Seeing Like an Algorithm: The Economic Value of Generative AI

Seeing Like an Algorithm: The Economic Value of Generative AI

In today's digital landscape, the success of companies often hinges on the capabilities of their machine learning algorithms. Whether it's TikTok's recommendation system or the AI-powered features of social networks like Facebook and Twitter, algorithms are becoming the driving force behind user experiences. Understanding how these algorithms work and how they can be optimized is crucial for businesses across all industries.

One of the key factors in algorithmic success is the training data it relies on. TikTok, for example, has a closed loop of feedback that allows its algorithm to continuously improve. Every action a user takes, from the moment a video starts playing, provides valuable signals that help the algorithm understand their preferences and interests. This feedback loop, combined with the use of relevant tags and labels, ensures that the algorithm can make accurate recommendations.

However, algorithm-friendly design doesn't always align with user preferences. Social networks like Facebook and Instagram use infinite scrolling feeds to provide a seamless browsing experience. While this approach reduces friction for users, it can also limit the algorithm's ability to accurately judge sentiment. By focusing on positive engagement, these platforms may miss out on signals of user disinterest, leading to a mismatch between the content shown and the user's true interests.

To truly serve users' interests, algorithm-friendly design needs to consider a different approach. The goal should be to help users achieve their desired outcomes, even if it means introducing some friction. By aligning every element of the design process with a single purpose and goal, companies can create a dataset that allows the algorithm to train itself into peak performance.

Generative AI, on the other hand, offers a different perspective on algorithmic power. It has the potential to transform industries ranging from language education to business operations. The economic value of generative AI is expected to be positively correlated with the median wage of each industry, driving a significant cost differential between traditional approaches and AI alternatives.

Historically, AI has struggled to build attractive business models in private markets. The escalating cost of progress often serves as an anti-moat, allowing fast-followers to close the gap at a fraction of the cost. Additionally, AI-powered solutions often rely on human intervention to ensure accuracy, which can be difficult to scale and burdensome on gross margins.

However, generative AI has shown tremendous potential for growth. Startups catering to emergent consumer needs have achieved unprecedented levels of adoption in a short period. For example, ChatGPT reached an estimated 230 million monthly active users in just six months, surpassing Facebook's early growth metrics. This level of adoption suggests that generative AI can fuel market transformations comparable to the advent of the microchip and the Internet.

Generative AI's applications are vast, encompassing content generation, companionship, and more. It has already been applied to industries like gaming, movies, music, and chat, with existing markets serving as proof points of value. However, the true potential lies in the emergence of entirely new behaviors and markets that are impossible to predict. As generative AI automates tasks that the human brain has spent less time evolving towards, it has the potential to perform these tasks cheaper, faster, and sometimes even better than humans.

The drop in the marginal cost of creation promised by generative AI will drive massive demand. The Jevons paradox, which has held true in the past, suggests that as the cost of a good decreases, the demand for it increases. This will lead to more jobs, economic expansion, and better goods for consumers. Just as the microchip brought the marginal cost of compute to zero and the Internet brought the marginal cost of distribution to zero, generative AI will bring the marginal cost of creation to zero.

In conclusion, understanding the inner workings of algorithms and harnessing the power of generative AI is essential for businesses looking to thrive in the digital age. To optimize algorithms, companies must focus on algorithm-friendly design that aligns with user interests while considering the potential for new behaviors and markets. Here are three actionable pieces of advice for businesses:

1. Embrace the feedback loop: Create a closed loop of feedback that allows your algorithm to continuously learn and improve from user interactions.

2. Prioritize accuracy over frictionless experiences: While reducing friction is important, it shouldn't come at the expense of accurate recommendations. Consider introducing some friction to ensure that the algorithm understands user sentiment accurately.

3. Explore the potential of generative AI: Look for opportunities to leverage generative AI in your industry. Consider the economic value it can bring and how it can transform existing markets or create entirely new ones.

By staying ahead of the curve and embracing the power of algorithms and generative AI, businesses can position themselves for success in an increasingly digital world.