Can AI solve medical mysteries? It’s worth finding out

Curated on Posted on by Stefaan Verhulst

Article by Bina Venkataraman: “Since finding a primary care doctor these days takes longer than finding a decent used car, it’s little wonder that people turn to Google to probe what ails them. Be skeptical of anyone who claims to be above it. Though I was raised by scientists and routinely read medical journals out of curiosity, in recent months I’ve gone online to investigate causes of a lingering cough, ask how to get rid of wrist pain and look for ways to treat a bad jellyfish sting. (No, you don’t ask someone to urinate on it.)

Dabbling in self-diagnosis is becoming more robust now that people can go to chatbots powered by large language models scouring mountains of medical literature to yield answers in plain language — in multiple languages. What might an elevated inflammation marker in a blood test combined with pain in your left heel mean? The AI chatbots have some ideas. And researchers are finding that, when fed the right information, they’re often not wrong. Recently, one frustrated mother, whose son had seen 17 doctors for chronic pain, put his medical information into ChatGPT, which accurately suggested tethered cord syndrome — which then led a Michigan neurosurgeon to confirm an underlying diagnosis of spina bifida that could be helped by an operation.

The promise of this trend is that patients might be able to get to the bottom of mysterious ailments and undiagnosed illnesses by generating possible causes for their doctors to consider. The peril is that people may come to rely too much on these tools, trusting them more than medical professionals, and that our AI friends will fabricate medical evidence that misleads people about, say, the safety of vaccines or the benefits of bogus treatments. A question looming over the future of medicine is how to get the best of what artificial intelligence can offer us without the worst.

It’s in the diagnosis of rare diseases — which afflict an estimated 30 million Americans and hundreds of millions of people worldwide — that AI could almost certainly make things better. “Doctors are very good at dealing with the common things,” says Isaac Kohane, chair of the department of biomedical informatics at Harvard Medical School. “But there are literally thousands of diseases that most clinicians will have never seen or even have ever heard of.”..(More)”.

Interwoven Realms: Data Governance as the Bedrock for AI Governance

Curated on Posted on by Stefaan Verhulst

Essay by Stefaan G. Verhulst and Friederike Schüür: “In a world increasingly captivated by the opportunities and challenges of artificial intelligence (AI), there has been a surge in the establishment of committees, forums, and summits dedicated to AI governance. These platforms, while crucial, often overlook a fundamental pillar: the role of data governance. As we navigate through a plethora of discussions and debates on AI, this essay seeks to illuminate the often-ignored yet indispensable link between AI governance and robust data governance.

The current focus on AI governance, with its myriad ethical, legal, and societal implications, tends to sidestep the fact that effective AI governance is, at its core, reliant on the principles and practices of data governance. This oversight has resulted in a fragmented approach, leading to a scenario where the data and AI communities operate in isolation, often unaware of the essential synergy that should exist between them.

This essay delves into the intertwined nature of these two realms. It provides six reasons why AI governance is unattainable without a comprehensive and robust framework of data governance. In addressing this intersection, the essay aims to shed light on the necessity of integrating data governance more prominently into the conversation on AI, thereby fostering a more cohesive and effective approach to the governance of this transformative technology.

Six reasons why Data Governance is the bedrock for AI Governance...(More)”.

New York City Takes Aim at AI

Curated on Posted on by Stefaan Verhulst

Article by Samuel Greengard: “As concerns over artificial intelligence (AI) grow and angst about its potential impact increase, political leaders and government agencies are taking notice. In November, U.S. president Joe Biden issued an executive order designed to build guardrails around the technology. Meanwhile, the European Union (EU) is currently developing a legal framework around responsible AI.

Yet, what is often overlooked about artificial intelligence is that it’s more likely to impact people on a local level. AI touches housing, transportation, healthcare, policing and numerous other areas relating to business and daily life. It increasingly affects citizens, government employees, and businesses in both obvious and unintended ways.

One city attempting to position itself at the vanguard of AI is New York. In October 2023, New York City announced a blueprint for developing, managing, and using the technology responsibly. The New York City Artificial Intelligence Action Plan—the first of its kind in the U.S.—is designed to help officials and the public navigate the AI space.

“It’s a fairly comprehensive plan that addresses both the use of AI within city government and the responsible use of the technology,” says Clifford S. Stein, Wai T. Chang Professor of Industrial Engineering and Operations Research and Interim Director of the Data Science Institute at Columbia University.

Adds Stefaan Verhulst, co-founder and chief research and development officer at The GovLab and Senior Fellow at the Center for Democracy and Technology (CDT), “AI localism focuses on the idea that cities are where most of the action is in regard to AI.”…(More)”.

Updates to the OECD’s definition of an AI system explained

Curated on Posted on by Stefaan Verhulst

Article by Stuart Russell: “Obtaining consensus on a definition for an AI system in any sector or group of experts has proven to be a complicated task. However, if governments are to legislate and regulate AI, they need a definition to act as a foundation. Given the global nature of AI, if all governments can agree on the same definition, it allows for interoperability across jurisdictions.

Recently, OECD member countries approved a revised version of the Organisation’s definition of an AI system. We published the definition on LinkedIn, which, to our surprise, received an unprecedented number of comments.

We want to respond better to the interest our community has shown in the definition with a short explanation of the rationale behind the update and the definition itself. Later this year, we can share even more details once they are finalised.

How OECD countries updated the definition

Here are the revisions to the current text of the definition of “AI System” in detail, with additions set out in bold and subtractions in strikethrough):

An AI system is a machine-based system that can, for a given set of human-defined explicit or implicit objectives, infers, from the input it receives, how to generate outputs such as makes predictions, content, recommendations, or decisions that can influenceing physical real or virtual environments. Different AI systems are designed to operate with varying in their levels of autonomy and adaptiveness after deployment…(More)”

Elon Musk is now taking applications for data to study X — but only EU risk researchers need apply…

Curated on Posted on by Stefaan Verhulst

Article by Natasha Lomas: “Lawmakers take note: Elon Musk-owned X appears to have quietly complied with a hard legal requirement in the European Union that requires larger platforms (aka VLOPs) to provide researchers with data access in order to study systemic risks arising from use of their services — risks such as disinformation, child safety issues, gender-based violence and mental heath concerns.

X (or Twitter as it was still called at the time) was designated a VLOP under the EU’s Digital Services Act (DSA) back in April after the bloc’s regulators confirmed it meets their criteria for an extra layer of rules to kick in that are intended to drive algorithmic accountability via applying transparency measures on larger platforms.

Researchers intending to study systemic risks in the EU now appear to at least be able to apply for access to study X’s data by accessing a web form through a button which appears at the bottom of this page on its developer platform. (Note researchers can be based in the EU but don’t have to be to meet the criteria; they just need to intend to study systemic risks in the EU.)…(More)”.

Internet use does not appear to harm mental health, study finds

Curated on Posted on by Stefaan Verhulst

Tim Bradshaw at the Financial Times: “A study of more than 2mn people’s internet use found no “smoking gun” for widespread harm to mental health from online activities such as browsing social media and gaming, despite widely claimed concerns that mobile apps can cause depression and anxiety.

Researchers at the Oxford Internet Institute, who said their study was the largest of its kind, said they found no evidence to support “popular ideas that certain groups are more at risk” from the technology.

However, Andrew Przybylski, professor at the institute — part of the University of Oxford — said that the data necessary to establish a causal connection was “absent” without more co-operation from tech companies. If apps do harm mental health, only the companies that build them have the user data that could prove it, he said.

“The best data we have available suggests that there is not a global link between these factors,” said Przybylski, who carried out the study with Matti Vuorre, a professor at Tilburg University. Because the “stakes are so high” if online activity really did lead to mental health problems, any regulation aimed at addressing it should be based on much more “conclusive” evidence, he added.

“Global Well-Being and Mental Health in the Internet Age” was published in the journal Clinical Psychological Science on Tuesday. 

In their paper, Przybylski and Vuorre studied data on psychological wellbeing from 2.4mn people aged 15 to 89 in 168 countries between 2005 and 2022, which they contrasted with industry data about growth in internet subscriptions over that time, as well as tracking associations between mental health and internet adoption in 202 countries from 2000-19.

“Our results do not provide evidence supporting the view that the internet and technologies enabled by it, such as smartphones with internet access, are actively promoting or harming either wellbeing or mental health globally,” they concluded. While there was “some evidence” of greater associations between mental health problems and technology among younger people, these “appeared small in magnitude”, they added.

The report contrasts with a growing body of research in recent years that has connected the beginning of the smartphone era, around 2010, with growing rates of anxiety and depression, especially among teenage girls. Studies have suggested that reducing time on social media can benefit mental health, while those who spend the longest online are at greater risk of harm…(More)”.

The Oligopoly’s Shift to Open Access. How the Big Five Academic Publishers Profit from Article Processing Charges 

Curated on Posted on by Stefaan Verhulst

Paper by Leigh-Ann Butler et al: “This study aims to estimate the total amount of article processing charges (APCs) paid to publish open access (OA) in journals controlled by the five large commercial publishers Elsevier, Sage, Springer-Nature, Taylor & Francis and Wiley between 2015 and 2018. Using publication data from WoS, OA status from Unpaywall and annual APC prices from open datasets and historical fees retrieved via the Internet Archive Wayback Machine, we estimate that globally authors paid $1.06 billion in publication fees to these publishers from 2015–2018. Revenue from gold OA amounted to $612.5 million, while $448.3 million was obtained for publishing OA in hybrid journals. Among the five publishers, Springer-Nature made the most revenue from OA ($589.7 million), followed by Elsevier ($221.4 million), Wiley ($114.3 million), Taylor & Francis ($76.8 million) and Sage ($31.6 million). With Elsevier and Wiley making most of APC revenue from hybrid fees and others focusing on gold, different OA strategies could be observed between publishers…(More)”.This study aims to estimate the total amount of article processing charges (APCs) paid to publish open access (OA) in journals controlled by the five large commercial publishers Elsevier, Sage, Springer-Nature, Taylor & Francis and Wiley between 2015 and 2018. Using publication data from WoS, OA status from Unpaywall and annual APC prices from open datasets and historical fees retrieved via the Internet Archive Wayback Machine, we estimate that globally authors paid $1.06 billion in publication fees to these publishers from 2015–2018. Revenue from gold OA amounted to $612.5 million, while $448.3 million was obtained for publishing OA in hybrid journals. Among the five publishers, Springer-Nature made the most revenue from OA ($589.7 million), followed by Elsevier ($221.4 million), Wiley ($114.3 million), Taylor & Francis ($76.8 million) and Sage ($31.6 million). With Elsevier and Wiley making most of APC revenue from hybrid fees and others focusing on gold, different OA strategies could be observed between publishers.

Meta is giving researchers more access to Facebook and Instagram data

Curated on Posted on by Stefaan Verhulst

Article by Tate Ryan-Mosley: “Meta is releasing a new transparency product called the Meta Content Library and API, according to an announcement from the company today. The new tools will allow select researchers to access publicly available data on Facebook and Instagram in an effort to give a more overarching view of what’s happening on the platforms. 

The move comes as social media companies are facing public and regulatory pressure to increase transparency about how their products—specifically recommendation algorithms—work and what impact they have. Academic researchers have long been calling for better access to data from social media platforms, including Meta. This new library is a step toward increased visibility about what is happening on its platforms and the effect that Meta’s products have on online conversations, politics, and society at large. 

In an interview, Meta’s president of global affairs, Nick Clegg, said the tools “are really quite important” in that they provide, in a lot of ways, “the most comprehensive access to publicly available content across Facebook and Instagram of anything that we’ve built to date.” The Content Library will also help the company meet new regulatory requirements and obligations on data sharing and transparency, as the company notes in a blog post Tuesday. 

The library and associated API were first released as a beta version several months ago and allow researchers to access near-real-time data about pages, posts, groups, and events on Facebook and creator and business accounts on Instagram, as well as the associated numbers of reactions, shares, comments, and post view counts. While all this data is publicly available—as in, anyone can see public posts, reactions, and comments on Facebook—the new library makes it easier for researchers to search and analyze this content at scale…(More)”.

Hypotheses devised by AI could find ‘blind spots’ in research

Curated on Posted on by Stefaan Verhulst

Article by Matthew Hutson: “One approach is to use AI to help scientists brainstorm. This is a task that large language models — AI systems trained on large amounts of text to produce new text — are well suited for, says Yolanda Gil, a computer scientist at the University of Southern California in Los Angeles who has worked on AI scientists. Language models can produce inaccurate information and present it as real, but this ‘hallucination’ isn’t necessarily bad, Mullainathan says. It signifies, he says, “‘here’s a kind of thing that looks true’. That’s exactly what a hypothesis is.”

Blind spots are where AI might prove most useful. James Evans, a sociologist at the University of Chicago, has pushed AI to make ‘alien’ hypotheses — those that a human would be unlikely to make. In a paper published earlier this year in Nature Human Behaviour4, he and his colleague Jamshid Sourati built knowledge graphs containing not just materials and properties, but also researchers. Evans and Sourati’s algorithm traversed these networks, looking for hidden shortcuts between materials and properties. The aim was to maximize the plausibility of AI-devised hypotheses being true while minimizing the chances that researchers would hit on them naturally. For instance, if scientists who are studying a particular drug are only distantly connected to those studying a disease that it might cure, then the drug’s potential would ordinarily take much longer to discover.

When Evans and Sourati fed data published up to 2001 to their AI, they found that about 30% of its predictions about drug repurposing and the electrical properties of materials had been uncovered by researchers, roughly six to ten years later. The system can be tuned to make predictions that are more likely to be correct but also less of a leap, on the basis of concurrent findings and collaborations, Evans says. But “if we’re predicting what people are going to do next year, that just feels like a scoop machine”, he adds. He’s more interested in how the technology can take science in entirely new directions….(More)”

Understanding AI jargon: Artificial intelligence vocabulary

Curated on Posted on by Stefaan Verhulst

Article by Kate Woodford: “Today, the Cambridge Dictionary announces its Word of the Year for 2023: hallucinate. You might already be familiar with this word, which we use to talk about seeing, hearing, or feeling things that don’t really exist. But did you know that it has a new meaning when it’s used in the context of artificial intelligence?

To celebrate the Word of the Year, this post is dedicated to AI terms that have recently come into the English language. AI, as you probably know, is short for artificial intelligence – the use of computer systems with qualities similar to the human brain that allow them to ‘learn’ and ‘think’. It’s a subject that arouses a great deal of interest and excitement and, it must be said, a degree of anxiety. Let’s have a look at some of these new words and phrases and see what they mean and how we’re using them to talk about AI…

As the field of AI continues to develop quickly, so does the language we use to talk about it. In a recent New Words post, we shared some words about AI that are being considered for addition to the Cambridge Dictionary…(More)”.