Curated resources to move from possibility to action.
The Turing Test, first introduced in the 1950s, offered a benchmark and foundation for assessing “artificial intelligence.” In 2014, a computer passed the test and successfully convinced an investigator that it was human. A few years later, Luminary Labs published its first AI explainer to separate hype from reality. The arrival of ChatGPT in late 2022 sent shockwaves across sectors, but we’ve been building toward this current moment for decades.
While AI is not new, the ubiquity of access and integration, the pace of change, and the potential impact for transformation are unlike anything we’ve experienced before. AI is suddenly everywhere, and it’s deeply integrated into everyday products and services. If you’re using a smartphone, posting on social media, shopping online, or watching movies on a streaming service, you’re using AI. Other technologies (electricity, cars, the internet) have similarly changed civilization, but this time around, the implications for life, work, and society are even more profound. It’s easy for an executive to feel overwhelmed, but it’s also increasingly easy to see the potential for transformation.
AI is here to stay, and the opportunities are massive. The transformation is well underway — but it’s not too late to pay attention and get up to speed. We’ve curated a selection of explainers that can serve as a starting point for conversations, experiments, and fresh thinking about our AI-shaped future.
What is AI?
Artificial intelligence describes technology that can perform tasks typically requiring human intelligence. Unlike traditional software, modern AI systems now have the ability to interpret language, generate creative content, and extract insights from vast datasets. These capabilities represent a fundamental shift in how technology functions and interacts with humans.
For organizations, AI offers powerful ways to transform operations, personalize experiences, and accelerate innovation. But the most successful implementations don’t just focus on the technology: They carefully consider how AI changes workflow, decision-making processes, and the human experience of work to solve problems and address real needs.
What is AI (artificial intelligence)? (McKinsey & Company, 2024) Understanding the history and origin of some fundamental AI capabilities is helpful for understanding how the technology is being deployed today. This explainer defines key concepts including generative AI, large language model (LLM), deep learning, and artificial general intelligence (AGI).
Transformer: a novel neural network architecture for language understanding (Google Research, 2023) This seminal 2017 paper is credited with introducing the transformer model that revolutionized natural language processing and other AI applications; it has become the foundation for most modern language models (like GPT, BERT, and T5), enabling breakthrough capabilities in language understanding and generation.
AI as normal technology (Knight First Amendment Institute, 2025) From the authors of “AI Snake Oil” — an early cautionary stance around the societal impacts of AI — this essay offers a generative technical and cultural critique of AI discourse. While AI differs from past technical innovations in key ways, it shares enough similarities with other technologies that established patterns like diffusion theory should apply unless proven otherwise.
What we talk about when we talk about AI (Careful Industries, 2025) In this provocative essay, Rachel Coldicutt insists that understanding AI’s social impacts is crucial; more than just a technology, AI is transforming society on macro and micro levels.
There is no AI (The New Yorker, 2023) Jaron Lanier, a Microsoft researcher who coined the term VR, offers a unique perspective on artificial intelligence. He argues that what is commonly referred to as AI is essentially a novel form of social collaboration, trained on human-developed content, rather than an independent human-like intelligence.
What about generative AI? The relatively recent rise and immense interest in generative AI has led some to treat “genAI” as synonymous with “AI” in public discourse, despite being just one subset of the broader field. Generative AI refers specifically to artificial intelligence systems designed to create new content — such as text, images, audio, code, or video.
Explained: generative AI (MIT News, 2023) “What do people actually mean when they refer to generative AI?” Experts from MIT answer this question, providing an overview of generative AI, how it works, how it’s different from other types of AI, and what the future may hold.
The cybernetic teammate: a field experiment on generative AI reshaping teamwork and expertise (SSRN, 2025) Researchers from Harvard Business School conducted a study with 776 professionals at Procter & Gamble to examine the impact of artificial intelligence on collaboration. The findings indicate that widespread AI adoption in knowledge work significantly alters how employees collaborate, necessitating a reevaluation of collaborative work structures.
How people are really using gen AI in 2025 (Harvard Business Review, 2025) As the AI hype continues to grow, Marc Zao-Sanders looks at how people are actually using generative AI in 2025 and ranks the top 100 use cases. Since the first report a year ago, the most popular uses have shifted toward personal and professional support and away from technical assistance and troubleshooting.
Strategies for an accelerating future (Ethan Mollick, 2024) Wharton School professor Ethan Mollick offers leaders four questions they should ask themselves when thinking about the rapidly evolving nature of AI. For a deeper exploration of how AI might affect our daily lives and work, consider reading Mollick’s 2024 book, “Co-intelligence: living and working with AI.”
What is applied AI?
When AI moves from theoretical research to solving real-world problems, it becomes applied AI. Unlike theoretical AI research that focuses on advancing fundamental concepts and algorithms, applied AI takes existing AI capabilities and puts them to work in specific contexts to create tangible value and solutions. In medical fields with workforce shortages, AI-powered workflow optimization and automated documentation systems could significantly reduce administrative burden, allowing existing staff to focus on direct patient care. In media and entertainment, AI tools can enhance creative workflows, personalize content at scale, and unlock new business models. In education, AI-powered tutors can extend quality, personalized learning to help students succeed.
As technology advances and more tools become commercially available, organizations must discern how to deploy AI to achieve business goals in a rapidly changing technology environment. Successful applied AI requires understanding both technical possibilities and organizational realities, including workflow integration, data infrastructure, staff capabilities, and ethical considerations.
Resources in this section bring to life the ways that AI is transforming organizations across a few of Luminary Labs’ focus areas.
Artificial intelligence in healthcare: transforming the practice of medicine (Future of Healthcare Journal, 2021) This 2021 journal article, still relevant today, reviews AI breakthroughs in healthcare and provides a roadmap for building effective, reliable, and safe AI systems, highlighting how advances in artificial intelligence can transform healthcare by enabling more personalized, precise, predictive, and portable medical approaches.
Accelerating scientific breakthroughs with an AI co-scientist (Google Research, 2025) This article introduces “AI co-scientist,” a multi-agent AI system built with Gemini 2.0 that functions as a virtual scientific collaborator, helping scientists generate novel hypotheses and research proposals to accelerate scientific and biomedical discoveries by leveraging AI’s ability to synthesize complex subjects and perform long-term planning and reasoning.
AI for climate and nature: landscape assessment (Columbia University’s Climate School and Engineering School) This broad landscape assessment examines how artificial intelligence can help address urgent global challenges like biodiversity loss and climate change by enhancing our understanding of these complex problems and accelerating the implementation of impactful, scalable, and ethical solutions in climate and nature conservation.
Artificial intelligence in education (National Education Association, 2024) Artificial intelligence is a transformative force for education. This NEA report documents the current AI uses and charts the future possibilities for educational transformation.
What is agentic AI?
Agentic AI describes systems that can independently complete multistep tasks, without human guidance, to achieve broader goals. Unlike basic AI that responds only to specific prompts, an AI agent can assess a human user’s intent, determine the necessary steps, and take actions across different systems with minimal supervision. For organizations, agentic AI promises to transform how work gets done. Teams can delegate entire workflows rather than just isolated tasks, freeing humans to focus on creative thinking, relationship-building, and strategic direction.
To fully leverage AI agents, organizations will first need to build the supportive infrastructure encompassing data systems, AI tools, workflow documentation, and appropriately skilled humans to guide and enable them. The promise of AI agents highlights that successful AI transformation hinges as much on systems and infrastructure development as on technological advancements.
Leading AI companies have operationalized agent concepts in their systems and products. The guides and developer tools from Anthropic and OpenAI offer a sense of how commercially available LLMs are making agentic tools available to their users. AI startups like Adept AI focus specifically on building AI systems that can take actions in software tools on behalf of users, and researchers are also building specialized agents in pursuit of a future “AI scientist.”
Navigating the AI frontier: a primer on the evolution and impact of AI agents (World Economic Forum, 2024) This white paper explains AI agents as autonomous systems that can sense, learn, and act on their environments, exploring their emergence from large language and multimodal models and highlighting their potential to enhance efficiency across industries.
AI agents in 2025: Expectations vs. reality (IBM, 2025) With tech media declaring 2025 as the” year of the AI agent” and predicting these technologies will transform both personal and professional aspects of human work, this IBM brief explores four narratives that could shape AI agent adoption in the year ahead.
A generational infrastructure buildout might hinge on AI agents (Goldman Sachs, 2025) Frank Long argues that the significant investment in AI infrastructure requires the development of complex, agentic AI systems for long-term viability, as simpler uses of AI (for example, chat experiences) may not justify the necessary infrastructure costs.
Will AI agents lead to freedom or surveillance? (Project Liberty, 2024) AI agents present a potential duality: They could increase individual autonomy while simultaneously enabling greater surveillance and control. This introduces societal considerations regarding data privacy, the spread of misinformation, and the necessity of regulatory frameworks.
Risks, governance, and ethics
AI governance can encompass the frameworks, policies, and practices organizations establish to ensure AI systems operate effectively, ethically, and in alignment with organizational values and societal expectations. As AI’s capabilities grow, so does the importance of thoughtful governance. AI developers on the frontier of model development must confront complex ethical dilemmas related to using training data, ensuring model alignment, and managing responsible user access — all while navigating intense global competition with significant geopolitical and economic stakes. Furthermore, developers and users cannot overlook the significant environmental impacts posed by data centers and heavy compute demands of AI.
We’ve written previously about ethical and policy questions specific to generative AI; while some of what was true two years ago remains relevant, new concerns and questions are arising — and will continue to surface — as AI reaches further into work and life.
What is AI governance? (IBM, 2024) This explainer and associated ebook educates readers about AI governance through practical examples and key principles, emphasizing the need for oversight that aligns AI behaviors with ethical standards.
Balancing innovation and governance in the age of AI (World Economic Forum, 2024) Building on the work of the World Economic Forum’s AI Governance Alliance, this article outlines a framework with three key pillars that identify critical areas where policymakers and regulators must focus to ensure resilient and adaptable AI governance.
What will remain for people to do? (Knight First Amendment Institute, 2025) This essay explores what paid work might remain for humans in a world where AI outperforms people in all economically useful tasks, arguing that employment opportunities will persist due to three constraints: general equilibrium limits, preference limits, and moral limits.
SAFE Benchmarks Framework (EDSAFE AI Alliance) This framework, backed by an industry council that includes Luminary Labs, is an example of how collaboration among diverse entities can produce actionable benchmarks to inform AI development and deployment. The alliance is working to realize a safer, more secure, and more trusted AI education ecosystem.
