PASTCF President’s Column February 2024
Last fall my wife and I finally made the trip to celebrate our 50th anniversary. We went to Jordan, arriving there on 8 October 2023(!). I was going to write about that in this column — Jordan is a marvelous country, full of (biblical) history: Edom, Moab, Ammon, Gerasene, Decapolis, Jabbok, Mount Nebo, and much more. But in the meantime, much has happened in the AI domain that I want to write about.
The PASTCF board meets every year in April with the board of the Lutheran Alliance for Faith, Science and Technology at the Ecumenical Roundtable. Last year we decided that in 2024 the theme would be Artificial Intelligence. Here are some troublesome applications.
Our trip to Jordan during the early weeks of the war in Gaza made me pay extra attention to what has been happening in that part of the world. Even more so, since I am a member of Friends Of Sabeel in North America (FOSNA) and have followed the PC(USA) Israel/Palestine Mission Network (IPNM). An interview on the PBS Newshour with Yuval Abraham on 8 December 2023 caught my attention. Investigative journalist Yuval Abraham was the lead reporter of an article in the non-partisan Israeli magazine +972 on the methods used by the Israeli military in attacks on Gaza, based on interviews with current and former sources in the Israeli security community[1]. The report focuses on how military targets are selected and how the number of acceptable civilian deaths is determined.
There always were rules about how many civilians could be allowed to be killed as collateral damage for the assassination of a single senior Hamas member. That number used to be a few dozen, but these days it has increased to several hundred. Here is where AI comes in. Whereas target selection and collateral damage estimates in the past were carefully researched by human analysts, the military are currently using and AI-based platform with the abominable name Gospel to draw up lists of targets with “acceptable” numbers of collateral casualties (i.e., murders). As a result, the troubling trend to conduct war remotely, like a video game, has now progressed to a state where the war planning is also conducted and commanded by an automatic AI process. It makes fighting a war much more efficient: many more targets can be added to the attack lists at very short notice, but it eliminates even basic meaningful human oversight and intervention. Scary. Is this something we should call a Gospel?
Around the same time the Boston Globe published an article about advanced DNA testing of embryos. It is being advertised as the next step after IVF: have your embryos’ DNA completely sequenced and fed into a software tool that determines their chances of polygenetic diseases and a host of personal traits. Imagine you can select the child with the lowest probability of genetically influenced health problems, and the highest probability of strength, intelligence, good looks, etc. Clearly, AI was involved in the development of all this. And it may be tempting to have the ability to select your perfect child. But this is also wrought with ethical problems. It is a form of eugenics. Militarized groups may want to use it to set up a factory of suitable army kids, for instance. There is also a serious question about equity. IVF and the test are not cheap. So, will rich parents be able to provide their offspring with even more advantages (brains, health, …) in addition to wealth? Could this lead to an all-powerful elite that outshines the unwashed masses in every aspect of their being?
With the quadrennial election season upon us, the specter is rising of political campaigns conducted (or interfered with by external players) through fake video. It is quite feasible to produce natural-looking and sounding video material showing candidates doing and saying things that they never said or did. A careful inspection of the material may reveal its fakeness, but that will be after many people have seen and integrated it into what they “know” about the candidate.
As a member of the IEEE Computer Society and the Association for Computer Machinery (ACM) I have noticed a marked uptick of articles on the potential as well as the problems of AI. A recent issue of Communications of the ACM focused on the Inherent Limitations of AI Fairness. Let me just quote a few comments that struck me as useful and they give a sense of the complexity of the issues involved.
“AI fairness should not be considered a panacea: It may have the potential to make society more fair than ever, but it needs critical thought and outside help to make it happen.” “In an unjust world, it is meaningless to talk about the fairness of an algorithm’s decision without considering the wider socio-technological context in which the algorithm is applied.” “The sheer complexity of fairness may be used as a cover to avoid fully committing to it in practice.” A case in point: how can we expect ChatGPT to be non-discriminatory and non-racist, if we don’t know whether the texts that were fed into the engine adhered to those criteria? And that is a problem for Large Language Models (LLM). Their basic function to generate statistically likely continuations of word sequences is ultimately determined by what they were fed. A whole different area of concern is anthropomorphizing computers and robots, but I won’t go there.
There are bright spots. The AI Advisory Body of the UN has published its Interim Report: Governing AI for Humanity[2]. So, there is work being done. I quote from its introduction:
- “Artificial intelligence (AI) increasingly affects us all. Though AI has been around for years, capabilities once hardly imaginable have been emerging at a rapid, unprecedented pace. AI offers extraordinary potential for good — from scientific discoveries that expand the bounds of human knowledge to tools that optimize finite resources and assist us in everyday tasks. It could be a game changer in the transition to a greener future, or help developing countries transform public health and leapfrog challenges of last mile access in education. Developed countries with ageing populations could use it to tackle labour shortages.”
- “Yet, there are also risks. AI can reinforce biases or expand surveillance; automated decision-making can blur accountability of public officials even as AI-enhanced disinformation threatens the process of electing them. The speed, autonomy, and opacity of AI systems challenge traditional models of regulation, even as ever more powerful systems are developed, deployed and used.”
- “The opportunities and the risks of AI for people and society are evident and have seized public interest. They also manifest globally, with geostrategic tensions over access to the data, compute, and talent that fuel AI, with talk of a new AI arms race. Nor are the benefits and risks equitably distributed. There is a real danger, even if humanity harnesses only the positive aspects of AI, that those will be limited to a club of the rich. Today’s AI benefits are accruing largely to a handful of states, companies, and individuals.”
“This technology cries out for governance, not merely to address the challenges and risks but to ensure we harness its potential in ways that leave no one behind. A key measure of our success is the extent to which AI technologies help achieve the Sustainable Development Goals (SDGs).”
[1] https://www.972mag.com/mass-assassination-factory-israel-calculated-bombing-gaza/
[2] https://www.un.org/en/ai-advisory-body
