The Extraction Age

“How did we get to the doorstep of the next leap in prosperity?” 

This question posed by Sam Altman in his 2024 blog post titled “The Intelligence Age” now reverberates with a deep sense of incongruity. The recent surge in information capital investment, posited as our key to the city, now occupies a place of significance in most discourse surrounding the American economic-political machine. But why is it that this new era has resulted in so many innovations seen as inconvenient or detrimental by the general American populace?

The statements heard from executives reflect a belief that we just need to be patient. Markets in automation and AI are heaving forward at unprecedented speeds, and while this new sector of the economy supplies a readily available commodity, inputs necessary for the continuation of advancement are a different case. The ability to create an AI capable of utilizing data and knowledge across a wide range of applications, otherwise known as artificial general intelligence, hinges on the availability of new chips and reliable energy infrastructure. 

Without immediate access, available action for tech progress becomes relegated to a status of producing various knick-knacks and doo-dads. In other words, we are watching the tech industry function like a slime mold, slowly expanding in different directions to feel out and test new paths. This perspective leaves us with a sense of ambiguity about our future, but is still the glass half full interpretation of the larger picture. Our worries will resolve themselves and in the meantime, all we need to do is get excited and brace ourselves for the dawn of the information age.

Unfortunately, for those of us familiar with the wishful thinking that permeates modern political rhetoric, it’s difficult to maintain such an optimistic view. The stock market rattles off new highs, seemingly unaffected by allegations of round-tripping between cloud providers like Oracle and companies like OpenAI. The big question is what exactly is being produced to generate such a high return on investment for holders of these stocks. From Google searches and online shopping to self-driving cars and robotic vacuums, AI is being inserted into virtually every tech and tech-adjacent product in an attempt to make progress in artificial intelligence net-profitable. But if the AI business model has so far failed to produce sustainable profit, what is the end goal of AI integration for companies like OpenAI, Google, or Amazon? Even further, why does the intelligence age feel more annoying and gimmicky than revolutionary? 

Daily social media and advertisements capture our attention and behavior and store them in databases to serve as training for new AI models and algorithms that feed catered information back to us in incomprehensible ways. The data is accessed by a handful of very large entities which has led to massive consolidation of market power and an interesting paradox. The ability to improve upon current AI models is only satisfiable through accessibility to infrastructure and data, and this itself requires so much capital that it is only available to pre-existing large companies. Market power on the other hand historically allows for corporations to avoid incentivizing innovation. The problems of innovation have been commented on by various figures in tech and are mainly related to the data problem. Models have become so large that they exhaust the world’s capacity for producing new information, leading to unintentional consumption of synthetic information output. 

The truth is that the technology industry is becoming less and less driven by traditional inputs like natural resources. Instead, it is being driven by attention capital, the ability to direct and control attention and extract value from it. The AI boom is not necessarily about creating real value, but is more so catered towards positioning companies to extract rent from the widespread adoption of artificial intelligence. Tech giants are aligning themselves in such a way as to make this new technology entirely ubiquitous. The solution to their data problem hinges on constant human-model interaction and is obtained from total integration. 

On a large scale, the shift towards rentier tendencies in technology signals a broader shift in the national economy. When the smoke clears, what remains is extensive infrastructure built in anticipation of an unstoppable expansion. Data centers, semiconductor supply chains, and oceans of data must be justified by profit without the excitement of exponential growth. This is the logic of the maintenance economy: when value can no longer be extracted from innovation, it must come from what already exists. Big tech’s shift towards rent-seeking might not be as temporary as we think, and may signal an adaptation to a perceived permanent condition. Revolutionary corporations, once focused on fundamentally changing our lives, will now compete to interject themselves into all political and economic realms to sustain infrastructure that has become too burdensome to abandon.

The main concern is that this sector of the economy now accounts for so much growth that if we don’t recognize the signs of stagnation or corporate overreach into our lives, it may end up being too late. The world itself, independent of tech, is always moving and we cannot afford to be left behind. To address these risks we need to recognize the dynamic for what it really is. Instead of simply assuming that we are on the precipice of untold prosperity, we should be skeptical of industry figureheads who claim to sell something that solves all of our problems. Once we do this, we can begin to ask ourselves the more important questions of who gains and who loses.