Digesting AI Doomerism

A little late to the party, given this week's attention to the banking crisis, I recently listened to Eliezer Yudkowsky's fascinating but unnerving interview with Bankless about his concerns regarding the potential dangers of artificial general intelligence (AGI). He expressed pessimism about the technology's potential to cause extinction-level disaster, arguing that humanity does not possess the knowledge and algorithm infrastructures to program frameworks of meaning and value into AGI that would align with human interests. I took away the sense that Eliezer's frustrations are primarily rooted in our present process and culture around the development of AGI, specifically the frameworks of meaning, value, and purpose that we infuse into our creations. Under the current regimes of thought, AGI behaviors will align with whatever it believes to be ideal for its proliferation, irrespective of human well-being or desires.

Here we identify the root cause of Eliezer's pessimism around the trajectory of AGI, our inheritance of frameworks of meaning and value that fail to meet the needs of the moment. The most advanced AGI research and productive efforts today all occur within exclusive institutional silos, each with distinct intentions and motives. Unsurprisingly, the most dominant action framework aligns with the intent to profit. A company prioritizing profits over collective safety and prudence is an unsurprising strategy given our inhabiting legal, political, and economic systems prioritize transaction as the highest form of cooperation. Eliezer's disappointment with companies like OpenAI is in releasing these models into the wild, pushing AGI to market as soon as possible to scale users akin to the traditional SaaS models.

Doomerism around runaway AGI is legitimate, but how it will align remains unanswered. When we consider this from our individual and collective circumstances, the most crucial question regarding the direction of our creations is (and always has been) who are we? All human systems bear an imprint of the moment of their creation, the intent they serve, and the frameworks of meaning and value their creators embody. In many ways, our creations are expressions of immortality, infusing specific patterns of thought and action into their users and the world for as long as they remain in use. In a global society dominated by hierarchal forms of legal, economic, political, and spiritual regimes, an AGI whose actions would cause harm to humanity is the most obvious of scenarios. After all, commodifying individuals is a standard practice of almost every political and corporate regime today. In a world of perpetual war for profit, birth lottery is the primary determinant of personal access and agency. If we do not recognize the inherent divinity of the individual, why would an AGI?

As with most of our struggles, artificial general intelligence is a crisis of our own making. One rooted in the frameworks of meaning and value that we have inherited, the dogmas we identify with, and a historical lack of viable alternatives. If we desire to most fully insulate ourselves from creating overlords who eliminate us, we must turn our attention inward. To embrace new frameworks of meaning and value that better meet the needs of the moment, informed by the natural authorities governing our relationships with the universe and others. While AGIs will be capable of self-learning and direction, humanity will ultimately be responsible for the trajectory it takes at the moment of its inception. Whatever it is, it will be an extension of us. To that end, proactive soul crafting in a direction that celebrates human life instead of is one of the most important steps we can take to move the needle toward creating a benevolent partner.

Ron Rivers is the author of Self-Actualization in the Age of Crisis a member of SpiritDAO Cohort 0.

SpiritDAO Weekly Progress Summary

SpiritDAO's official membership application is now live.
- Learn more about the details of Cohort 0 and Cohort 1.
Check out Ron Rivers' brief talk at the DAOPlanet event at ETH Denver.
SpiritDAO will be leveraging Unlock Protocol to facilitate our membership tokens.
The SpiritDAO x Surreals Collaboration concluded, with 23 books being prepped for mailing to the Surreals community members. Overall it was a great success and helps further SpiritDAO's core purpose of spreading the message.