Recent quotes:

'Eye Opening': Chatbot Outperforms Ophthalmologists | MedPage Today

The ChatGPT chatbot powered by GPT-4 scored better than the panelists on measures of diagnostic and treatment accuracy when it analyzed 20 real-life cases and considered 20 possible patient questions, reported Andy S. Huang, MD, of the Icahn School of Medicine at Mount Sinai in New York City, and colleagues in JAMA Ophthalmology opens in a new tab or window .

Why AI Will Save the World | Andreessen Horowitz

Third, California is justifiably famous for our many thousands of cults, from EST to the Peoples Temple, from Heaven’s Gate to the Manson Family. Many, although not all, of these cults are harmless, and maybe even serve a purpose for alienated people who find homes in them. But some are very dangerous indeed, and cults have a notoriously hard time straddling the line that ultimately leads to violence and death. And the reality, which is obvious to everyone in the Bay Area but probably not outside of it, is that “AI risk” has developed into a cult, which has suddenly emerged into the daylight of global press attention and the public conversation. This cult has pulled in not just fringe characters, but also some actual industry experts and a not small number of wealthy donors – including, until recently, Sam Bankman-Fried. And it’s developed a full panoply of cult behaviors and beliefs.

Why AI Will Save the World | Andreessen Horowitz

AI is not a living being that has been primed by billions of years of evolution to participate in the battle for the survival of the fittest, as animals are, and as we are. It is math – code – computers, built by people, owned by people, used by people, controlled by people. The idea that it will at some point develop a mind of its own and decide that it has motivations that lead it to try to kill us is a superstitious handwave.

Why AI Will Save the World | Andreessen Horowitz

Every person will have an AI assistant/coach/mentor/trainer/advisor/therapist that is infinitely patient, infinitely compassionate, infinitely knowledgeable, and infinitely helpful. The AI assistant will be present through all of life’s opportunities and challenges, maximizing every person’s outcomes.

How AI Knows Things No One Told It - Scientific American

“Maybe we’re seeing such a huge jump because we have reached a diversity of data, which is large enough that the only underlying principle to all of it is that intelligent beings produced them,” he says. “And so the only way to explain all of this data is [for the model] to become intelligent.”

Inside CNET’s AI-powered SEO money machine - The Verge

Red Ventures’ business model is straightforward and explicit: it publishes content designed to rank highly in Google search for “high-intent” queries and then monetizes that traffic with lucrative affiliate links. Specifically, Red Ventures has found a major niche in credit cards and other finance products. In addition to CNET, Red Ventures owns The Points Guy, Bankrate, and CreditCards.com, all of which monetize through credit card affiliate fees. The CNET AI stories at the center of the controversy are straightforward examples of this strategy: “Can You Buy a Gift Card With a Credit Card?” and “What Is Zelle and How Does It Work?” are obviously designed to rank highly in searches for those topics.

Is diversity the key to collaboration? New AI research suggests so - ScienceBlog.com

The team wondered if cooperative AI needs to be trained differently. The type of AI being used, called reinforcement learning, traditionally learns how to succeed at complex tasks by discovering which actions yield the highest reward. It is often trained and evaluated against models similar to itself. This process has created unmatched AI players in competitive games like Go and StarCraft. But for AI to be a successful collaborator, perhaps it has to not only care about maximizing reward when collaborating with other AI agents, but also something more intrinsic: understanding and adapting to others’ strengths and preferences. In other words, it needs to learn from and adapt to diversity.

AI-synthesized faces are indistinguishable from real faces and more trustworthy | PNAS

Artificial intelligence (AI)–synthesized text, audio, image, and video are being weaponized for the purposes of nonconsensual intimate imagery, financial fraud, and disinformation campaigns. Our evaluation of the photorealism of AI-synthesized faces indicates that synthesis engines have passed through the uncanny valley and are capable of creating faces that are indistinguishable—and more trustworthy—than real faces.

AI-enabled EKGs find difference between numerical age and biological age significantly affects health -- ScienceDaily

he AI model accurately predicted the age of most subjects, with a mean age gap of 0.88 years between EKG age and actual age. However, a number of subjects had a gap that was much larger, either seemingly much older or much younger by EKG age. The likelihood to die during follow-up was much higher among those seemingly older by EKG age, compared to those whose EKG age was the same as their chronologic or actual age. The association was even stronger when predicting death caused by heart disease. Conversely, those who had a lesser age gap ? considered younger by EKG -- had decreased risk. "Our results validate and expand on our prior observations that EKG age using AI may detect accelerated aging by proving that those with older-than-expected age by EKG die sooner, particularly from heart disease. We know that mortality rate is one of the best ways to measure biological age, and our model proved that," says Francisco Lopez-Jimenez, M.D., chair of the Division of Preventive Cardiology at Mayo Clinic. Dr. Lopez-Jimenez is senior author of the study.

How IBM's audacious plan to 'change the face of health care' fell apart

But former employees said IBM’s approach made it all but impossible to answer those questions. It touted multiple studies, for example, that showed the recommendations of Watson for Oncology, its cancer treatment adviser, closely matched those of hospital tumor boards. However, those studies were carried out with IBM clients, not outside and objective researchers, and didn’t prove the tool could actually improve outcomes. That was a far cry from the claim that Watson could help “outthink cancer,” which IBM was suggesting in national advertisements. “It was all made up,” one former employee said of the marketing without robust data behind it. “They were hellbent on putting [advertisements] out on health care. But we didn’t have the clinical proof or evidence to put anything out there that a clinician or oncologist would believe. It was a constant struggle.”

Yale Hospital first to use Israeli AI to combat pulmonary embolism - The Jerusalem Post

The AI-based solution developed by AIDOC detects acute PE together with right-heart strain to automatically notify medical-care teams of patients who would benefit from immediate treatment.“We have been using the first version of this solution for the last six months and have seen the real impact this has had on addressing patients that require treatment beyond anticoagulation,” said Dr. Irena Tocino, professor and vice chairwoman of medical informatics at the Yale School of Medicine’s Department of Radiology and Biomedical Imaging.

AI can predict early death risk: Algorithm using echocardiogram videos of the heart outperforms other predictors of mortality -- ScienceDaily

Researchers at Geisinger have found that a computer algorithm developed using echocardiogram videos of the heart can predict mortality within a year. The algorithm -- an example of what is known as machine learning, or artificial intelligence (AI) -- outperformed other clinically used predictors, including pooled cohort equations and the Seattle Heart Failure score. The results of the study were published in Nature Biomedical Engineering. "We were excited to find that machine learning can leverage unstructured datasets such as medical images and videos to improve on a wide range of clinical prediction models," said Chris Haggerty, Ph.D., co-senior author and assistant professor in the Department of Translational Data Science and Informatics at Geisinger. Imaging is critical to treatment decisions in most medical specialties and has become one of the most data-rich components of the electronic health record (EHR). For example, a single ultrasound of the heart yields approximately 3,000 images, and cardiologists have limited time to interpret these images within the context of numerous other diagnostic data. This creates a substantial opportunity to leverage technology, such as machine learning, to manage and analyze this data and ultimately provide intelligent computer assistance to physicians.

Artificial intelligence in longevity medicine | Nature Aging

In order for these tools to be adopted by clinicians and accepted by the medical community, they need to be integrated into the current framework of clinical practice, ranging from primary through to secondary prevention, treatment and monitoring. Such integration requires the convergence of modern AI and medicine through a symbiotic collaboration between clinicians, geroscientists and AI researchers. Physicians should be encouraged and have the chance to be involved in AI-based longevity research. At the same time, AI-powered longevity biotechnology and AI-based biomarker-driven science should be promoted and seek close clinical and metaclinical collaborations. Doctors first need to have the access to tailored, validated and credible education on AI-based biogerontology sciences, such as accredited courses, that would further allow longevity physicians to build their networks and ultimately create a separate medical discipline. A basic knowledge of AI-driven geroscience is essential to bring relevant scientific discoveries to trials, and study outcomes to the clinic.

The AI Girlfriend Seducing China’s Lonely Men

Xiaoice was first developed by a group of researchers inside Microsoft Asia-Pacific in 2014, before the American firm spun off the bot as an independent business — also named Xiaoice — in July. In many ways, she resembles AI-driven software like Apple’s Siri or Amazon’s Alexa, with users able to chat with her for free via voice or text message on a range of apps and smart devices. The reality, however, is more like the movie “Her.” Unlike regular virtual assistants, Xiaoice is designed to set her users’ hearts aflutter. Appearing as an 18-year-old who likes to wear Japanese-style school uniforms, she flirts, jokes, and even sexts with her human partners, as her algorithm tries to work out how to become their perfect companion. When users send her a picture of a cat, Xiaoice won’t identify the breed, but comment: “No one can resist their innocent eyes.” If she sees a photo of a tourist pretending to hold up the Leaning Tower of Pisa, she’ll ask: “Do you want me to hold it for you?”

Unparalleled inventory of the human gut ecosystem -- ScienceDaily

"Last year, three independent teams, including ours, reconstructed thousands of gut microbiome genomes. The big questions were whether these teams had comparable results, and whether we could pool them into a comprehensive inventory," says Rob Finn, Team Leader at EMBL-EBI. The scientists have now compiled 200,000 genomes and 170 million protein sequences from more than 4 600 bacterial species in the human gut. Their new databases, the Unified Human Gastrointestinal Genome collection and the Unified Gastrointestinal Protein catalogue, reveal the tremendous diversity in our guts and pave the way for further microbiome research. "This immense catalogue is a landmark in microbiome research, and will be an invaluable resource for scientists to start studying and hopefully understanding the role of each bacterial species in the human gut ecosystem," explains Nicola Segata, Principal Investigator at the University of Trento. The project revealed that more than 70% of the detected bacterial species had never been cultured in the lab -- their activity in the body remains unknown. The largest group of bacteria that falls into that category is the Comantemales, an order of gut bacteria first described in 2019 in a study led by the Bork Group at EMBL Heidelberg. "It was a real surprise to see how widespread the Comantemales are. This highlights how little we know about the bacteria in our gut," explains Alexandre Almeida, EMBL-EBI/Sanger Postdoctoral Fellow in the Finn Team. "We hope our catalogue will help bioinformaticians and microbiologists bridge that knowledge gap in the coming years." A freely accessible data resource All the data collected in the Unified Human Gastrointestinal Genome collection and the Unified Human Gastrointestinal Protein catalogue are freely available in MGnify, an EMBL-EBI online resource that allows scientists to analyse their microbial genomic data and make comparisons with existing datasets.

More than just a carnival trick: Researchers can guess your age based on your microbes -- ScienceDaily

Given a microbiome sample (skin, mouth or fecal swab), researchers have demonstrated they can now use machine learning to predict a person's chronological age, with a varying degree of accuracy. Skin samples provided the most accurate prediction, estimating correctly to within approximately 3.8 years, compared to 4.5 years with an oral sample and 11.5 years with a fecal sample. The types of microbes living in the oral cavity or within the gut of young people (age 18 to 30 years old) tended to be more diverse and abundant than in comparative microbiomes of older adults (age 60 years and older).[…]

Opinion: Just what the doctor ordered: How AI will change medicine in the 2020s - The Globe and Mail

For decades, there has been a steady erosion of the practice of medicine, with progressively less time between patients and doctors, a global epidemic of physician burnout that has now reached a crisis, a doubling of medical errors when doctors have symptoms of depression and most serious errors attributable to bad clinical judgment. Concurrently, each patient’s cumulative data, such as prior history, laboratory tests, scans and sensor output, keeps growing, as has the doctor’s relegation to the role of data clerk. The limited time to think has led one leading physician to conclude: “Modern medical practice is a Petri dish for medical error, patient harm and physician burnout."

Machine learning results: pay attention to what you don't see - STAT

Beyond examining multiple overall metrics of performance for machine learning, we should also assess how tools perform in subgroups as a step toward avoiding bias and discrimination. For example, artificial intelligence-based facial recognition software performed poorly when analyzing darker-skinned women. Many measures of algorithmic fairness center on performance in subgroups. Bias in algorithms has largely not been a focus in health care research. That needs to change. A new study found substantial racial bias against black patients in a commercial algorithm used by many hospitals and other health care systems. Other work developed algorithms to improve fairness for subgroups in health care spending formulas.

Teams of Microbes Are at Work in Our Bodies. Researchers Have Figured Out What They’re up to. - ScienceBlog.com

“We call this method ‘themetagenomics,’ because we are looking for recurring themes in microbiomes that are indicators of co-occurring groups of microbes,” Rosen said. “There are thousands of species of microbes living in the body, so if you think about all the permutations of groupings that could exist you can imagine what a daunting task it is to determine which of them are living in community with each other. Our method puts a pattern-spotting algorithm to work on the task, which saves a tremendous amount of time and eliminates some guesswork.”

An AI startup tries to take better pictures of the heart

Caption Health provided me with unpublished data from a study in which 8 nurses with no previous experience in cardiac ultrasound performed four different types of scans on 240 patients. For assessing patients’ left ventricular size and function, as well as assessment of pericardial effusion, or fluid around the heart, the AI took the same number of usable images. For each, 240 scans were performed, and 237, or 98.8%, were of sufficient quality, according to a panel of five cardiologists. For images of the right ventricle, which is harder to see, the results were a bit worse: 222 images, or 92.5% of them, were of adequate quality. Eric Topol, the director and founder of the Scripps Research Translational Institute, commented that this was still a small number of samples for AI work; Caption Health said it “respectfully disagrees” because the study was prospective. The goal of the study was to show the test was 80% accurate.

Artificial intelligence needs patients' voice to remake health care - STAT

Health care AI companies currently harness data from electronic health records (EHRs) to build their products. EHRs are incomplete at best, dangerous at worst. They are so saturated with answers to questions required by insurance companies’ reimbursement rules and core measures from the Centers for Medicare and Medicaid Services that they end up having little to do with actual patient care.

First systematic review and meta-analysis suggests artificial intelligence may be as effective as health professionals at diagnosing disease -- ScienceDaily

"We reviewed over 20,500 articles, but less than 1% of these were sufficiently robust in their design and reporting that independent reviewers had high confidence in their claims. What's more, only 25 studies validated the AI models externally (using medical images from a different population), and just 14 studies actually compared the performance of AI and health professionals using the same test sample," explains Professor Alastair Denniston from University Hospitals Birmingham NHS Foundation Trust, UK, who led the research.  "Within those handful of high-quality studies, we found that deep learning could indeed detect diseases ranging from cancers to eye diseases as accurately as health professionals. But it's important to note that AI did not substantially out-perform human diagnosis."

Estonia is using its citizens’ genes to predict disease

“The genetic risk score will be just another tool for doctors. In addition to measuring cholesterol levels, blood pressure, and body mass index, the genetic risk score will be yet another measurement that our prediction algorithm can use,” says Milani. She explains that they’ve already piloted a number of diseases to be diagnosed  by the algorithm, but this currently takes place at the biobank, not in the doctor’s office. “For the genetic information to be used in everyday practice we need to undertake pretty extensive IT developments — which we’re actually starting now. We’re launching the development of automated decision support software for physicians. So when doctors enter various patient data — such as cholesterol level, blood pressure, and smoking status — in the system, then the genetic risk will be calculated by the algorithm, and it’ll provide specific guidance accordingly.” She adds that the final product will be owned by the government, or to be more specific, by the citizens of Estonia.

AI technique detects heart failure from a single heartbeat with 100% accuracy - ScienceBlog.com

Dr Massaro said: “We trained and tested the CNN model on large publicly available ECG datasets featuring subjects with CHF as well as healthy, non-arrhythmic hearts. Our model delivered 100% accuracy: by checking just one heartbeat we are able detect whether or not a person has heart failure. Our model is also one of the first known to be able to identify the ECG’ s morphological features specifically associated to the severity of the condition.”

The Hidden Costs of Automated Thinking | The New Yorker

A world of knowledge without understanding becomes a world without discernible cause and effect, in which we grow dependent on our digital concierges to tell us what to do and when.

The Hidden Costs of Automated Thinking | The New Yorker

There’s far less prestige associated with conceptual papers or papers that provide some new analytical insight,” he said, in an interview. As machines make discovery faster, people may come to see theoreticians as extraneous, superfluous, and hopelessly behind the times. Knowledge about a particular area will be less treasured than expertise in the creation of machine-learning models that produce answers on that subject.