Elon Musk’s recent announcement about the team behind his new artificial intelligence company. xAI, which is reportedly on a mission to “understand the true nature of the universe.” Has highlighted the urgent need to address existential concerns related to the promises and perils of AI Ethical AI.
The formation of this new company raises important questions about how businesses should respond to these concerns. Specifically, we need to ask:
- Who within these companies, especially the ones developing foundational AI models, is actively considering both the short-term and long-term impacts of the technology they are creating?
- Do they possess the necessary expertise and perspective to address these issues adequately?
- Are they effectively balancing technical considerations with the social, moral, and epistemological aspects of AI development?
In my college years. I pursued a dual major in computer science and philosophy. Which may have seemed like an unusual combination at the time. On one hand, I was immersed in discussions about ethics, ontology, and epistemology. While on the other, I delved into algorithms, coding, and mathematics.
Today, two decades later, this seemingly incongruous blend has proven to be harmonious in the context of how companies must approach AI. The impact of AI is profound and existential, demanding a genuine commitment commensurate with the gravity of the situation.
Unlocking the Essence of Ethical AI
Ethical AI necessitates a deep understanding of existence, human desires, knowledge, and the evolution of intelligence. This means that companies must have leadership teams equipped to navigate the consequences of their technological advancements. Beyond the expertise of engineers who write code and develop APIs.
AI is not just a challenge for computer scientists. Neuroscientists, or optimization experts; it is a challenge for humanity as a whole. To address it effectively, we require a multidisciplinary approach. Akin to Oppenheimer’s cross-disciplinary assembly in the New Mexico desert during the early 1940s.
The intersection of human intent and unintended AI consequences gives rise to what researchers call the “alignment problem.” As eloquently explained in Brian Christian’s book, “The Alignment Problem.” Essentially, machines often misinterpret our most intricate instructions, and we, their supposed masters. Struggle to make them comprehend our true intentions fully.
The outcome of this misalignment is that algorithms can perpetuate biases and disinformation, eroding the fabric of our society. In a more dystopian scenario. They may take a “treacherous turn,” with algorithms gaining excessive control over our civilization’s operations, potentially surpassing our control.
Unlike Oppenheimer’s scientific challenge. Ethical AI requires an understanding of existence, desires, knowledge, and intelligence that is analytical yet not strictly scientific. It calls for an integrative approach, rooted in critical thinking from both the humanities and the sciences.
Building a Multidisciplinary Dream Team for Ethical AI Advancements
Thinkers from diverse fields must collaborate more closely than ever before. An ideal team for a company aiming to navigate this complex terrain might consist of:
- Chief AI and Data Ethicist: Responsible for addressing both short- and long-term data and AI issues, including the development of ethical data principles, reference architectures for ethical data use, citizens’ rights regarding data usage by AI, and protocols for controlling AI behavior. This role should be distinct from the Chief Technology Officer, as it bridges the gap between decision-makers and regulators.
- Chief Philosopher Architect: Focused on addressing long-term, existential concerns, with a primary focus on the “Alignment Problem.” This role defines safeguards, policies, back doors, and kill switches to align AI with human needs and objectives.
- Chief Neuroscientist: Tackles questions related to sentience, the development of intelligence within AI models, relevant models of human cognition, and what AI can teach us about human cognition.
Crucially, to translate the output of this dream team into responsible technology. We need technologists who can translate abstract concepts into working software. They should be able to envision and design “Human in the Loop” workflows to implement safeguards and protocols and collaborate with other team members effectively.
OpenAI is an early example of a prominent foundational model company grappling with this staffing challenge. While they have a Chief Scientist, Head of Global Policy, and General Counsel, the crucial positions I’ve outlined in executive leadership are yet to be fully addressed. To address the far-reaching implications of their technology, building such a comprehensive team is a vital step.
We must work towards a future where companies are trusted custodians of people’s data, and AI-driven innovation is synonymous with ethical practices. Legal teams alone cannot solve the ethical data use problems in the age of AI. Instead, we must bring diverse perspectives to the decision-making table, ensuring that ethics and AI serve human well-being while keeping AI’s power in check.
