Late Summer Reading Recs -- The AI Edition

The dog days of summer are upon us. Sure, you could browse the NYT Bestsellers in bookstores, but as Labor Day weekend approaches, I’ve assembled a list of readings (both books and articles) on AI recommended by Rob May and me. Let’s dive in.

—oo—

My list includes:

Competing in the Age of AI Marco Iansiti and Karim Lakhani are professors at the Harvard Business School and wrote an awesome book about innovation and reinvention — which can be summed up by the title. It’s an indispensable resource for CEOs and board members of startups and growth stage AI companies. The book argues companies need to focus less on traditional assets and strategies, and instead build their competitive advantage through data, network effects, computational power, and self-reinforcing algorithms. 

Ethical Decision-Making in the Digital Age discusses how new technologies are creating novel ethical challenges that existing frameworks struggle to address adequately. Key emerging issues include privacy, accessibility, automation, biases in AI systems, and the spread of misinformation online. A core argument is that ethical thinking needs to adapt to guide decision-making in the digital age. Overall, the book calls for ethical, human-centered technology design and policies to align innovations with democratic values.

Unique Aspects of Building Deep Tech Businesses An excellent article that is indispensable for startup founders and ELT in the AI, data and related tech sectors. A short summary: deep tech solutions are based on cutting-edge scientific advances and engineering innovations, so developing them requires extensive R&D, multidisciplinary expertise, large capital investments, longer time horizons, and overcoming technical uncertainty.

What Is ChatGPT Doing ... and Why Does It Work? explains how ChatGPT works by using a deep learning model called a transformer to generate human-like text, trained on massive datasets to predict the next word in a sequence. It’s a great ChatGPT primer by Stephen Wolfram – a leading computer scientist, physicist, and founder and CEO of Wolfram Research where he works as chief designer of Mathematica and the Wolfram Alpha answer engine.

—oo—

Rob May, a close friend and angel investing partner, has over 25 years of experience founding and leading AI and big data companies, including serving as CEO of Pique AI and as founding CEO of Talla. He also writes an excellent SubStack called Investing In AI. Here are his suggestions:

Two Books

The AI-First Company: How to Compete and Win with Artificial Intelligence argues that companies need to put AI at the core of their business strategy in order to gain a competitive advantage, and provides a blueprint for companies to integrate AI throughout their organization to drive innovation, efficiency, and growth.

The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake Our World explores the quest to develop a master algorithm that can teach itself everything there is to know from data, outlining the different schools of machine learning and their limitations. Domingos proposes ways to integrate deep learning, evolution, Bayes, analogies, and other methods into one all-encompassing algorithm.

Articles

The Interplay of AI, Power, and Regulation advocates for an adaptive, collaborative approach between tech companies, governments, and civil society to develop ethical AI systems that enhance human agency.

The Economic Case for Generative AI and Foundation Models makes the case that generative AI models like ChatGPT represent a new general-purpose technology that will drive economic growth by enhancing productivity across many sectors, but realizing its full potential requires overcoming challenges like high compute costs and thoughtfully managing societal risks.

Does it Make Sense to Invest in a Foundation Model Startup? examines whether it still makes sense for companies to invest in developing proprietary AI systems versus utilizing public AI models like ChatGPT, arguing the value has shifted towards better dataset curation and integration.

Google "We Have No Moat, And Neither Does OpenAI" argues that the rise of large language models like ChatGPT, which can be easily accessed by anyone, represents an existential threat to Google's dominance, as proprietary AI and data is no longer a defensible moat. 

The two “not-to-be-missed” readings are “Unique Aspects” and Google "No Moat” articles.

Happy reading!

---