AI and The Burden of Knowledge: Seeing Like an Algorithm — Remains of the Day

AI and The Burden of Knowledge: Seeing Like an Algorithm — Remains of the Day

As humans continue to innovate and push the frontier of knowledge further out, each successive generation takes longer to reach that frontier. This growing "burden" of knowledge is slowing down innovation, as great inventors have less time to actually innovate. However, AI, on the other hand, does not suffer from this burden. It does not die or degrade, it simply improves. This raises the question of what will happen when AI no longer needs us. As artificial intelligence accelerates past us, it will accumulate knowledge that we cannot comprehend.

Humans are the only species that pass significant knowledge from generation to generation. We rely on the wisdom of prior ages and build upon it. This mechanism is the essence of progress. But AI, with its ability to continuously improve, does not rely on this mechanism. It does not need to stand on the shoulders of giants to see further and innovate. It simply keeps getting better.

On a generational basis, animals make essentially no progress against the mountain of knowledge. They learn the same biological lessons over and over again. Humans, infected with the drive for knowledge and advancement, have a different relationship with the mountain of knowledge. Generation after generation, we tunnel further and learn more. But as economist Benjamin Jones argues, the age at which notable inventions occur is significantly increasing. The more time it takes someone to reach the end of the tunnel, the less time they have to actually dig.

AI, however, does not face this limitation. It does not experience lag or degradation in reaching the rockface. As our efforts decelerate, AI's efforts accelerate. It will be fully unleashed in the mountain of knowledge, leaving an ever-expanding gap between us and it.

But what about our understanding of the world? Will we become secondary creatures, incapable of comprehending the world we live in? AI may reveal wonders and accelerate our learning, but without ancillary support, we will be outmatched. We may benefit from the advancements brought by AI without truly grasping how they have arrived.

Understanding how algorithms achieve their accuracy is crucial in today's world. Competitors in all industries will be running up against a competitor whose advantage centers around a machine learning algorithm. While TikTok is an example in the short video space, the effectiveness of a machine learning algorithm is not solely based on the algorithm itself, but on the algorithm after being trained on a dataset.

The design of TikTok is a closed loop of feedback that inspires and enables the creation and viewing of videos on which its algorithm can be trained. Every action a user takes on the app serves as a signal to the algorithm. From the moment a video begins playing, the algorithm considers the user's sentiment towards that video. Relevant tags or labels are added by humans on TikTok's operations team before the video is even sent down to the user's phone. Vision AI is also invoked at the point of creation with filters designed to track human faces, hands, or gestures.

The default UI of our largest social networks today, such as Facebook or Twitter, is the infinite vertically scrolling feed. This design choice prioritizes lower friction scanning for users but sacrifices a more accurate read on negative signals. Positive engagement is more visible, leading to content derived from a social graph drifting away from a user's true interests.

Algorithm-friendly design takes a different approach to best serve the user's interests. It is not about minimizing friction but about helping the user achieve their desired outcome. This approach recognizes that true competitive advantages or moats in the software era are increasingly illusory. Features and UI designs can be easily copied, but the true magic lies in how every element of design and processes connect with each other to create a dataset with which the algorithm trains itself into peak performance.

In conclusion, the burden of knowledge and the rapid advancement of AI present both challenges and opportunities. While humans may be limited by time and the need to build upon prior knowledge, AI has the potential to surpass us and accumulate knowledge we cannot comprehend. However, understanding how algorithms work and designing with the algorithm in mind can lead to better outcomes for businesses and users alike. Here are three actionable advice:

1. Embrace AI as a tool for innovation: Rather than seeing AI as a threat, businesses should embrace it as a tool that can enhance innovation and accelerate progress. By leveraging AI's ability to continuously improve, companies can stay at the forefront of their industries.

2. Invest in AI education and training: As AI becomes increasingly prevalent, it's crucial for individuals and organizations to invest in education and training to understand how AI works and how to leverage its capabilities effectively. This will ensure that humans can work alongside AI and harness its potential.

3. Prioritize user-centric design: When designing products or services, consider how best to serve the algorithm that powers them. By aligning design choices with the algorithm's needs, businesses can create a better user experience and optimize the performance of their AI-driven systems.

By taking these actions, we can navigate the challenges and opportunities presented by the burden of knowledge and the rise of AI, ensuring that we continue to progress and innovate in this rapidly evolving landscape.