Eventually and some months after its presentation at the Geneva Centre for Security Policy, I have had the chance of diving into the book “The AI Wave in Defense Innovation. Assessing Military Artificial Intelligence Strategies, Capabilities, and Trajectories.” Edited by Michael Raska and Richard A. Bitzinger and with the contribution from experts in different fields -AI, intelligence, technology governance, defense innovation…-, this volume is an international and interdisciplinary perspective on the adoption and governance of AI in defense and military innovation by major and middle powers.
From the different chapters, I deem it should be stressed the one titled “AI Ethics and Governance in Defense Innovation. Implementing AI Ethics Framework” by Cansu Canca, owing to the meaningful insights included in it.
For the author, AI use within a military context raises various ethical concerns due to high ethical risks often inherent in these technologies. A comprehensive account of the ethical landscape of AI in the military must also account for the potential ethical benefits. AI ethics within a military context is hence a double-edged sword. To illustrate this dual aspect, Canca lists some ethical pairs of pros and cons:
Precision vs. Aggression
A major area of AI applications for the military is about increasing precision -the Pentagon-funded Project Maven is probable the best example of this.
The military benefits of increased precision in surveillance and potentially in targeting are clear. AI may help search and rescue mission, predict and prevent deadly attacks from the opponent, and eliminate or decrease errors in defense. However, these increased capabilities might also boost the confidence of armed forces and make them more aggressive in their offensive and defensive attacks, resulting in more armed conflicts and more casualties.
Efficiency vs. Lowered Barriers to Conflict
AI systems that reduce the need for human resources, keep humans safe, and allow human officers to use their expertise are beneficial for the military and military personnel. The other side of the coin is the concern that increasing efficiency and safety for the military will also lower the barriers to entering a conflict: if war is both safe and low cost, what would stand in the way of starting one?
Those who lack the technology would be deterred from starting a war, whereas those equipped with the technology could become even more eager to escalate a conflict.
Protecting combatants vs. Neglecting Responsibility and Humanity
Sparing military personnel’s lives and keeping them safe from death and physical and mental harm would be a clear benefit for the military. However, it is never that simple: Can an officer remotely operating a weapon feel the weight of responsibility of “pulling the trigger” as strongly when they are distant from the “other”? Can they view the “other” as possessing humanity, when they are no longer face-to-face with them?
The ethical decision-making of a human is inevitably intertwined with human psychology. How combatants feel about the situation and the other parties necessarily feature in their ethical reasoning. The question is: where, if anywhere, should we draw the line on the spectrum of automation and remote control to ensure that human officers still engage in ethical decision-making, acknowledging the weight of their decisions and responsibility as they operate and collaborate with AI tools?
Military use of AI is not good or bad; it is a mixed bag. For that reason, more than ever it is needed the creation and implementation of frameworks for the ethical design, development, and deployment of AI systems.
AI Ethics Framework
AI ethics has increasingly been a core area of concern across sectors, particularly since 2017, when technology-related scandals slowly started to catch the public’s attention. AI ethics is concerned with the whole AI development and deployment cycle. This includes research, development, design, deployment, use, and even the updating stages of the AI life cycle.
Each stage presents its unique ethical questions, being some of them:
Does the dataset represente the world and, if not, who is left out?
Does the algorithm prioritize a value explicitly or implicitly? And if it does, is this justified?
Does the dashboard provide information for a user to understand the core variables of an AI decision-aid tool?
Did research subjects consent to have their data used?
Do professional such as military officers, managers, physicians, and administrators understand the limitation of the AI tools they are handed?
As users engage with AI tools in their professional or personal daily lives, do they agree to have their data collected and used, and do they understand its risks?
What makes ethical issues AI-specific are the extremely short gap between R&D and practice, the wide-scale and systematic use of AI systems, the increased capabilities of AI systems, and the façade of computational objectivity. Concerning the latter, this point is extremely dangerous because it allows users -such as judges and officers- to put aside their critical approach and prevents them from questioning the system’s results. Inadvertently the AI system leads them to make ethical errors and do so systematically due to their over-reliance on the AI tool.
The ethical questions and issues that AI raises cannot be addressed through this traditional ethics compliance and oversight model. Neither can they be solved solely through regulation. Instead, and according to the author, we need a comprehensive AI ethics framework to address all aspects of AI innovation and use in organizations. A proper AI ethics strategy consists of the playbook component, which includes all ethical guiding materials such as ethics principles, use cases, and tools; the process component, which includes ethics analysis and ethics-by-design components, and structures how and when to integrate these and other ethics components into the organizational operations; and the people component, which structures a network of different ethics roles within the organization.
AI Ethics Playbook
The AI ethics playbook forms the backbone of the AI ethics strategy, consisting of all guidelines and tools to aid ethical decision-making at all levels; i.e., fundamental principles to create a solid basis, accesible and user-friendly tools that help apply these principles, and use cases that demonstrate how to use these tools and bring the principles into action. The AI ethics playbook should be a living document.
As an example, these are the AI ethics principles recommended by the US Department of Defense in 2019:
Responsibility: DoD personnel will exercise appropriate levels of judgement and care concerning the development, deployment, and use of AI capabilities.
Equitability: The DoD will take the required steps to minimize unintended bias in AI capabilities.
Traceability: DoD relevant personnel must possess an appropriate understanding of the technology, development processes, and operational methods applicable to AI capabilities.
Reliability: the safety, security, and effectiveness of AI capabilities will be subject to testing and assurance.
Governability: DoD will design AI capabilities to fulfill their intended functions while processing the ability to detect and avoid unintended consequences, and the ability to disengage or deactivate deployed systems that demonstrate unintended behavior.
AI Ethics Process
AI playbook is inconsequential if there is no process to integrate it into the organization’s operations. Therefore an essential part of an AI ethics strategy is figuring out how to add the playbook into the organization’s various operations seamlessly. The objective of the AI ethics process is to create systematic procedures for ethics decisions which include: structuring the workflow to add “ethics checkpoints”; ethics analysis and ethics-by-design sessions; documenting ethics decisions and adding them into the playbook as use cases; and finally ensuring a feedback loop between new decisions and the updating of the playbook when needed.
AI Ethics People
A network of people who carry different roles concerning ethics should be embedded into every layer of the organization. It should be clear from the tasks at hand that the ethics team’s work is distinct from the legal team. Whilst there should be close collaboration between legal and ethics experts, the focus of legal experts and their skills and tools differ significantly from those of ethics experts.
And last but not least, from Canca’s standpoint, an ethics framework without an accountability mechanism would be limited to individuals’ motivation to “do good” rather than functioning as a reliable and strategic organizational tool. To that end, an AI ethics framework should be supported by AI regulations. AI regulations would function as an external enforcing mechanism to define the boundaries of legally acceptable action and to establish a fair playing field for competition. Regulations should also function to support the AI ethics framework by requiring basic components of this framework to be implemented such as an audit mechanism for the safety, reliability, and fairness of AI systems. Having such a regulatory framework for AI ethics governance would also help create citizen trust.
In 2020 Meta (then known as Facebook) published the paper Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks, which came up with a framework called retrieval-augmented generation (RAG) to give LLMs access to information beyond their training data.
Large language models can be inconsistent. Sometimes they can grant a perfect answer to a question, but other times they regurgitate aleatory facts from their training data.
Retrieval-augmented generation (RAG) is a technique for enhancing the accuracy and reliability of generative AI models with facts fetched from external sources. In other words, it fills a gap in how LLMs work. Under the hood, LLMs are neural networks, typically measured by how many parameters they contain. An LLM’s parameters essentially represent the general patterns of how humans use words to form sentences. That deep understanding makes LLMs useful in responding to general prompts extremely fast. Nonetheless, it does not serve users who want a deeper dive into a current or more specific topic. Retrieval-augmented generation (RAG) gives models sources they can cite, so users can check any claims. That builds trust. What’s more, the technique can help models clear up ambiguity in a user query.
The roots of the technique go back at least to the early 1970s. That’s when researchers in information retrieval (IR) prototyped what they called question-answering systems, apps that use natural language processing to access text initially in narrow topics.
Implementing RAG in an LLM-based question answering system has two main benefits: It ensures that the model has access to the most current, reliable facts, and that users have access to the model’s sources, ensuring that the accuracy of responses can be easily checked.
By grounding an LLM on a set of external, verifiable facts, the model has fewer opportunities to “hallucinate” or mislead information. RAG allows LLMs to build on a specialized body of knowledge to answer questions in more accurate way. It also reduces the need for users to continuously train the model on new data and update its parameters, as circumstances evolve. In this way, RAG can lower the computational and financial costs of running LLM-powered chatbots in an enterprise setting.
RAG has two phases: retrieval and content generation. In the retrieval phase, algorithms search for and retrieve snippets of information relevant to the user’s prompt or question. In an open-domain, consumer setting, those facts can come from indexed documents on the internet; in a closed-domain, enterprise setting, a narrower set of sources are typically used for added security and reliability.
This collection of outside information is sent to the language model along with the user’s request. During the generative phase, the LLM creates an appealing answer that is customized for the user currently using the enhanced prompt and its internal representation of its training data. A chatbot can then be given the response together with connections to its original sources. The entire procedure can be represented graphically as follows:
Summing up, customer queries are not always straightforward. They can be ambiguously worded, complex, or require knowledge the model either doesn’t have or can’t easily parse. These are the conditions in which LLMs are prone to making things up. LLMs need to be explicitly trained to recognize questions they can’t answer, it may need though to see thousands of examples of questions that can and can’t be answered. Only then can the model learn to identify an unanswerable question, and probe for more detail until it hits on a question that it has the information to answer. RAG is currently the best-known tool for grounding LLMs on the latest, verifiable information, and it allows LLMs to go one step further by greatly reducing the need to feed and retrain the model on fresh examples.
For much time it seemed that in the computing landscape the main application of graphs were only related to ontology engineering, so when my colleague Mihael shared with me the paper “Graph of Thoughts: Solving Elaborate Problems with Large Language Models” -published by the end of August-, I thought we might be in the right path to re-discover the power to representing knowledge of these structures. In the afore-mentioned paper, the authors harness the graph abstraction as a key mechanism that enhances prompting capabilities in LLMs.
Prompt engineering is one of the central new domains of the large language model research. However, designing effective prompts is a challenging task. Graph of Thoughts (GoT) is a new paradigm that enables the LLM to solve different tasks effectively without any model updates.The key idea is to model the LLM reasoning as a graph, where thoughts are vertices and dependencies between thoughts are edges.
Human’s task solving is often non-linear, and it involves combining intermediate solutions into final ones, or changing the flow of reasoning upon discovering new in sights. For example, a person could explore a certain chain of reasoning, backtrack and start a new one, then realize that a certain idea from the previous chain could be combined with the currently explored one, and merge them both into a new solution, taking advantage of their strengths and eliminating their weaknesses. GoT reflects this, so to say, anarchic reason process with its graph structure.
Nonetheless, let’s take a step back: besides Graph of Thoughts, there are other approaches for prompting:
Input-Output (IO): a straightforward approach in which we use an LLM to turn an input sequence x into the output y directly, without any intermediate thoughts.
Chain-of-Thought (CoT): one introduces intermediate thoughts a1, a2,… between x and y. This strategy was shown to significantly enhance various LLM tasks over the plain IO baseline, such as mathematical puzzles or general mathematical reasoning.
Multiple CoTs: generating several (independent) k CoTs, and returning the one with the best output, according to certain metrics.
Tree of Thoughts (ToT): it enhances Multiple CoTs by modeling the process of reasoning as a tree of thoughts. A single tree node represents a partial solution. Based on a given node, the thought generator constructs a given number k of new nodes. Then, the state evaluator generates scores for each such new node.
Explained in a more visual way:
Image taken from the paper “Graph of Thoughts: Solving Elaborate Problems with Large Language Models”
The design and implementation of GoT, according to the authors, consists of four main components: the Prompter, the Parser, the Graph Reasoning Schedule (GRS), and the Thought Transformer:
The Prompter prepares the prompt to be sent to the LLM, using a use-case specific graph encoding.
The Parser extracts information from the LLM’s thoughts, and updates the graph structure accordingly.
The GRS specifies the graph decomposition of a given task, i.e., it prescribes the transformations to be applied to LLM thoughts, together with their order and dependencies.
The Thought Transformer applies the transformations to the graph, such as aggregation, generation, refinement, or backtracking.
Finally, the authors evaluate GoT on four use cases -sorting, keyword counting, set operations, and document merging-, and compare it to other prompting schemes in terms of quality, cost, latency, and volume. The authors show that GoT outperforms other schemes, especially for tasks that can be naturally decomposed into smaller subtasks, are solved individually, and then merged for a final solution.
Summing up, another breath of fresh air in this hecticly evolving world of AI; this time combining abstract reasoning, linguistics, and computer sciences. Pas mal at all.
Much has been stated about the impact of Generative AI on the workforce for the past months, and many more pages will be written in the coming ones.
Last September, 15th 2023, appeared the paper “Navigating the Jagged Technological Frontier: Field Experimental Evidence of the Effects of AI on KnowledgeWorker Productivity and Quality” by, amongst others, Harvard Business School and MIT Sloan School of Management. Some months in advance -in June- McKinsey & Company published a meaningful report titled “The Economic Potential of Generative AI”
Some of the thoughts exposed in them are worthy to be pondered.
The paper by Harvard Business School and MIT Sloan School of Management examined the performance implications of AI on complex knowledge-intensive tasks, using controlled field experiments with highly skilled professionals. The experiments involved two sets of tasks: one inside and one outside the potential technological frontier of GPT-4, OpenAI large language model. The participants were randomly assigned to one of three experimental conditions: no AI, AI only, or AI plus overview. The “overview condition” provided participants with additional materials that explained the strengths and limitations of GPT-4, as well as effective usage strategies.
Understanding the implications of LLMs for the work of organizations and individuals has taken on urgency among scholars, workers, companies, and even governments. Outside of their technical differences from previous forms of machine learning, there are three aspects of LLMs that suggest they will have a much more rapid, and widespread impact on work.
The first is that LLMs have surprising capabilities that they were not specifically created to have, and ones that are growing rapidly over time as model size and quality improve. Trained as general models, LLMs nonetheless demonstrate specialist knowledge and abilities as part of their training process and during normal use.
The general ability of LLMs to solve domain-specific problems leads to the second differentiating factor of LLMs compared to previous approaches to AI: their ability to directly increase the performance of workers who use these systems, without the need for substantial organizational or technological investment. Early studies of the new generation of LLMs suggest direct performance increases from using AI, especially for writing tasks and programming, as well as for ideation and creative work. As a result, the effects of AI are expected to be higher on the most creative, highly paid, and highly educated workers.
The final relevant characteristic of LLMs is their relative opacity. The advantages of AI, while substantial, are similarly unclear to users. It performs well at some jobs, and fails in other circumstances in ways that are difficult to predict in advance.
Taken together, these three factors suggest that currently the value and downsides of AI may be difficult for workers and organizations to grasp. Some unexpected tasks (like idea generation) are easy for AI, while other tasks that seem to be easy for machines to do (like basic math) are challenges for some LLMs. This creates an uneven blurred frontier, in which tasks that appear to be of similar difficulty may either be performed better or worse by humans using AI.
The future of understanding how AI impacts work involves understanding how human interaction with AI changes depending on where tasks are placed on this frontier, and how the frontier will change over time. The current generation of LLMs are highly capable of causing significant increases in quality and productivity, or even completely automating some tasks, but the actual tasks that AI can do are surprising and not immediately obvious to individuals or even to producers of LLMs themselves, because this frontier is expanding and changing. According to the research performed, on tasks within the frontier AI significantly improved human performance. Outside of it, humans relied too much on the AI and were more likely to make mistakes. Not all users navigated the uneven frontier with equal adeptness. Understanding the characteristics and behaviors of these participants may prove important, as organizations think about ways to identify and develop talent for effective collaboration with AI tools.
Two predominant models that encapsulate human approach to AI could be identified, according to the Harvard and MIT research. The first is Centaur behavior: this approach involves a similar strategic division of labor between humans and machines closely fused together. Users with this strategy switch between AI and human tasks, allocating responsibilities based on the strengths and capabilities of each entity. They discern which tasks are best suited for human intervention and which can be efficiently managed by AI. The second approach is Cyborg behavior. In this stance users don’t just delegate tasks; they intertwine their efforts with AI at the very frontier of capabilities.
Upon identifying tasks outside the frontier, performance decreases were observed as a result of AI. Professionals who had a negative performance when using AI tended to blindly adopt its output and interrogate it less. Navigating the frontier requires expertise, which will need to be built through formal education, on-the-job training, and employee-driven up-skilling. Moreover, the optimal AI strategy might vary based on a company’s production function. While some organizations might prioritize consistently high average outputs, others might value maximum exploration and innovation.
Overall, AI seems poised to significantly impact human cognition and problem-solving ability. Similarly to how the internet and web browsers dramatically reduced the marginal cost of information sharing, AI may also be lowering the costs associated with human thinking and reasoning.
The era of generative AI is just beginning. Excitement over this technology is palpable, and earlypilots are compelling. It is extremely complex to forecast the adoption curves of this technology. According to historical findings, technologies take 8 to 27 years from commercial availability to reach a plateau in adoption. Some argue that the adoption of generative AI will be faster due to the ease of deployment. For McKinsey, it will be needed a minimum of eight years -in its earliest scenario- for reaching a global plateau in adoption.
One final note about the labor impact of GenAI: economists have often noted that the deployment of automation technologies tends to have the most impact on workers with the lowest skill levels -as measured by educational attainment-, or what is called skill biased. From the McKinsey analysis, we can reach the conclusion that generative AI has the opposite pattern—it is likely to have the most incremental impact through automating some of the activities of more-educated workers:
Notwithstanding, a full realization of the technology’s benefits will take time, and leaders in business and society still have considerable challenges to address. These include managing the risks inherent in generativeAI, determining what new skills and capabilities the workforce will need, and rethinking core business processes such as retraining and developing new skills.
CICERO is an AI agent that can use language to negotiate, persuade, and work with people to achieve strategic goals similar to the way humans do. It was the first AI to achieve human-level performance in the strategy game No-press Diplomacy.
No-press Diplomacy is a complex strategy game, involving both cooperation and competition, that has served as a benchmark for multi-agent AI research. It is a 7-player zero-sum cooperative/competitive board game, featuring simultaneous moves and a heavy emphasis on negotiation and coordination. In the game a map of Europe is divided into 75 provinces. 34 of these provinces contain supply centers, and the goal of the game is for a player to control a majority (18) of the SCs. Each players begins the game controlling three or four supply centers and an equal number of units. Importantly, all actions occur simultaneously: players write down their orders and then reveal them at the same time. This makes Diplomacy an imperfect-information game in which an optimal policy may need to be stochastic in order to prevent predictability.
Diplomacy is a game about people rather than pieces. It is designed in such a way that cooperation with other players is almost essential to achieve victory, even though only one player can ultimately win. It requires players to master the art of understanding other people’s motivations and perspectives; to make complex plans and adjust strategies; and then to use natural language to reach agreements with other people and to persuade them to form partnerships and alliances.
How Was Cicero Developed by FAIR?
In two-player zero-sum (2p0s) settings, principled self-play algorithms ensures that a player will not lose in expectation regardless of the opponent’s strategy, as exposed by John von Neumann in 1928 in his work Zur Theorie der Gesellschaftsspiele.
Theoretically, any finite 2p0s game -such as chess, go, or poker- can be solved via self-play given sufficient computing power and memory. However, in games involving cooperation, self-play alone no longer guarantees good performance when playing with humans, even with infinite computing power and memory. The clearest example of this is language. A self-play agent trained from scratch without human data in a cooperative game involving free-form communication channels would almost certainly not converge to using English, for instance, as the medium of communication. Owing to this, the afore-mentioned researchers developed a self-play reinforcement learning algorithm -named RL-DiL-piKL-, that provided a model of human play while simultaneously training an agent that responds well to this human model. The RL-DiL-piKL was used to train an agent, named Diplodocus. In a 200-game No-press Diplomacy tournament involving 62 human participants, two Diplodocus agents both achieved a higher average score than all other participants who played more than two games, and ranked first and third according to an Elo rating system -a method for calculating the relative skill levels of players in zero-sum games.
Which Are the Implications of this Breakthrough?
Despite almost silenced by the advent of GPT in its different versions, firstly this is an astonishing advance in the field of negotiation, and more particularly in the realm of diplomacy. Never an AI model has had such a brilliant performance in a fuzzy environment, seasoned by information asymmetries, common sense reasoning, ambiguous natural language, and statistical modeling. Secondly and more importantly, this is another evidence we are in a completely new AI era in which machines can and are scaling knowledge.
These LLMs have caused a deep shift: we went from attempting to encode human-distilled insights into machines to delegating the learning process itself to machines. AI is ushering in a world in which decisions are made in three primary ways: by humans (which is familiar), by machines (which is becoming familiar), and by collaboration between humans and machines (which is not only unfamiliar but also unprecedented). We will begin to give AI fewer specific instructions about how exactly to achieve the goals we assign it. Much more frequently we will present AI with ambiguos goals and ask: “How, based on your conclusions, should we proceed?”
AI promises to transform all realms of human experience. And the core of its transformations will ultimately occur at the philosophical level, transforming how humans understand reality and our roles within it. In an age in which machines increasingly perform tasks only humans used to be capable of: what, then, will constitute our identity as human beings?
With the rise of AI, the definition of the human role, human aspirations, and human fulfillment will change. For humans accustomed to monopoly on complex intelligence, AI will challenge self-perception. To make sense of our place in this world, our emphasis may need to shift from the centrality of human reason to the centrality of human dignity and autonomy. Human-AI collaboration does not occur between peers. Our task will be to understand the transformations that AI brings to human experience, the challenges it presents to human identity, and which aspects of these developments require regulation or counterbalancing by other human commitments.
The AI revolution has come to stay. Unless we develop new concepts to explain, interpret, and organize its consequent transformations, we will be unprepared to navigate them. We must rely on our most solid resources -reason, moral and ethical values, tradition…- to adapt our relationship with reality so it keeps on being human.
