Arquivo da tag: Ética em pesquisa

AI Is Deciphering Animal Speech. Should We Try to Talk Back? (Gizmodo)

Uncategorized, , , , , , ,

gizmodo.com – original article

As scientists use machine learning to decode the sounds of whales, dogs, and dolphins, opinions vary on how best to deploy the technology.

Isaac Schultz

May 17, 2025

Chirps, trills, growls, howls, squawks. Animals converse in all kinds of ways, yet humankind has only scratched the surface of how they communicate with each other and the rest of the living world. Our species has trained some animals—and if you ask cats, animals have trained us, too—but we’ve yet to truly crack the code on interspecies communication.

Increasingly, animal researchers are deploying artificial intelligence to accelerate our investigations of animal communication—both within species and between branches on the tree of life. As scientists chip away at the complex communication systems of animals, they move closer to understanding what creatures are saying—and maybe even how to talk back. But as we try to bridge the linguistic gap between humans and animals, some experts are raising valid concerns about whether such capabilities are appropriate—or whether we should even attempt to communicate with animals at all.

Using AI to untangle animal language

Towards the front of the pack—or should I say pod?—is Project CETI, which has used machine learning to analyze more than 8,000 sperm whale “codas”—structured click patterns recorded by the Dominica Sperm Whale Project. Researchers uncovered contextual and combinatorial structures in the whales’ clicks, naming features like “rubato” and “ornamentation” to describe how whales subtly adjust their vocalizations during conversation. These patterns helped the team create a kind of phonetic alphabet for the animals—an expressive, structured system that may not be language as we know it but reveals a level of complexity that researchers weren’t previously aware of. Project CETI is also working on ethical guidelines for the technology, a critical goal given the risks of using AI to “talk” to the animals.

Meanwhile, Google and the Wild Dolphin Project recently introduced DolphinGemma, a large language model (LLM) trained on 40 years of dolphin vocalizations. Just as ChatGPT is an LLM for human inputs—taking visual information like research papers and images and producing responses to relevant queries—DolphinGemma intakes dolphin sound data and predicts what vocalization comes next. DolphinGemma can even generate dolphin-like audio, and the researchers’ prototype two-way system, Cetacean Hearing Augmentation Telemetry (fittingly, CHAT), uses a smartphone-based interface that dolphins employ to request items like scarves or seagrass—potentially laying the groundwork for future interspecies dialogue.

“DolphinGemma is being used in the field this season to improve our real-time sound recognition in the CHAT system,” said Denise Herzing, founder and director of the Wild Dolphin Project, which spearheaded the development of DolphinGemma in collaboration with researchers at Google DeepMind, in an email to Gizmodo. “This fall we will spend time ingesting known dolphin vocalizations and let Gemma show us any repeatable patterns they find,” such as vocalizations used in courtship and mother-calf discipline.

In this way, Herzing added, the AI applications are two-fold: Researchers can use it both to explore dolphins’ natural sounds and to better understand the animals’ responses to human mimicking of dolphin sounds, which are synthetically produced by the AI CHAT system.

Expanding the animal AI toolkit

Outside the ocean, researchers are finding that human speech models can be repurposed to decode terrestrial animal signals, too. A University of Michigan-led team used Wav2Vec2—a speech recognition model trained on human voices—to identify dogs’ emotions, genders, breeds, and even individual identities based on their barks. The pre-trained human model outperformed a version trained solely on dog data, suggesting that human language model architectures could be surprisingly effective in decoding animal communication.

Of course, we need to consider the different levels of sophistication these AI models are targeting. Determining whether a dog’s bark is aggressive or playful, or whether it’s male or female—these are perhaps understandably easier for a model to determine than, say, the nuanced meaning encoded in sperm whale phonetics. Nevertheless, each study inches scientists closer to understanding how AI tools, as they currently exist, can be best applied to such an expansive field—and gives the AI a chance to train itself to become a more useful part of the researcher’s toolkit.

And even cats—often seen as aloof—appear to be more communicative than they let on. In a 2022 study out of Paris Nanterre University, cats showed clear signs of recognizing their owner’s voice, but beyond that, the felines responded more intensely when spoken to directly in “cat talk.” That suggests cats not only pay attention to what we say, but also how we say it—especially when it comes from someone they know.

Earlier this month, a pair of cuttlefish researchers found evidence that the animals have a set of four “waves,” or physical gestures, that they make to one another, as well as to human playback of cuttlefish waves. The group plans to apply an algorithm to categorize the types of waves, automatically track the creatures’ movements, and understand the contexts in which the animals express themselves more rapidly.

Private companies (such as Google) are also getting in on the act. Last week, China’s largest search engine, Baidu, filed a patent with the country’s IP administration proposing to translate animal (specifically cat) vocalizations into human language. The quick and dirty on the tech is that it would intake a trove of data from your kitty, and then use an AI model to analyze the data, determine the animal’s emotional state, and output the apparent human language message your pet was trying to convey.

A universal translator for animals?

Together, these studies represent a major shift in how scientists are approaching animal communication. Rather than starting from scratch, research teams are building tools and models designed for humans—and making advances that would have taken much longer otherwise. The end goal could (read: could) be a kind of Rosetta Stone for the animal kingdom, powered by AI.

“We’ve gotten really good at analyzing human language just in the last five years, and we’re beginning to perfect this practice of transferring models trained on one dataset and applying them to new data,” said Sara Keen, a behavioral ecologist and electrical engineer at the Earth Species Project, in a video call with Gizmodo.

The Earth Species Project plans to launch its flagship audio-language model for animal sounds, NatureLM, this year, and a demo for NatureLM-audio is already live. With input data from across the tree of life—as well as human speech, environmental sounds, and even music detection—the model aims to become a converter of human speech into animal analogues. The model “shows promising domain transfer from human speech to animal communication,” the project states, “supporting our hypothesis that shared representations in AI can help decode animal languages.”

“A big part of our work really is trying to change the way people think about our place in the world,” Keen added. “We’re making cool discoveries about animal communication, but ultimately we’re finding that other species are just as complicated and nuanced as we are. And that revelation is pretty exciting.”

The ethical dilemma

Indeed, researchers generally agree on the promise of AI-based tools for improving the collection and interpretation of animal communication data. But some feel that there’s a breakdown in communication between that scholarly familiarity and the public’s perception of how these tools can be applied.

“I think there’s currently a lot of misunderstanding in the coverage of this topic—that somehow machine learning can create this contextual knowledge out of nothing. That so long as you have thousands of hours of audio recordings, somehow some magic machine learning black box can squeeze meaning out of that,” said Christian Rutz, an expert in animal behavior and cognition and founding president of International Bio-Logging Society, in a video call with Gizmodo. “That’s not going to happen.”

“Meaning comes through the contextual annotation and this is where I think it’s really important for this field as a whole, in this period of excitement and enthusiasm, to not forget that this annotation comes from basic behavioral ecology and natural history expertise,” Rutz added. In other words, let’s not put the horse before the cart, especially since the cart—in this case—is what’s powering the horse.

But with great power…you know the cliché. Essentially, how can humans develop and apply these technologies in a way that is both scientifically illuminating and minimizes harm or disruption to its animal subjects? Experts have put forward ethical standards and guardrails for using the technologies that prioritize the welfare of creatures as we get closer to—well, wherever the technology is going.

As AI advances, conversations about animal rights will have to evolve. In the future, animals could become more active participants in those conversations—a notion that legal experts are exploring as a thought exercise, but one that could someday become reality.

“What we desperately need—apart from advancing the machine learning side—is to forge these meaningful collaborations between the machine learning experts and the animal behavior researchers,” Rutz said, “because it’s only when you put the two of us together that you stand a chance.”

There’s no shortage of communication data to feed into data-hungry AI models, from pitch-perfect prairie dog squeaks to snails’ slimy trails (yes, really). But exactly how we make use of the information we glean from these new approaches requires thorough consideration of the ethics involved in “speaking” with animals.

A recent paper on the ethical concerns of using AI to communicate with whales outlined six major problem areas. These include privacy rights, cultural and emotional harm to whales, anthropomorphism, technological solutionism (an overreliance on technology to fix problems), gender bias, and limited effectiveness for actual whale conservation. That last issue is especially urgent, given how many whale populations are already under serious threat.

It increasingly appears that we’re on the brink of learning much more about the ways animals interact with one another—indeed, pulling back the curtain on their communication could also yield insights into how they learn, socialize, and act within their environments. But there are still significant challenges to overcome, such as asking ourselves how we use the powerful technologies currently in development.

The Genetic Engineering Genie Is Out of the Bottle (Financial Times)

Uncategorized, , , ,

foreignpolicy.com

Vivek Wadhwa, September 11, 2020

An infrared microscope image shows mosquito larvae with red-glowing eyes, part of an experiment using CRISPR gene-editing technology. MediaNews Group/Orange County Register via Getty Images

Usually good for a conspiracy theory or two, U.S. President Donald Trump has suggested that the virus causing COVID-19 was either intentionally engineered or resulted from a lab accident at the Wuhan Institute of Virology in China. Its release could conceivably have involved an accident, but the pathogen isn’t the mishmash of known viruses that one would expect from something designed in a lab, as a research report in Nature Medicine conclusively lays out. “If someone were seeking to engineer a new coronavirus as a pathogen, they would have constructed it from the backbone of a virus known to cause illness,” the researchers said.

But if genetic engineering wasn’t behind this pandemic, it could very well unleash the next one. With COVID-19 bringing Western economies to their knees, all the world’s dictators now know that pathogens can be as destructive as nuclear missiles. What’s even more worrying is that it no longer takes a sprawling government lab to engineer a virus. Thanks to a technological revolution in genetic engineering, all the tools needed to create a virus have become so cheap, simple, and readily available that any rogue scientist or college-age biohacker can use them, creating an even greater threat. Experiments that could once only have been carried out behind the protected walls of government and corporate labs can now practically be done on the kitchen table with equipment found on Amazon. Genetic engineering—with all its potential for good and bad—has become democratized.

To design a virus, a bio researcher’s first step is to obtain the genetic information of an existing pathogen—such as one of the coronaviruses that cause the common cold—which could then be altered to create something more dangerous. In the 1970s, the first genetic sequencing of a bacterium, Escherichia coli, took weeks of effort and cost millions of dollars just to determine its 5,836 base pairs, the building blocks of genetic information. Today, sequencing the 3,000,000,000 base pairs that make up the human genome, which dictates the construction and maintenance of a human being, can be done in a few hours for about $1000 in the United States. Xun Xu, the CEO of Chinese genomics research company BGI Group, told me by email that he expects to offer full human-genome sequencing in supermarkets and online for about $290 by the end of this year.

The next step in engineering a virus is to modify the genome of the existing pathogen to change its effects. One technology in particular makes it almost as easy to engineer life forms as it is to edit Microsoft Word documents. CRISPR gene editing, developed only a few years ago, deploys the same natural mechanism that bacteria use to trim pieces of genetic information from one genome and insert it into another. This mechanism, which bacteria developed over millennia to defend themselves from viruses, has been turned into a cheap, simple, and fast way to edit the DNA of any organism in the lab.

If experimenting with DNA once required years of experience, sophisticated labs, and millions of dollars, CRISPR has changed all that. To set up a CRISPR editing capability, the experimenter need only order a fragment of RNA and purchase off-the-shelf chemicals and enzymes, costing only a few dollars, on the Internet. Because it’s so cheap and easy to use, thousands of scientists all over the world are experimenting with CRISPR-based gene editing projects. Very little of this research is limited by regulations, largely because regulators don’t yet understand what has suddenly become possible.

China, with its emphasis on technological progress ahead of safety and ethics, has made the most astonishing breakthroughs. In 2014, Chinese scientists announced they had successfully produced monkeys that had been genetically modified at the embryonic stage. In April 2015, another group of researchers in China detailed the first ever-effort to edit the genes of a human embryo. While the attempt failed, it shocked the world: This wasn’t supposed to happen so soon.

In April 2016, yet another group of Chinese researchers reported having succeeded in modifying the genome of a human embryo in an effort to make it resistant to HIV infection, though the embryo was not brought to term. But then, in November 2018, Chinese researcher He Jiankui announced that he had created the first “CRISPR babies”—healthy infants whose genomes were edited before they were born. The People’s Daily gushed over the “historical breakthrough,” but after a global uproar, the Chinese authorities—who, He claims, had supported his efforts—arrested and later sentenced him to three years in prison for unethical conduct. But the Rubicon of biomedical science had been crossed.

China’s legion of rogue scientists is certainly a worry. But gene-editing technology has become so accessible that we could conceivably see teenagers experimenting with viruses. In the United States, anyone who wants to start modifying the genome in their garage can order a do-it-yourself CRISPR kit online for $169, for example. This comes with “everything you need to make precision genome edits in bacteria at home.” For $349, the same company is also offering a human engineering kit, which comes with embryonic kidney cells from a tissue culture originally taken from an aborted female human fetus. Shipment is advertised to take no longer than three days—no special couriers or ice packs needed.

Mail-order DNA fragments enabled a team at the University of Alberta, in 2017, to resurrect an extinct relative of the smallpox virus, horsepox, from scratch by stitching together the fragments. Horsepox is not known to harm humans, but experts warned that the same method could be used by scientists without much specialized knowledge to recreate smallpox—a horrific virus finally eradicated in 1980—within six months at a cost of about $100,000. Had the Canadian scientists used CRISPR, their cost would have been reduced to a fraction.

In my book, The Driver in the Driverless Car, published before the Canadian horsepox resurrection and the Chinese gene-edited babies, I warned about the dangers of gene editing, predicting we would have to make difficult choices about whether to restrict synthetic biology technologies. When used for good purposes, these technologies can help solve the problems of humanity—by quickly finding cures for diseases, for example. When used for evil, they can wreak global havoc of exactly the kind we are now fighting. That is why many people, myself included, have advocated for a moratorium on human gene editing.

But not just a moratorium: There should have been international treaties to prevent the use of CRISPR for gene editing on humans or animals. The U.S. Food and Drug Administration should have kept companies from selling DIY gene-editing kits. Governments should have placed restrictions on labs such as the University of Alberta’s. But none of this happened, nor were there any other checks and balances. It is now too late to stop the global spread of these technologies—the genie is out of the bottle.

Now, the only solution is to accelerate the good side of these technologies while building our defenses. As we are seeing with the development of vaccines for COVID-19, this is possible. In the past, vaccines took decades to create. Now, we are on track to have them within months, thanks to advances in genetic engineering. The Moderna Therapeutics and Pfizer/BioNTech vaccines, which are now in third-stage clinical trials, took only weeks to develop. It is conceivable that this could be reduced to hours once the technologies are perfected.

We can also accelerate the process of testing vaccines and treatments, which has become the slowest part of the development cycle. To test greater numbers of potential cancer drugs more quickly, for example, labs all over the world are creating three-dimensional cell cultures called “patient-derived organoids” from tumor biopsies. The leading company in this field, SEngine Precision Medicine, is able to test more than 100 drugs on these organoids, removing the need to use human subjects as the guinea pigs. Researchers at Harvard University’s Wyss Institute announced in January 2020 that they had developed the first human “organ-on-a-chip” model of the lung that accurately replicates a human organ’s physiology and pathophysiology. Engineers at the Massachusetts Institute of Technology have been developing a microfluidic platform that connects engineered tissues from up to 10 organs, allowing the replication of human-organ interactions for weeks at a time in order to measure the effects of drugs on different parts of the body. Many more such systems are being developed that could accelerate testing and treatment. All these technologies will greatly strengthen our biodefense.

There really is no turning back to correct the mistakes of the past. The genie cannot be put back in the bottle. We must treat the coronavirus pandemic as a full dress rehearsal of what is to come—unfortunately, that includes not only viruses that erupt from nature, but also those that will be deliberately engineered by humans. We must learn very quickly to build the same types of types of defenses that our computers have against their invaders. The good that might ultimately come from this is the cure for all disease. The bad is just about too terrible to think about.

Face masks and the coronavirus: Masks probably slow the spread of covid-19 (The Economist)

Uncategorized, , , , ,

But wearing one is mainly an act of altruism

Science & technology – May 28th 2020 edition

Editor’s note: Some of our covid-19 coverage is free for readers of The Economist Today, our daily newsletter. For more stories and our pandemic tracker, see our coronavirus hub

“THIS IS A, I would say, senseless dividing line,” said Doug Burgum, governor of North Dakota, his voice catching as he talked of the rows that have broken out in his state over the wearing of face-coverings. There are similar spats elsewhere in America, for masks have become the latest aspect of the culture war that has emerged there over how to deal with covid-19. Some shops refuse entry to maskwearers and Mike DeWine, the governor of Ohio, has rescinded an order requiring people to wear them, saying that he “went too far”.

Elsewhere in the world, by contrast, there is increasing acceptance that mask-wearing is a good thing. On May 5th, for example, the Royal Society, Britain’s top science academy, concluded that masks “could be an important tool for managing community transmission”. This is not so much because they protect the wearer—the normal reason people may put them on in times of pestilence—but rather because they stop the wearer infecting others.

In this context covid-19’s particular peculiarity—that people who test positive for it often do not have symptoms—is important. Research published last month in Nature Medicine, by Xi He of Guangzhou Medical University and Eric Lau of Hong Kong University, suggests that 44% of cases are caused by transmission from people without symptoms at the time of transmission.

Taking cover

Those who do have symptoms should not, of course, be out and about at all. In their case masks are irrelevant. But to break the chain, it behoves even the symptomless to assume that they might be infected. Covid-19 is transmitted, above all, by virus-laden droplets of spit. Experiments show that face-coverings as simple as tea-towels are effective. One study found that a tea-towel worn around the face captured 60% of droplets. At 75%, a surgical mask did better, but not overwhelmingly so.

Governments are beginning to take this on board. As part of the loosening their lockdown, the Dutch are required to wear face-coverings on public transport—but not ones of medical grade, which should be reserved for professionals. This encourages people to make their own.

Neither laboratory studies nor the data on asymptomatic transmission provide watertight evidence of the efficacy of masks. That would need randomised controlled trials, in which one group wore masks and the other did not. This would be ethically tricky, since it might condemn one of the groups to a higher death rate. Hamsters, which are susceptible to covid-19, are the next best thing to people. So researchers at Hong Kong University put cages of healthy hamsters next to cages of infected ones, with a fan in between drawing air from the infected to the healthy cage. They sometimes also placed a stretched-out face mask in the air stream. With no interposed mask, two-thirds of the healthy animals were infected within a week. With a mask interposed close to the healthy hamsters (the equivalent of a healthy person wearing a mask), one-third were. With the mask close to the infected hamsters, only a sixth were.

Although scientists cannot experiment on human beings deliberately, some wonder if the world is now carrying out a natural experiment that tests the value of mask-wearing. In many East Asian countries it was common practice to sport masks, even before covid-19, to protect against respiratory diseases and pollution. A lot of people in these places therefore took immediately to wearing masks when the epidemic started. Countries that adopted masks early on did not, by and large, shut their economies down. Yet they suppressed the disease more effectively than those that locked down but did not wear masks.

There is a correlation between mask-wearing and rapid suppression of covid-19. According to Patricia Greenhalgh, professor of primary health care sciences at Oxford University, “there is not a single country in which mask wearing was introduced early and with high compliance, where the disease wasn’t quickly brought under control.” Sceptics point out that this does not prove masks work, since countries in which they are widely worn also tend to be those which have been threatened by epidemics in the past, and therefore have well-established systems of testing and contact tracing.

In the West nobody normally wears a mask, though the practice is spreading. Universal masking started in the Czech Republic after Petr Ludwig, a Czech YouTube star, posted a video on March 14th recommending the practice, and it went viral. Other social-media influencers posted pictures of themselves wearing masks. “Mask trees”, where people would hang home-made masks for others to use, sprang up on street corners. By March 19th masks were mandatory in the country. Slovakia and Slovenia followed swiftly.

The World Health Organisation has not advocated widespread mask-wearing, and has received some criticism for this. Jeremy Howard, a research scientist at the University of San Francisco and co-founder of Masks4all, a charity, says “they did a good job of recommending handwashing and social distancing, but they have been slow on masks.”

In light of all this, regulations requiring people to wear masks have spread, as an increasing number of governments view the evidence as strong enough to warrant compulsion. India now requires them to be worn in crowded public spaces, as do France, Germany, Italy and Spain. In most of the world, people either wear them in such spaces without being told to, or are required to by their governments.

Among big countries, Britain and America are outliers. In Britain the government advises people to wear masks, but to little effect. On the London Underground around a third of travellers do so. On the Paris metro where people risk a €135 fine if they fail to cover their faces, everybody does. In America the Centres for Disease Control and Prevention, which previously recommended mask-wearing only for health workers, changed its mind in early April. It now recommends that everybody should wear them in places where it is hard for people to stay far enough apart. Several states have passed regulations along those lines, as has New York City. But, as Governor Burgum noted, the rows go on.■

This article appeared in the Science & technology section of the print edition under the headline “We interrupt this transmission…”