What’s a Fact, Anyway?

Curated on Posted on by Stefaan Verhulst

Essay by Fergus McIntosh: “…For journalists, as for anyone, there are certain shortcuts to trustworthiness, including reputation, expertise, and transparency—the sharing of sources, for example, or the prompt correction of errors. Some of these shortcuts are more perilous than others. Various outfits, positioning themselves as neutral guides to the marketplace of ideas, now tout evaluations of news organizations’ trustworthiness, but relying on these requires trusting in the quality and objectivity of the evaluation. Official data is often taken at face value, but numbers can conceal motives: think of the dispute over how to count casualties in recent conflicts. Governments, meanwhile, may use their powers over information to suppress unfavorable narratives: laws originally aimed at misinformation, many enacted during the COVID-19 pandemic, can hinder free expression. The spectre of this phenomenon is fuelling a growing backlash in America and elsewhere.

Although some categories of information may come to be considered inherently trustworthy, these, too, are in flux. For decades, the technical difficulty of editing photographs and videos allowed them to be treated, by most people, as essentially incontrovertible. With the advent of A.I.-based editing software, footage and imagery have swiftly become much harder to credit. Similar tools are already used to spoof voices based on only seconds of recorded audio. For anyone, this might manifest in scams (your grandmother calls, but it’s not Grandma on the other end), but for a journalist it also puts source calls into question. Technologies of deception tend to be accompanied by ones of detection or verification—a battery of companies, for example, already promise that they can spot A.I.-manipulated imagery—but they’re often locked in an arms race, and they never achieve total accuracy. Though chatbots and A.I.-enabled search engines promise to help us with research (when a colleague “interviewed” ChatGPT, it told him, “I aim to provide information that is as neutral and unbiased as possible”), their inability to provide sourcing, and their tendency to hallucinate, looks more like a shortcut to nowhere, at least for now. The resulting problems extend far beyond media: election campaigns, in which subtle impressions can lead to big differences in voting behavior, feel increasingly vulnerable to deepfakes and other manipulations by inscrutable algorithms. Like everyone else, journalists have only just begun to grapple with the implications.

In such circumstances, it becomes difficult to know what is true, and, consequently, to make decisions. Good journalism offers a way through, but only if readers are willing to follow: trust and naïveté can feel uncomfortably close. Gaining and holding that trust is hard. But failure—the end point of the story of generational decay, of gold exchanged for dross—is not inevitable. Fact checking of the sort practiced at The New Yorker is highly specific and resource-intensive, and it’s only one potential solution. But any solution must acknowledge the messiness of truth, the requirements of attention, the way we squint to see more clearly. It must tell you to say what you mean, and know that you mean it…(More)”.

AI for Social Good

Curated on Posted on by Stefaan Verhulst

Essay by Iqbal Dhaliwal: “Artificial intelligence (AI) has the potential to transform our lives. Like the internet, it’s a general-purpose technology that spans sectors, is widely accessible, has a low marginal cost of adding users, and is constantly improving. Tech companies are rapidly deploying more capable AI models that are seeping into our personal lives and work.

AI is also swiftly penetrating the social sector. Governments, social enterprises, and NGOs are infusing AI into programs, while public treasuries and donors are working hard to understand where to invest. For example, AI is being deployed to improve health diagnostics, map flood-prone areas for better relief targeting, grade students’ essays to free up teachers’ time for student interaction, assist governments in detecting tax fraud, and enable agricultural extension workers to customize advice.

But the social sector is also rife with examples over the past two decades of technologies touted as silver bullets that fell short of expectations, including One Laptop Per Child, SMS reminders to take medication, and smokeless stoves to reduce indoor air pollution. To avoid a similar fate, AI-infused programs must incorporate insights from years of evidence generated by rigorous impact evaluations and be scaled in an informed way through concurrent evaluations.

Specifically, implementers of such programs must pay attention to three elements. First, they must use research insights on where AI is likely to have the greatest social impact. Decades of research using randomized controlled trials and other exacting empirical work provide us with insights across sectors on where and how AI can play the most effective role in social programs.

Second, they must incorporate research lessons on how to effectively infuse AI into existing social programs. We have decades of research on when and why technologies succeed or fail in the social sector that can help guide AI adopters (governments, social enterprises, NGOs), tech companies, and donors to avoid pitfalls and design effective programs that work in the field.

Third, we must promote the rigorous evaluation of AI in the social sector so that we disseminate trustworthy information about what works and what does not. We must motivate adopters, tech companies, and donors to conduct independent, rigorous, concurrent impact evaluations of promising AI applications across social sectors (including impact on workers themselves); draw insights emerging across multiple studies; and disseminate those insights widely so that the benefits of AI can be maximized and its harms understood and minimized. Taking these steps can also help build trust in AI among social sector players and program participants more broadly…(More)”.

What Could Citizens’ Assemblies Do for American Politics?

Curated on Posted on by Stefaan Verhulst

Essay by Nick Romeo: “Last July, an unusual letter arrived at Kathryn Kundmueller’s mobile home, in central Oregon. It invited her to enter a lottery that would select thirty residents of Deschutes County to deliberate for five days on youth homelessness—a visible and contentious issue in an area where the population and cost of living have spiked in recent years. Those chosen would be paid for their time—almost five hundred dollars—and asked to develop specific policy recommendations.

Kundmueller was being invited to join what is known as a citizens’ assembly. These gatherings do what most democracies only pretend to: trust normal people to make decisions on difficult policy questions. Many citizens’ assemblies follow a basic template. They impanel a random but representative cross-section of a population, give them high-quality information on a topic, and ask them to work together to reach a decision. In Europe, such groups have helped spur reform of the Irish constitution in order to legalize abortion, guided an Austrian pharmaceutical heiress on how to give away her wealth, and become a regular part of government in Paris and Belgium. Though still rare in America, the model reflects the striking idea that fundamental problems of politics—polarization, apathy, manipulation by special interests—can be transformed through radically direct democracy.

Kundmueller, who is generally frustrated by politics, was intrigued by the letter. She liked the prospect of helping to shape local policy, and the topic of housing insecurity had a particular resonance for her. As a teen-ager, following a falling-out with her father, she spent months bouncing between friends’ couches in Vermont. When she moved across the country to San Jose, after college, she lived in her car for a time while she searched for a stable job. She worked in finance but became disillusioned; now in her early forties, she ran a small housecleaning business. She still thought about living in a van and renting out her mobile home to save money…(More)”.

Will Artificial Intelligence Replace Us or Empower Us?

Curated on Posted on by Stefaan Verhulst

Article by Peter Coy: “…But A.I. could also be designed to empower people rather than replace them, as I wrote a year ago in a newsletter about the M.I.T. Shaping the Future of Work Initiative.

Which of those A.I. futures will be realized was a big topic at the San Francisco conference, which was the annual meeting of the American Economic Association, the American Finance Association and 65 smaller groups in the Allied Social Science Associations.

Erik Brynjolfsson of Stanford was one of the busiest economists at the conference, dashing from one panel to another to talk about his hopes for a human-centric A.I. and his warnings about what he has called the “Turing Trap.”

Alan Turing, the English mathematician and World War II code breaker, proposed in 1950 to evaluate the intelligence of computers by whether they could fool someone into thinking they were human. His “imitation game” led the field in an unfortunate direction, Brynjolfsson argues — toward creating machines that behaved as much like humans as possible, instead of like human helpers.

Henry Ford didn’t set out to build a car that could mimic a person’s walk, so why should A.I. experts try to build systems that mimic a person’s mental abilities? Brynjolfsson asked at one session I attended.

Other economists have made similar points: Daron Acemoglu of M.I.T. and Pascual Restrepo of Boston University use the term “so-so technologies” for systems that replace human beings without meaningfully increasing productivity, such as self-checkout kiosks in supermarkets.

People will need a lot more education and training to take full advantage of A.I.’s immense power, so that they aren’t just elbowed aside by it. “In fact, for each dollar spent on machine learning technology, companies may need to spend nine dollars on intangible human capital,” Brynjolfsson wrote in 2022, citing research by him and others…(More)”.

AI Is Bad News for the Global South

Curated on Posted on by Stefaan Verhulst

Article by Rachel Adams: “…AI’s adoption in developing regions is also limited by its design. AI designed in Silicon Valley on largely English-language data is not often fit for purpose outside of wealthy Western contexts. The productive use of AI requires stable internet access or smartphone technology; in sub-Saharan Africa, only 25 percent of people have reliable internet access, and it is estimated that African women are 32 percent less likely to use mobile internet than their male counterparts.

Generative AI technologies are also predominantly developed using the English language, meaning that the outputs they produce for non-Western users and contexts are oftentimes useless, inaccurate, and biased. Innovators in the global south have to put in at least twice the effort to make their AI applications work for local contexts, often by retraining models on localized datasets and through extensive trial and error practices.

Where AI is designed to generate profit and entertainment only for the already privileged, it will not be effective in addressing the conditions of poverty and in changing the lives of groups that are marginalized from the consumer markets of AI. Without a high level of saturation across major industries, and without the infrastructure in place to enable meaningful access to AI by all people, global south nations are unlikely to see major economic benefits from the technology.

As AI is adopted across industries, human labor is changing. For poorer countries, this is engendering a new race to the bottom where machines are cheaper than humans and the cheap labor that was once offshored to their lands is now being onshored back to wealthy nations. The people most impacted are those with lower education levels and fewer skills, whose jobs can be more easily automated. In short, much of the population in lower- and middle-income countries may be affected, severely impacting the lives of millions of people and threatening the capacity of poorer nations to prosper…(More)”.

The world of tomorrow

Curated on Posted on by Stefaan Verhulst

Essay by Virginia Postrel: “When the future arrived, it felt… ordinary. What happened to the glamour of tomorrow?

Progress used to be glamorous. For the first two thirds of the twentieth-century, the terms modern, future, and world of tomorrow shimmered with promise.

Glamour is more than a synonym for fashion or celebrity, although these things can certainly be glamorous. So can a holiday resort, a city, or a career. The military can be glamorous, as can technology, science, or the religious life. It all depends on the audience. Glamour is a form of communication that, like humor, we recognize by its characteristic effect. Something is glamorous when it inspires a sense of projection and longing: if only . . .

Whatever its incarnation, glamour offers a promise of escape and transformation. It focuses deep, often unarticulated longings on an image or idea that makes them feel attainable. Both the longings – for wealth, happiness, security, comfort, recognition, adventure, love, tranquility, freedom, or respect – and the objects that represent them vary from person to person, culture to culture, era to era. In the twentieth-century, ‘the future’ was a glamorous concept…

Much has been written about how and why culture and policy repudiated the visions of material progress that animated the first half of the twentieth-century, including a special issue of this magazine inspired by J Storrs Hall’s book Where Is My Flying Car? The subtitle of James Pethokoukis’s recent book The Conservative Futurist is ‘How to create the sci-fi world we were promised’. Like Peter Thiel’s famous complaint that ‘we wanted flying cars, instead we got 140 characters’, the phrase captures a sense of betrayal. Today’s techno-optimism is infused with nostalgia for the retro future.

But the most common explanations for the anti-Promethean backlash fall short. It’s true but incomplete to blame the environmental consciousness that spread in the late sixties…

How exactly today’s longings might manifest themselves, whether in glamorous imagery or real-life social evolution, is hard to predict. But one thing is clear: For progress to be appealing, it must offer room for diverse pursuits and identities, permitting communities with different commitments and values to enjoy a landscape of pluralism without devolving into mutually hostile tribes. The ideal of the one best way passed long ago. It was glamorous in its day but glamour is an illusion…(More)”.

The AI tool that can interpret any spreadsheet instantly

Curated on Posted on by Stefaan Verhulst

Article by Duncan C. McElfresh: “Say you run a hospital and you want to estimate which patients have the highest risk of deterioration so that your staff can prioritize their care1. You create a spreadsheet in which there is a row for each patient, and columns for relevant attributes, such as age or blood-oxygen level. The final column records whether the person deteriorated during their stay. You can then fit a mathematical model to these data to estimate an incoming patient’s deterioration risk. This is a classic example of tabular machine learning, a technique that uses tables of data to make inferences. This usually involves developing — and training — a bespoke model for each task. Writing in Nature, Hollmann et al.report a model that can perform tabular machine learning on any data set without being trained specifically to do so.

Tabular machine learning shares a rich history with statistics and data science. Its methods are foundational to modern artificial intelligence (AI) systems, including large language models (LLMs), and its influence cannot be overstated. Indeed, many online experiences are shaped by tabular machine-learning models, which recommend products, generate advertisements and moderate social-media content3. Essential industries such as healthcare and finance are also steadily, if cautiously, moving towards increasing their use of AI.

Despite the field’s maturity, Hollmann and colleagues’ advance could be revolutionary. The authors’ contribution is known as a foundation model, which is a general-purpose model that can be used in a range of settings. You might already have encountered foundation models, perhaps unknowingly, through AI tools, such as ChatGPT and Stable Diffusion. These models enable a single tool to offer varied capabilities, including text translation and image generation. So what does a foundation model for tabular machine learning look like?

Let’s return to the hospital example. With spreadsheet in hand, you choose a machine-learning model (such as a neural network) and train the model with your data, using an algorithm that adjusts the model’s parameters to optimize its predictive performance (Fig. 1a). Typically, you would train several such models before selecting one to use — a labour-intensive process that requires considerable time and expertise. And of course, this process must be repeated for each unique task.

Figure 1 | A foundation model for tabular machine learning. a, Conventional machine-learning models are trained on individual data sets using mathematical optimization algorithms. A different model needs to be developed and trained for each task, and for each data set. This practice takes years to learn and requires extensive time and computing resources. b, By contrast, a ‘foundation’ model could be used for any machine-learning task and is pre-trained on the types of data used to train conventional models. This type of model simply reads a data set and can immediately produce inferences about new data points. Hollmann et al. developed a foundation model for tabular machine learning, in which inferences are made on the basis of tables of data. Tabular machine learning is used for tasks as varied as social-media moderation and hospital decision-making, so the authors’ advance is expected to have a profound effect in many areas…(More)”

Government reform starts with data, evidence

Curated on Posted on by Stefaan Verhulst

Article by Kshemendra Paul: “It’s time to strengthen the use of data, evidence and transparency to stop driving with mud on the windshield and to steer the government toward improving management of its programs and operations.

Existing Government Accountability Office and agency inspectors general reports identify thousands of specific evidence-based recommendations to improve efficiency, economy and effectiveness, and reduce fraud, waste and abuse. Many of these recommendations aim at program design and requirements, highlighting specific instances of overlap, redundancy and duplication. Others describe inadequate internal controls to balance program integrity with the experience of the customer, contractor or grantee. While progress is being reported in part due to stronger partnerships with IGs, much remains to be done. Indeed, GAO’s 2023 High Risk List, which it has produced going back to 1990, shows surprisingly slow progress of efforts to reduce risk to government programs and operations.

Here are a few examples:

  • GAO estimates recent annual fraud of between $233 billion to $521 billion, or about 3% to 7% of federal spending. On the other hand, identified fraud with high-risk Recovery Act spending was held under 1% using data, transparency and partnerships with Offices of Inspectors General.
  • GAO and IGs have collectively identified hundreds of billions in potential cost savings or improvements not yet addressed by federal agencies.
  • GAO has recently described shortcomings with the government’s efforts to build evidence. While federal policymakers need good information to inform their decisions, the Commission on Evidence-Based Policymaking previously said, “too little evidence is produced to meet this need.”

One of the main reasons for agency sluggishness is the lack of agency and governmentwide use of synchronized, authoritative and shared data to support how the government manages itself.

For example, the Energy Department IG found that, “[t]he department often lacks the data necessary to make critical decisions, evaluate and effectively manage risks, or gain visibility into program results.” It is past time for the government to commit itself to move away from its widespread use of data calls, the error-prone, costly and manual aggregation of data used to support policy analysis and decision-making. Efforts to embrace data-informed approaches to manage government programs and operations are stymied by lack of basic agency and governmentwide data hygiene. While bright pockets exist, management gaps, as DOE OIG stated, “create blind spots in the universe of data that, if captured, could be used to more efficiently identify, track and respond to risks…”

The proposed approach starts with current agency operating models, then drives into management process integration to tackle root causes of dysfunction from the bottom up. It recognizes that inefficiency, fraud and other challenges are diffused, deeply embedded and have non-obvious interrelationships within the federal complex…(More)”

Academic writing is getting harder to read—the humanities most of all

Curated on Posted on by Stefaan Verhulst

The Economist: “Academics have long been accused of jargon-filled writing that is impossible to understand. A recent cautionary tale was that of Ally Louks, a researcher who set off a social media storm with an innocuous post on X celebrating the completion of her PhD. If it was Ms Louks’s research topic (“olfactory ethics”—the politics of smell) that caught the attention of online critics, it was her verbose thesis abstract that further provoked their ire. In two weeks, the post received more than 21,000 retweets and 100m views.

Although the abuse directed at Ms Louks reeked of misogyny and anti-intellectualism—which she admirably shook off—the reaction was also a backlash against an academic use of language that is removed from normal life. Inaccessible writing is part of the problem. Research has become harder to read, especially in the humanities and social sciences. Though authors may argue that their work is written for expert audiences, much of the general public suspects that some academics use gobbledygook to disguise the fact that they have nothing useful to say. The trend towards more opaque prose hardly allays this suspicion…(More)”.

Once It Has Been Trained, Who Will Own My Digital Twin?

Curated on Posted on by Stefaan Verhulst

Article by Todd Carpenter: “Presently, if one ignores the hype around Generative AI systems, we can recognize that software tools are not sentient. Nor can they (yet) overcome the problem of coming up with creative solutions to novel problems. They are limited in what they can do by the training data that they are supplied. They do hold the prospect for making us more efficient and productive, particularly for wrote tasks. But given enough training data, one could consider how much farther this could be taken. In preparation for that future, when it comes to the digital twins, the landscape of the ownership of the intellectual property (IP) behind them is already taking shape.

Several chatbots have been set up to replicate long-dead historical figures so that you can engage with them in their “voice”.  Hellohistory is an AI-driven chatbot that provides people the opportunity to, “have in-depth conversations with history’s greatest.” A different tool, Historical Figures Chat, was widely panned not long after its release in 2023, and especially by historians who strongly objected. There are several variations on this theme of varying quality. Of course, with all things GenAI, they will improve over time and many of the obvious and problematic issues will be resolved either by this generation of companies or the next. Whether there is real value and insight to be gained, apart from the novelty, of engaging with “real historical figures” is the multi-billion dollar question. Much like the World Wide Web in the 1990s, very likely there is value, but it will be years before it can be clearly discerned what that value is and how to capitalize upon it. In anticipation of that day, many organizations are positioning themselves to capture that value.

While many universities have taken a very liberal view of ownership of the intellectual property of their students and faculty — far more liberal than many corporations might — others are quite more restrictive…(More)”.