South Africa, Sri Lanka and the island of Zanzibar have no cases but get streams of visitors from Mozambique, India and mainland Tanzania.

And when a region is close to eliminating malaria, dogs could sweep through villages, nosing out silent carriers — people who are not ill but have parasites in their blood that mosquitoes could pass on to others.

. . .

If just one chemical indicated cancer or malaria, “we’d have discovered it by now,” said Claire Guest, who founded Medical Detection Dogs in 2008 and oversaw dog training in the study. “It’s more like a tune of many notes, and the dogs can pick it up.”

For the full commentary see:

Donald G. McNeil Jr. “Global Health; Sniffing Out Malaria in Its Tracks.” The New York Times (Tuesday, November 6, 2018 [sic]): D4.

(Note: ellipses added.)

(Note: the online version of the commentary has the date Nov. 5, 2018 [sic], and has the title “Global Health; Dogs Can Detect Malaria. How Useful Is That?”)

A later-published version of the initial “small pilot study” discussed above is:

Guest, Claire, Margaret Pinder, Mark Doggett, Chelci Squires, Muna Affara, Balla Kandeh, Sarah Dewhirst, Steven V. Morant, Umberto D’Alessandro, James G. Logan, and Steve W. Lindsay. “Trained Dogs Identify People with Malaria Parasites by Their Odour.” The Lancet Infectious Diseases 19, no. 6 (June 2019): 578-80.

Universally Applicable Egg Guidelines Are Impossible Because Some Are Hypo-Responders and Others Are Hyper-Responders to Dietary Cholesterol

(p. D5) “Intervention studies have shown that moderate egg consumption doesn’t appreciably raise cholesterol levels,” Dr. Hu [chairman of nutrition and epidemiology at the Harvard T.H. Chan School of Public Health] said. “Low to moderate consumption of three or four eggs a week doesn’t appear to have a major effect on blood cholesterol unless the person has high cholesterol or Type 2 diabetes.”

He added, “In most previous studies of healthy people, moderate egg consumption was not associated with a significant increase in cardiovascular risk.” However, among 21,275 participants in the Physicians’ Health Study who were followed for more than 20 years, those who ate one or more eggs a day were more likely to develop heart failure than those who ate eggs infrequently.

“Contradictory findings among different studies are not unusual — it’s part of the scientific process,” Dr. Hu said. “In forming guidelines, you have to look at the totality of evidence rather than overreact to a single new study.”

Zachary S. Clayton, author of a comprehensive review of research on egg consumption and heart health published in Nutrition in 2017, said in an interview that giving two eggs a day for 12 weeks to healthy people didn’t raise any of their cardiovascular risk factors and “actually decreased their triglyceride levels.”

But, Dr. Clayton, a postdoctoral fellow in nutrition at the University of Colorado, Boulder, said, “It’s important to distinguish between hypo-responders and hyper-responders to dietary cholesterol. If someone is a hyper-responder, eating two eggs a day would increase the risk of cardiovascular disease.”

For the full commentary see:

Jane E. Brody. “Cracking the Code on Eggs and Your Diet.” The New York Times (Tuesday, April 23, 2019 [sic]): D5.

(Note: bracketed words quoted from earlier in the commentary.)

(Note: the online version of the commentary has the date April 22, 2019 [sic], and has the title “Should You Be Eating Eggs?”)

Clayton’s co-authored academic review article on the effects of egg consumption, mentioned above, is:

Clayton, Zachary S., Elizabeth Fusco, and Mark Kern. “Egg Consumption and Heart Health: A Review.” Nutrition 37 (May 2017): 79-85.

The Ingenuity of Scientists and Entrepreneurs Can Find New Uses for the Previously Useless or Underutilitized

(p. D2) It may be unpleasant to contemplate the ultimate fate of all the material from your own body that you flush down the pipes. But it’s time we talk about biosolids — the disinfected leftovers from the water treatment process.

This sandy material contains nutrient-rich organic content that’s good for agriculture. But it also makes nice bricks, according to Abbas Mohajerani, a civil engineer at Royal Melbourne Institute of Technology University in Australia. He’s talking about the kind we use for building.

“Biosolids bricks look the same, smell the same and have similar physical and mechanical properties as normal fired clay bricks,” he said.

For the full story see:

JoAnna Klein. “Ultimate Recycling: When You Flushed The Toilet, They Made A Few Bricks.” The New York Times (Tuesday, February 5, 2019 [sic]): D2.

(Note: the online version of the story has the date Jan. 31, 2019 [sic], and has the title “You Flushed the Toilet. They Made Some Bricks.” Where the versions differ, the passages quoted above are from the more detailed online version.)

Mohajerani’s co-authored academic article proposing that human manure can usefully be turned into bricks for buildings is:

Mohajerani, Abbas, Aruna Ukwatta, Tristan Jeffrey-Bailey, Michael Swaney, Mohtashim Ahmed, Glen Rodwell, Simon Bartolo, Nicky Eshtiaghi, and Sujeeva Setunge. “A Proposal for Recycling the World’s Unused Stockpiles of Treated Wastewater Sludge (Biosolids) in Fired-Clay Bricks.” Buildings 9, no. 1 (2019): article #14.

Bioprospecting Tweaks Venom to Cure Diseases

(p. C3) One of the earliest treatments for ailments from gout to baldness was apitherapy, the medical application of bee venom, which was used in ancient Greece, China and Egypt. The ancient Greeks associated snakes and their venoms with medicine through the god Asclepius, whose followers prescribed venoms as cures and whose staff had a snake wrapped around it—the inspiration for the well-known symbol of medicine today.

Even so, scientists have only recently started to intensively explore the healing powers of venom. “In the 1980s and ’90s, people weren’t saying, ‘We should use venoms as a drug source,’ ” says Glenn King, a biologist at the University of Queensland in Brisbane, Australia. That changed at the beginning of this century: Scientists started to look at venoms as “complex molecular libraries,” he says. The bodily mechanisms that venoms derail often turn out to be the same ones that we need to manipulate to cure deadly diseases.

. . .

Chemical engineers have taken to mining living organisms, fine-tuning their chemicals to be more potent and precise. This process, known as bioprospecting, has had increasing appeal for scientists eager to tackle incurable diseases. Bioprospecting involves selecting a species with a type of venom known to have a specific effect on the human body—say, a snake with venom that causes a steep drop in blood pressure. The scientists will adjust the level of the toxin or tweak it biochemically so that it becomes not harmful but therapeutic.

. . .

Cancer is a natural target, and treatments may be lurking not just in scorpion venom but in the venoms of bees, snakes, snails, and even mammals. A compound derived from venomous shrews concluded a Phase I trial last year. This innovative peptide blocks a calcium channel called TRPV6, which is abundant in cancer cells, starving them of an essential element needed to grow and divide.

. . .

Each venomous animal is an artisanal mixologist, crafting chemical cocktails that can contain thousands of ingredients. The wealth of potential in venoms—each with its unique recipe—is hard to overstate.

For the full commentary see:

Christie Wilcox. “The Healing Powers of Venom.” The Wall Street Journal (Saturday, July 23, 2016 [sic]): C3.

(Note: ellipses added.)

(Note: the online version of the commentary was updated July 25, 2016 [sic], and has the title “The Healing Power of Venom.”)

The commentary quoted above is related to the author’s book:

Wilcox, Christie. Venomous: How Earth’s Deadliest Creatures Mastered Biochemistry. New York: Scientific American/Farrar, Straus and Giroux, 2016.

After a Century an Important Serendipitous Health Hunch Is Pursued

All of us (you, me, dogs, and physicians) observe patterns all the time. Some of the patterns, if pursued, could make the world much better. When a physician observes a pattern, even one they cannot articulately describe or justify, they could change their practices, curing more patients, saving more lives. But they are constrained from deviating from mainstream protocols by government regulations, insurance company rules, hospital administrators, and potential lawsuits. How many serendipitous discoveries that would help us flourish are delayed a century, or even totally snuffed out?

(p. C2) . . . my eye was drawn to a new study in the New England Journal of Medicine finding that hysterosalpingography cured some cases of infertility. Hystero refers to the uterus. Salpingo, I knew, relates to the fallopian tubes that funnel eggs to the uterus. Ography relates to imaging—but how could taking a picture of reproductive organs cure anything?

Doctors use hysterosalpingography to see if there are blockages that could be causing fertility problems.

. . .

To look at blockages, technicians have to introduce a teaspoon or two of a dye that’s opaque to X-rays. How that material is introduced, it turns out, is the key to the procedure’s effect on childlessness.

. . .

Smaller studies had given the scientists an idea of what to do next. They randomly chose half of the women to get the X-ray-opaque dye dissolved in oil, while the other half got the dye in water.

. . .

In an average of three months, whether treated or not, about 40% of the women receiving the oil-based dye material became pregnant, while only 29% of the women who got the water-based dye material conceived.

Hysterosalpingography is exactly a century old this year. Luckily, some astute doctors guessed that the method of taking a picture was having an unintended fertility effect, and now research has backed this up. Such serendipity in medical progress is neatly captured by a saying of the great French biologist Louis Pasteur about the need to be ready to see the unexpected: “In the fields of observation, chance only favors the prepared mind.”

The realization that supposedly inert oil could help to fulfill some couples’ dreams has built slowly. No one knows exactly how it works.

For the full commentary see:

Melvin Konner. “Mind & Matter; Can Just Taking a Picture Help to Treat Infertility?” The Wall Street Journal (Saturday, July 29, 2017 [sic]): C2.

(Note: ellipses added.)

(Note: the online version of the commentary has the date July 26, 2017 [sic], and has the same title as the print version. The Latin words in the first quoted sentence appear in italics in the original version.)

The New England Journal of Medicine article discussed in the passages above is:

Dreyer, Kim, Joukje van Rijswijk, Velja Mijatovic, Mariëtte Goddijn, Harold R. Verhoeve, Ilse A.J. van Rooij, Annemieke Hoek, Petra Bourdrez, Annemiek W. Nap, Henrike G.M. Rijnsaardt-Lukassen, Catharina C.M. Timmerman, Mesrure Kaplan, Angelo B. Hooker, Anna P. Gijsen, Ron van Golde, Cathelijne F. van Heteren, Alexander V. Sluijmer, Jan-Peter de Bruin, Jesper M.J. Smeenk, Jacoba A.M. de Boer, Eduard Scheenjes, Annette E.J. Duijn, Alexander Mozes, Marie J. Pelinck, Maaike A.F. Traas, Machiel H.A. van Hooff, Gijsbertus A. van Unnik, Cornelia H. de Koning, Nan van Geloven, Jos W.R. Twisk, Peter G.A. Hompes, and Ben W.J. Mol. “Oil-Based or Water-Based Contrast for Hysterosalpingography in Infertile Women.” New England Journal of Medicine 376, no. 21 (May 25, 2017): 2043-52.

Formal and Tacit Knowledge Are Located in Different Parts of the Brain

Brenda Milner turned 106 on July 15, 2024.

(p. D5) At 98, Dr. Milner is not letting up in a nearly 70-year career to clarify the function of many brain regions — frontal lobes, and temporal; vision centers and tactile; the left hemisphere and the right — usually by painstakingly testing people with brain lesions, often from surgery. Her prominence long ago transcended gender, and she is impatient with those who expect her to be a social activist. It’s science first with Dr. Milner, say close colleagues, in her lab and her life.

Perched recently on a chair in her small office, resplendent in a black satin dress and gold floral pin and banked by moldering towers of old files, she volleyed questions rather than answering them. “People think because I’m 98 years old I must be emerita,” she said. “Well, not at all. I’m still nosy, you know, curious.”

. . .

Dr. Milner changed the course of brain science for good as a newly minted Ph.D. in the 1950s by identifying the specific brain organ that is crucial to memory formation.

She did so by observing the behavior of a 29-year-old Connecticut man who had recently undergone an operation to relieve severe epileptic seizures. The operation was an experiment: On a hunch, the surgeon suctioned out two trenches of tissue from the man’s brain, one from each of his medial temporal lobes, located deep below the skull about level with the ears. The seizures subsided.

But the patient, an assembly line worker named Henry Molaison, was forever altered. He could no longer form new memories.

. . .

In a landmark 1957 paper Dr. Milner wrote with Mr. Molaison’s surgeon, she concluded that the medial temporal areas — including, importantly, an organ called the hippocampus — must be critical to memory formation. That finding, though slow to sink in, would upend the accepted teaching at the time, which held that no single area was critical to supporting memory.

Dr. Milner continued to work with Mr. Molaison and later showed that his motor memory was intact: He remembered how to perform certain physical drawing tests, even if he had no memory of learning them.

The finding, reported in 1962, demonstrated that there are at least two systems in the brain for processing memory: one that is explicit and handles names, faces and experiences; and another that is implicit and incorporates skills, like riding a bike or playing a guitar.

“I clearly remember to this day my excitement, sitting there with H. M. and watching this beautiful learning curve develop right there in front of me,” Dr. Milner said. “I knew very well I was witnessing something important.”

. . .

For Dr. Milner, after a lifetime exploring the brain, the motive for the work is personal as well as professional. “I live very close; it’s a 10-minute walk up the hill,” she said. “So it gives me a good reason to come in regularly.”

For the full story see:

Benedict Carey. “At 98, ‘Still Nosy’ About the Brain.” The New York Times (Tuesday, May 16, 2017 [sic]): D5.

(Note: ellipses added.)

(Note: the online version of the story has the date May 15, 2017 [sic], and has the title “Brenda Milner, Eminent Brain Scientist, Is ‘Still Nosy’ at 98.”)

The “landmark 1957 paper” mentioned above is:

Scoville, William Beecher, and Brenda Milner. “Loss of Recent Memory after Bilateral Hippocampal Lesions.” Journal of Neurology, Neurosurgery & Psychiatry 20, no. 1 (Feb. 1957): 11-21.

Dogs Pass a Smell Test–Locating Ancient Buried Human Remains

(p. D1) On a sunny summer day in Croatia several years ago, an archaeologist and two dog handlers watched as two dogs, one after another, slowly worked their way across the rocky top of a wind-scoured ridge overlooking the Adriatic Sea.

. . .

Panda, a Belgian Malinois with a “sensitive nose,” according to her handler, Andrea Pintar, had begun exploring the circular leftovers of a tomb when she suddenly froze, her nose pointed toward a stone burial chest. This was her signal that she had located the scent of human remains.

Ms. Pintar said the hair on her arms rose. “I was skeptical, and I was like, ‘She is kidding me,’” she recalled thinking about her dog that day.

Archaeologists had found fragments of human bone and teeth in the chest, but these had been removed months earlier for analysis and radiocarbon dating. All that was left was a bit of dirt, the stone slabs of the tomb and the cracked limestone of the ridge.

. . .

(p. D6) . . . the experiment in Croatia marked the start of one of the most careful inquiries yet carried out of an unusual archaeological method. If such dogs could successfully locate the burial sites of mass executions, dating from World War II through the conflicts in the Balkans in the 1990s, might they be effective in helping archaeologists find truly ancient burials?

. . .

That “test run” was the beginning of a careful study on whether human-remains detection dogs could be an asset to archaeologists. Setting up a controlled study was difficult. Dr. Glavaš had to learn the scientific literature, such as scent theory, far outside the standard confines of archaeology; the same was true for Ms. Pintar and the field of archaeology.

. . .

“I think dogs are really capable of this, but I think it’s a logistical challenge,” said Adee Schoon, a scent-detection-animal expert from the Netherlands who was not involved in the study. “It’s not something you can replicate again and again. It’s hard to train.”

And, as Dr. Schoon noted, dogs are “great anomaly detectors.” Something as subtle as recently disturbed soil can elicit a false alert from a dog that is not rigorously trained.

Nonetheless, the team returned to the necropolis for the first controlled tests in September 2015, and again a full year later. Both times, they used all four of Ms. Pintar and Mr. Nikolić’s cadaver dogs: Panda, Mali, a third Belgian Malinois and a German shepherd. They worked them on both known and double-blind searches, in areas where nobody knew if tombs were located.

The dogs located four tombs new to the archaeologists. Dr. Glavaš had suspected that a fifth site might hold a burial chest, and the dogs’ alerts, combined with excavation, proved her suspicion correct.

In September 2019, the Journal of Archaeological Method and Theory published the results of their study: “This research has demonstrated that HRD dogs are able to detect very small amounts of specific human decomposition odor as well as to indicate to considerably older burials than previously assumed,” Dr. Glavaš and Ms. Pintar wrote.

Dr. Schoon, who researches and helps create protocols to train scent-detection animals worldwide, said the Iron Age necropolis study was nicely designed and “really controlled.”

. . .

Cadaver dogs are also helping archaeologists at some especially challenging sites. Mike Russo and Jeff Shanks, archaeologists with the National Park Service’s Southeast Archeological Center, had created at least 14 test holes near a promising site in northwest Florida that had been flattened during an earlier era of less diligent archaeology. They found nothing.

“We knew where it should be, but when we went there, there was absolutely no mound,” Mr. Russo said.

They then asked Suzi Goodhope, a longtime cadaver-dog handler in Florida, to bring her experienced detection dog, Shiraz, a Belgian Malinois, to the site in 2013. Shiraz and Ms. Goodhope worked the flat, brushy area for a long time. Then, Shiraz sat. Once.

“I was pretty skeptical,” Mr. Shanks said.

Nonetheless, the archaeologists dug. And dug. They went down nearly three feet — and there they found a human toe bone more than 1,300 years old.

Passing sniff tests

What is the future of using human-remains detection dogs as a noninvasive tool in archaeology?

Some archaeologists, forensic anthropologists, geologists, scientists — and even H.R.D. dog handlers who know how challenging the work is — say they have great potential. But challenges abound.

Although researchers are learning ever more about the canine olfactory system, they are still trying to pinpoint what volatile organic compounds in human remains are significant to trained dogs.

. . .

Detection dogs also must be trained for archaeology with more consistency. Often humans are the limiting factor. Sometimes, Dr. Schoon said, she can almost see a dog thinking, “Is that all you want me to do? I can do much more!”

For the full story see:

Cat Warren. “Sniffing Out New (Old) Digs.” The New York Times (Tuesday, May 19, 2020 [sic]): D1 & D6.

(Note: ellipses added.)

(Note: the online version of the story was updated May 25, 2020 [sic], and has the title “When Cadaver Dogs Pick Up a Scent, Archaeologists Find Where to Dig.”)

The academic article documenting that dogs are able use their hypercapable noses to smell ancient human remains is:

Glavaš, Vedrana, and Andrea Pintar. “Human Remains Detection Dogs as a New Prospecting Method in Archaeology.” Journal of Archaeological Method and Theory 26, no. 3 (Sept. 2019): 1106-24.

AI Algorithms Lack Intelligence Since They Are “Just Predicting the Next Word in a Text”

(p. B5) Yann LeCun helped give birth to today’s artificial-intelligence boom. But he thinks many experts are exaggerating its power and peril, and he wants people to know it.

. . .

On social media, in speeches and at debates, the college professor and Meta Platforms AI guru has sparred with the boosters and Cassandras who talk up generative AI’s superhuman potential, from Elon Musk to two of LeCun’s fellow pioneers, who share with him the unofficial title of “godfather” of the field. They include Geoffrey Hinton, a friend of nearly 40 years who on Tuesday was awarded a Nobel Prize in physics, and who has warned repeatedly about AI’s existential threats.

. . .

LeCun thinks AI is a powerful tool.

. . .

At the same time, he is convinced that today’s AIs aren’t, in any meaningful sense, intelligent—and that many others in the field, especially at AI startups, are ready to extrapolate its recent development in ways that he finds ridiculous.

If LeCun’s views are right, it spells trouble for some of today’s hottest startups, not to mention the tech giants pouring tens of billions of dollars into AI. Many of them are banking on the idea that today’s large language model-based AIs, like those from OpenAI, are on the near-term path to creating so-called “artificial general intelligence,” or AGI, that broadly exceeds human-level intelligence.

OpenAI’s Sam Altman last month said we could have AGI within “a few thousand days.” Elon Musk has said it could happen by 2026.

LeCun says such talk is likely premature. When a departing OpenAI researcher in May talked up the need to learn how to control ultra-intelligent AI, LeCun pounced. “It seems to me that before ‘urgently figuring out how to control AI systems much smarter than us’ we need to have the beginning of a hint of a design for a system smarter than a house cat,” he replied on X.

He likes the cat metaphor. Felines, after all, have a mental model of the physical world, persistent memory, some reasoning ability and a capacity for planning, he says. None of these qualities are present in today’s “frontier” AIs, including those made by Meta itself.

Léon Bottou, who has known LeCun since 1986, says LeCun is “stubborn in a good way”—that is, willing to listen to others’ views, but single-minded in his pursuit of what he believes is the right approach to building artificial intelligence.

Alexander Rives, a former Ph.D. student of LeCun’s who has since founded an AI startup, says his provocations are well thought out. “He has a history of really being able to see gaps in how the field is thinking about a problem, and pointing that out,” Rives says.

. . .

The large language models, or LLMs, used for ChatGPT and other bots might someday have only a small role in systems with common sense and humanlike abilities, built using an array of other techniques and algorithms.

Today’s models are really just predicting the next word in a text, he says. But they’re so good at this that they fool us. And because of their enormous memory capacity, they can seem to be reasoning, when in fact they’re merely regurgitating information they’ve already been trained on.

“We are used to the idea that people or entities that can express themselves, or manipulate language, are smart—but that’s not true,” says LeCun. “You can manipulate language and not be smart, and that’s basically what LLMs are demonstrating.”

For the full commentary see:

Christopher Mims. “Keywords: This AI Pioneer Thinks AI Is Dumber Than a Pet Cat.” The Wall Street Journal (Saturday, Oct. 12, 2024): B5.

(Note: ellipses added.)

(Note: the online version of the commentary was updated Oct. 11, 2024, and has the title “Keywords: This AI Pioneer Thinks AI Is Dumber Than a Cat.” The sentence starting with “Léon Bottou” appears in the online, but not the print, version. Where there are small differences between the versions, the passages quoted above follow the online version.)