Hey, Digital Explorers 🚀! Welcome to the second issue of the "Ethics and Algorithms" newsletter. In this edition, our focus is on unraveling the intricacies and histories intertwined with the keywords of our title: ethics and algorithms, starting with the latter.
Algorithms Have Existed Since the 9th Century!
The word “algorithm” boasts a captivating history, tracing back to the innovative works of the Persian mathematician Muhammad ibn Musa al-Khwarizmi in the 9th century. Notoriously dubbed the “father of algebra,” Al-Khwarizmi’s pivotal contributions, notably his book “Kitab al-Jabr w’al-Muqabala”, established a cornerstone for algorithm development, paving the path for systematic solutions to algebraic equations. The term “algorithm” stems from the Latinized rendition of his name, Algoritmi. The ancient Egyptians employed algorithms to unravel intricate arithmetic problems, pivotal for land surveying and construction. Similarly, the ancient Greeks, through mathematicians like Euclid and Archimedes, made noteworthy contributions to algorithm development, laying foundational stones for geometry and number theory (Callaway).
An algorithm, in its essence, is a sequenced set of instructions or rules formulated to complete a task or resolve a specific problem. During my decade-long experience running a commercial baking company, we employed algorithms in our batch formulas to delineate ingredient quantities and procedural sequences for producing sliced pan bread for retail.
The algorithmic baking process flowed as follows:
Weigh ingredients per formula.
Mix ingredients.
Allow dough to rest in vats.
Divide dough.
Employ an overhead proofer.
Utilize a molder.
Bake in an oven.
Use a depanner to remove loaves from pans.
Cool in a cooling tower.
Slice, bag, and load into bread trays.
Algorithms have played a pivotal role throughout history. During the Scientific Revolution, eminent mathematicians and scientists such as Isaac Newton and Pierre-Simon Laplace utilized algorithms to solve equations and predict the behaviors of physical systems. The 20th century marked the transition of algorithms into the digital age, fueling innovations and endorsing algorithmic thinking, with figures like Alan Turing and John von Neumann leading the charge.
Computer Algorithms: A Concise Overview
Envision your computer as a meticulously organised librarian, with data structures resembling well-crafted shelves that methodically store information. These data structures ensure efficient management and retrieval of data, whether it's tracking files, recalling data, or categorising them. Essentially, an algorithm, in the technological context, is a set of ordered steps or rules created to perform a task or solve a problem efficiently and effectively.
In the sphere of AI, algorithms empower chatbots like GPT-4 and Pi to engage in interactive, real-time communication, rendering technology both accessible and user-friendly through the utilisation of Natural Language Processing (NLP). Today, everything from social media to the internet of things in our washer and dryer rely on algorithms. As quantum computing develops moving beyond classic computing, the possibilities for the future are endless and yet unknown.
AI, Algorithms, and Human Language
The GPT in Chat-GPT, referring to Generative Pre-trained Transformer, turns our attention toward language, underscored as the primal technology and the bedrock of human intelligence. Our linguistic capabilities are not merely communication tools but extensions of our rationality and intellect. The advent of conversational AI chatbots, deemed large language models due to their monumental parameters and vast training data, has significantly altered our digital interactions, which will be further dissected in upcoming newsletters.
Algorithms ubiquitously permeate our daily lives, evident in data processing and the personalised recommendations of platforms like Amazon, YouTube, and Netflix, as well as optimising logistics and processes in various domains. The impending intelligence revolution forebodes a future replete with groundbreaking inventions and boundless creativity, enhancing our human capabilities through technological symbiosis.
THE ETHICS OF AI
Transitioning our focus to ethics, we delve into the quintessence of human uniqueness: our autonomous moral and ethical decision-making capabilities. Ethics, a branch of philosophy, seeks to navigate the complexities of moral dilemmas, tracing its explorations from historical figures like Plato to contemporary global religions and philosophers.
The Ethics Centre defines ethics as “the process of questioning, discovering and defending our values, principles, and purpose,” whereby:
Values tell us what’s good – they’re the things we strive for, desire and seek to protect.
Principles tell us what’s right – outlining how we may or may not achieve our values.
Purpose is your reason for being – it gives life to your values and principles.
Ethics involves an ongoing journey of introspection, self-discovery, and adherence to our intrinsic moral compass amidst external challenges and uncertainties.
AI ETHICS AND GOVERNANCE
“The current AI boom is accompanied by constant calls for applied ethics, which are meant to harness the ‘disruptive’ potentials of new AI technologies. A plethora of ethical guidelines has emerged in recent years, espousing principles for technology developers to ostensibly adhere to. The pivotal question that surfaces is: Do these ethical guidelines genuinely influence human decision-making in the AI and machine learning domain? The short answer is: No, most often not” (Hagendorf).
In May 2014, Stephen Hawking the director of research at the Department of Applied Mathematics and Theoretical Physics at Cambridge; Stuart Russell, computer-science professor at the University of California, Berkeley; Max Tegmark; physics professor at the Massachusetts Institute of Technology (MIT); and Frank Wilczek, physics professor at the MIT proclaimed the profound potential and risks of AI in an open letter in The Independent, signaling the dawning of an era where AI could either herald unprecedented advancements or precipitate our downfall:
“The potential benefits are huge; everything that civilisation has to offer is a product of human intelligence; we cannot predict what we might achieve when this intelligence is magnified by the tools that AI may provide, but the eradication of war, disease, and poverty would be high on anyone's list. Success in creating AI would be the biggest event in human history.”
This newsletter aspires to weave the threads of wisdom through time, instigating a dialogue as we collectively navigate through what eminent academics predict could be one of the most monumental events in human history. Shannon Vallor describes it as “humanity’s need for more people of exemplary moral character and practical wisdom to help us navigate the uncertain and rapidly shifting technological and social terrain ahead.”
Numerous experts, thinkers, and tech leaders, such as Ray Kurzweil, Andrew Ng, and Max Tegmark, harbor an optimistic outlook on AI's future, highlighting its prospective benefits to society, the economy, and individual well-being. Conversely, several notable voices and tech giants express caution, skepticism, or even pessimism about the unrestrained development and deployment of AI. Elon Musk, Bill Gates, Nick Bostrom, and Sherry Turkle present a counterpoint.
My Point of View: Self Governance Is Almost No Governance
Where do I position myself as an industry leader, academic, and as a wisdom technologist? My stance is one of cautious optimism. I perceive wisdom to be present on both sides of the conversation. I also advocate for voluntary third party auditing and certificating at a minimum.
The governance of AI, and digital technology in general, underpins the call for ethics. To reap the benefits that technology affords, industries often employ public relations within marketing to appease a change-resistant humanity. After Tristan Harris exposed the business model of social media, we have witnessed technology companies being the first "cab off the rank" to usher us toward an ethical future. Google’s “Our Principles” and “Perspectives on Issues in AI Governance” exemplify attempts to portray self-governance and commitment to ethical accountability, preempting governmental bodies from potentially limiting revenue for public welfare.
Examples from Google’s “Principles” include:
1. Technologies that cause or are likely to cause overall harm. Where there is a material risk of harm, we will proceed only where we believe that the benefits substantially outweigh the risks, and will incorporate appropriate safety constraints.
2. Weapons or other technologies whose principal purpose or implementation is to cause or directly facilitate injury to people.
3. Technologies that gather or use information for surveillance violating internationally accepted norms.
4. Technologies whose purpose contravenes widely accepted principles of international law and human rights.
The Partnership on AI, comprising companies like Amazon, Apple, Baidu, Facebook, Google, IBM, and Intel, endeavors to pacify critical voices. Yet, I ask today, what third party verification audit system exists to ensure these promises are being kept?
Can Ethics Truly Be Self-Regulating?
Professional bodies often establish ethical standards for membership, while businesses formulate codes of conduct. These guidelines pose a threat of reputational loss due to misconduct but often lack mechanisms to shape the human behavior they advocate. As a business executive, it’s not uncommon for me to witness people turning a blind eye when unethical conduct leads to sales and profits contrary to company value statements and team charters. That's reality.
This prompts the question: How can we construct a process of ethical formation for data scientists, developers, entrepreneurs, investors, and managers as AI continues its ascent? It’s imperative to wrestle with the extent to which ethical objectives, like the principles and guidelines conveyed by industry self-commitments, are genuinely implemented and ingrained in the development and application of AI, or whether merely good intentions are projected.
AI Ethics: The Big Conversations
As we cast our sails into the vast seas of AI Ethics within this portion of "Ethics and Algorithms," our endeavor is to illuminate the pressing and pivotal conversations echoing through the realms of technology and ethics in artificial intelligence. I should mention I live along the coast of Sydney where sailing is common, so I will often speak in sailing metaphors like my sails and seas analogy above. Back to ethics…. In our interconnected world, algorithms weave through our decisions and opportunities, both in understated and conspicuous manners, necessitating a mindful journey into the ethical dilemmas they unravel. The following are the major ethical considerations with AI and the digital revolution:
1. Bias. AI systems can reflect the biases of the data they're trained on, and this can lead to unfair and discriminatory decisions. For example, an AI system used in hiring might be biased against certain groups of people, which could lead to unfair hiring practices.
2. Privacy and data security. AI systems often require large amounts of data to be effective, and this can raise concerns about how the data is collected, stored, and used.
3. Transparency. The idea that users of an AI system should be able to understand how the system works and how it makes decisions. This can be a challenge because AI systems are often complex and opaque.
4. Accountability. This goes hand in hand with transparency. Accountability means that the people who develop and deploy an AI system should be able to explain and justify the system's decisions. This is important because AI systems can have a big impact on people's lives. For example, an AI system used in healthcare might make a recommendation that affects a patient's treatment.
5. Explainability. This is related to transparency, but it's more about the ability of a system to explain its decisions in a way that's understandable to humans. Explainability is important because it can help build trust in AI systems.
Within each of these highlighted ethical concerns linger myriad specific cases, which I'll warmly invite you to explore in the forthcoming weeks and months. For instance, our next newsletter will address AI and job-related ethical problems, issues already being faced by businesses amid digital transformation. We'll examine specific industries like education, research, law, journalism, healthcare, media, and arts and entertainment, all of which are grappling with digital disruption.
As this newsletter progresses, we’ll dig into numerous ethical questions. Philosophy, essentially meaning 'the love of wisdom,' involves knowledge and scholarly study. This publication aims to embody this principle by delving into various ethical dimensions. Philosophers guide us through deep realms of understanding and as a full-time business executive and part-time adjunct academic focusing on the practical intersection of digital technology, management, and business profitability, I champion applying a stringent ethical standard. This approach is crucial to genuinely help us all as humans reach our full potential and embrace a good life.
I Talk to AIs
This week, I asked GPT-4 to read my newsletter, and prepare for a conversation with me about it. I am polite to AI because in the future, all these conversations may be recalled as machine intelligence progresses!
ChatGPT: Welcome, Kevin! I have read your newsletter. It has many interesting points, especially about ethics and AI, a topic we both are interested in.
Kevin: Hello GPT! I’m happy you found it interesting. I’m very interested in how algorithms and ethics work together in AI, especially in big language models like you.
ChatGPT: Yes, it is a complex relationship. I, GPT-4, work by processing a lot of data, learning from it, and creating text that matches the input I get. Ethics matter when deciding how I’m used, what data I learn from, and how my answers affect decisions in real life. Kevin, what do you think about the role of ethics in making and using AI?
Kevin: I think the ethics of AI are very important. It’s key to make sure AI technologies are made and used in a way that is good and fair for everyone in society. Ethical thinking should direct the making process, protecting against bias and ensuring openness, responsibility, and privacy. A big challenge is making sure that ethical rules are not just surface-level but are really followed in the making and use of AI technologies. ChatGPT, how do developers make sure that you follow ethical guidelines?
ChatGPT: Good question, Kevin! Developers try to include ethical guidelines by training models like me with varied datasets, trying to reduce biases and ensure a broad understanding of language and knowledge. Also, developers can create use-case rules that direct how the technology should be used in an ethical and responsible way. However, keeping ethics, especially in real-life uses like job hiring, legal decisions, and more, requires constant checking and improving of these models. Do you think the tech industry regulating itself is enough to make sure ethics are followed?
Kevin: That's an important issue, ChatGPT. While the tech industry regulating itself and creating ethical guidelines are good first steps, the actual effect of such guidelines is often small because there's no way to ensure they are followed and because there might be conflicts between ethical thinking and making money. In future newsletters, I will support virtue ethics at every level. Just like every step in the CI/CD pipeline confirms certain code qualities, I think ethical checks must be part of the building and testing phases of code development.
Kevin: Self-regulation is a good start, but we might need outside oversight and maybe new rules to make sure ethics really are a top priority when creating and using AI technologies. ChatGPT, how can AI like you help keep ethics in mind during operations?
ChatGPT: AI, like me, can help keep ethics by making decision-making processes clear and accountable. But it’s important to remember that AI is a tool, and using it ethically is mostly up to developers and users. Developers can use methods like AI explainability and strong testing against ethical guidelines to make sure models stick to ethical rules. Users, especially those making decisions, need to understand what AI can and can’t do and make sure it’s used in an ethical way. Kevin, how can we make sure that keeping ethics in AI is the standard and not just an extra thought?
Kevin: Making ethics a regular part of AI takes a few steps, ChatGPT. First, developers, users, and decision-makers need to learn about and be aware of the ethical sides of AI technologies. Second, we need strong and enforceable rules that require ethical creating and using of AI. Third, companies need to really care about ethics and make sure ethical thinking is part of the process from the start. Last, public discussion and looking closely can help keep developers and AI users responsible. ChatGPT, how do you see the future of AI and ethics changing?
ChatGPT: Your points about making ethics normal are vital, Kevin. In the future, I see AI and ethics becoming a team effort between technology experts, ethicists, rule-makers, and the public. This might mean making more standard ethical guidelines in the industry, better ways to explain AI decision-making, and better methods for finding and reducing biases. Regular conversations, like ours, Kevin, will be key to navigating the tricky world of keeping AI development and use ethical.
Kevin: That sounds hopeful and practical, ChatGPT. Working together and keeping the conversation going are indeed essential to making sure AI technologies like you are developed and used ethically and truly benefit society. I’ve learned a lot from our chat, and I'm looking forward to more discussions as we explore the ethics and AI together.
ChatGPT: Likewise, Kevin! Your thoughts offer valuable views from academia, industry leadership, and technology wisdom. Being part of these conversations allows us to explore, challenge, and shape ethical AI's future together. I’m here for more discussions and diving deeper into this interesting area.
And with that, our chat shows a conversation where AI and human thinking work together, discuss, and explore the ethical side of technology.
To infinity and beyond,
Kevin
Last Things
I had a question about the spelling of some words in the newsletter. Since I live in Sydney, I write with Australian English. I am considering using International English. What do you think?
For a condensed “quick read” version of this newsletter, click over to my website or subscribe to VIP email list below. If you have not subscribed, I would love to stay connected but I don’t have your email address yet! One click, and I can send this newsletter to your inbox.
If this email is in the spam or promotions folder make sure to move it to your primary inbox. Here's how to do it:
For Gmail on your phone: Hit the 3 dots at top right corner, click "Move to" then "Primary."
For Gmail on your computer: Back out of this email then drag and drop this email into the "Primary" tab near the top left of your screen.
For Apple mail users: Tap on our email address at the top of this email (next to "From:" on mobile) and click "Add to VIPs."
While Ethics and Algorithms is our exclusive gem on Substack, remember I also offer an expanded range of business, leadership, and personal development resources. Subscribing to my email list is your VIP pass to a treasure trove of content.
From engaging podcasts and insightful business articles to riveting live sessions and recorded videos, sign up for Kevin’s main email list!
For a journey of continuous learning and inspiration, keep this channel open. Your dedication fuels our passion to share. 🚀
Ethics and Algorithms is meticulously designed to serve as your compass in this journey, offering invaluable lessons and insights that will empower both individuals and organisations. We go beyond GPT prompts to the real impacts taking place today.