Paper Summary
Title: The Returns to Science In the Presence of Technological Risks
Source: arXiv (0 citations)
Authors: Matt Clancy
Published Date: 2023-12-21
Podcast Transcript
Hello, and welcome to paper-to-podcast.
Today, we delve into the fascinating world of fiscal science fiction—well, not entirely fiction—as we explore a recent paper titled "The Returns to Science In the Presence of Technological Risks." Authored by Matt Clancy and published on December 21st, 2023, this paper is like a rollercoaster for the nerdy economist in all of us.
Clancy takes us on an economic adventure, arguing that the social impact of chucking a dollar into the science hat is not just double, triple, or even quadruple, but a whopping 230 times more beneficial than just handing that same dollar to someone earning a cool $50,000 a year. Talk about getting bang for your buck or, in this case, your research grant.
But wait, before you start throwing money at scientists left, right, and center, Clancy adds a dash of cautionary spice to the stew. With great power comes great responsibility, and with great science comes the risk of creating new bioweapons. Cue the dramatic music because we're entering what Clancy calls the "time of perils," which sounds like a rejected Harry Potter book title but is actually about the potential for more frequent and deadly pandemics.
Here's the kicker, though: even when we factor in the possibility of getting knocked out by our own scientific curiosity, the benefits of science still seem to outweigh the costs—as long as the grim reaper doesn't start working overtime with an increase in annual mortality of less than 0.2-0.5%.
Now, we've all heard of superforecasters, right? No? Well, they're like the weather forecasters of doom and gloom, predicting when technology might backfire. According to these folks and other domain experts, the long-run risks of genetically engineered pandemics will eventually go the way of the dodo. That's right; science might clean up its own mess, mitigating the time of perils.
But wait, there's more! Clancy cleverly suggests that if we put the pedal to the metal on scientific progress, we could zip past these perilous times faster. It's like saying the best way to get over a hangover is to keep the party going.
How did Clancy come up with these insights, you ask? Through a quantitative economic model that makes a high school calculus class look like kindergarten. This model looks at the long-term effects of science on income and health and compares it to giving a dollar to someone who's doing alright financially.
The paper's strength lies in its comprehensive approach, using a utility framework inspired by Open Philanthropy's work. It's like having a super calculator that not only counts your money but also measures happiness and life expectancy. The researchers even put a price tag on the risk of pandemics brought about by advanced biotechnology. The use of forecasts from a diverse group of experts ensures that they're not putting all their eggs in one basket.
Now, no study is perfect, and this one is no exception. It relies on forecasts, which are about as certain as a chameleon's color in a disco. The models used are based on assumptions that could be as wobbly as a Jenga tower in an earthquake. The study simplifies the social return on investment for science, which could overlook some nuances in economic growth and health improvements.
Despite these limitations, the potential applications are as wide as the Grand Canyon. This research could be the North Star for science policy decisions, philanthropic investments, and government funding priorities. It's about finding that sweet spot between chasing scientific breakthroughs and not accidentally unleashing a zombie apocalypse.
Imagine a world where we can enjoy the fruits of scientific labor without worrying about turning into Frankenstein's monster. That's the future this paper is pointing towards.
And with that thought, we wrap up another episode of paper-to-podcast. You can find this paper and more on the paper2podcast.com website. Stay curious, stay safe, and remember: science is like a box of chocolates; it's mostly sweet, but watch out for the occasional nut.
Supporting Analysis
The paper suggests that the social impact of spending on scientific research is hundreds of times more beneficial than just giving a dollar to someone earning $50,000 per year. Specifically, the preferred model finds that the social return on investment (ROI) of science is roughly 230 times the social ROI of the benchmark scenario. However, this positive outlook is tempered by the consideration of technological risks, particularly the potential for scientific advances to enable the creation of new bioweapons. If such risks lead to more frequent and deadly pandemics, known as the "time of perils," it could lower the social welfare benefits of scientific progress. Despite this, even with the introduction of risks from advanced biotechnology, the paper indicates that the benefits of science are likely to outweigh the costs as long as the expected increase in annual mortality due to these scientific perils is less than 0.2-0.5% per year. Forecasts by superforecasters and domain experts predict that the long-run risks of genetically engineered pandemics will eventually decline, suggesting that science could eventually mitigate the time of perils. Surprisingly, the paper argues that faster scientific progress could be net welfare improving because it might accelerate us to the end of the perilous times more quickly.
The research paper employs a quantitative economic model to evaluate the social impact of spending on scientific progress. The model integrates the long-term effects of science on per capita income and health into a utility framework. The model then compares the benefits of the average dollar spent on science to the social impact of giving a dollar to an individual earning $50,000 per year. The paper examines two main aspects: the social impact of spending on science and the conditions under which accelerating scientific progress could raise or lower social welfare. To estimate the social impact, the study incorporates a long-run view on the growth impact of science on health and income. For the conditions affecting social welfare, the model considers the dilemma that scientific advancements may soon enable a broader range of actors to engineer viruses, potentially leading to more frequent and deadly pandemics, referred to as the "time of perils." The research develops a framework to compare the historical benefits of science against the forecasted costs of these technological risks, using a variety of expert and superforecaster estimates to model the potential harms from advanced biotechnology. The model also includes a discount rate to weigh the utility of future people versus those alive today, accounting for the uncertainty about the future's resemblance to the past.
The most compelling aspect of the research is its comprehensive approach to assessing the societal impact of science spending, especially its potential to accelerate both benefits and risks. The researchers used a utility framework inspired by Open Philanthropy's work, which allowed them to integrate the long-term effects of science on per capita income and health into a cohesive model. They compared the benefits of science spending against the social impact of simply giving a dollar to an individual earning $50,000 per year, providing a clear and relatable benchmark for their analysis. One of the best practices followed in the research was the thorough effort to quantify the risks associated with accelerating scientific progress, particularly focusing on the potential for enabling more frequent and deadly pandemics through advanced biotechnology. The use of a quantitative economic model to balance historical benefits against forecasted costs demonstrates a methodical approach. Additionally, the researchers incorporated forecasts from a diverse group of experts and superforecasters, ensuring a wide range of expert opinions was considered. This robust approach to forecasting, combined with a neutral stance on various assumptions in the face of evidence scarcity, illustrates a commitment to objectivity and thoroughness in their methodology.
Possible limitations of the research include the reliance on forecasts and models for predicting future risks, which inherently contain uncertainty and may be influenced by the subjective judgments of experts. The models used also make several assumptions, such as a constant probability of exiting the current epistemic regime and a fixed mortality rate during the "time of perils," which may not fully capture the dynamic nature of scientific progress and its societal impacts. Additionally, the research uses a simplified approach to estimate the social return on investment for science, which may not account for all variables and complexities of economic growth and health improvements. The findings also hinge on the accuracy of the assumed parameters, such as the rate of scientific progress, the effectiveness of science in reducing mortality, and the probability of civilization-ending events, which are difficult to quantify with precision. Lastly, the research abstracts from the heterogeneity of scientific endeavors by considering an average dollar spent on science, potentially overlooking the varying impacts of different scientific fields and types of research.
The research explores the societal impact of investing in scientific progress and weighs it against the potential risks associated with enabling the creation of dangerous biological technologies. Potential applications of this research are wide-ranging and include informing science policy decisions, guiding philanthropic investment strategies, and shaping government funding priorities. By quantifying the return on investment (ROI) for science spending and modeling the risks of biocatastrophes, the findings can be used to balance the pursuit of scientific advancement with the need for biosecurity measures. This can lead to the development of more robust frameworks for evaluating the long-term benefits and risks of scientific research, ultimately contributing to safer and more effective science governance. The research could also influence public health strategies by highlighting the importance of preparing for and mitigating the risks of potential pandemics, including those that could be man-made. Additionally, the insights gained could be pivotal in discussions about the ethical implications of emerging technologies and the importance of responsible innovation.