This post is about two stories — separated in time by perhaps 4,000+/- years. Whether historical or allegorical, the tale of the Tower of Babel offers a moral perspective worth considering amid the current global rush to artificial intelligence (AI).

the Cautionary Babel account

The biblical story emerges in the wake of massive flooding in the days of Noah. The narrative suggests that post-flood repopulation is accompanied by one large urban cluster previously unknown. So as the Hebrew scriptures tell it:

Now the whole earth had one language and the same words. And as they migrated from the east, they came upon a plain in the land of Shinar (in southern Mesopotamia) and settled there.

And they said to one another, “Come, let us make bricks and fire them thoroughly.” And they had brick for stone and bitumen for mortar.

Then they said, “Come, let us build ourselves a city and a tower with its top in the heavens, and let us make a name for ourselves; otherwise we shall be scattered abroad upon the face of the whole earth.”

The Tower of Babel Envisioned

The Lord came down to see the city and the tower, which mortals had built. And the Lord said, “Look, they are one people, and they have all one language, and this is only the beginning of what they will do; nothing that they propose to do will now be impossible for them.

Come, let us go down and confuse their language there, so that they will not understand one another’s speech.” So the Lord scattered them abroad from there over the face of all the earth, and they left off building the city.

Therefore it was called Babel, because there the Lord confused the language of all the earth, and from there the Lord scattered them abroad over the face of all the earth.

-- Genesis 11:1-9

The AI Connection

What in the world does an ancient and potentially mythical story have to do with today’s unprecedented technological and global rush to artificial intelligence (AI)? Parallels abound.

Babel and AI as experienced to date both share features of:

Massive leaps and bounds progress to previously unrecognizable technological and cultural achievement.
Urban agglomeration in globally dominant metros as centers of talent, innovation, financial resources, and governance.
A single language of tech-speak, excluding those of the surrounding economic and geographic hinterlands.
Pride and arrogance to make a name for themselves, with the resources to reshape and dominate global civilization.

Then and now, the perception would be that the infrastructure and edifices of Babel and AI (especially Large Language Models or LLMs) would prove too big to fail. However, as Babel learned, the reality is of becoming so big that massive failure proves inevitable.

Structural defects may go unnoticed to the point where repair is no longer readily manageable. Collapse looms — whether partial or complete. And the economics of rebuilding are so formidable that the bust following boom world reverts back to simpler technologies accompanied by impoverished communities.

Leadership Gone Awry

As with Babel, a particular challenge with AI is the willingness of the public and civic leadership to ascribe inordinate influence and power to the leading AI purveyors. As would be surmised centuries later by the 1st century AD historian Flavius Josephus, blame for Babel’s demise is focused on one named “Nimrod who excited them to such an affront and contempt of God … as the grandson of Ham, the son of Noah, a bold man, and of great strength of hand.”

Echoes of past experience. Sound familiar?

AI Risks & Mitigation

There has been enough experience with LLMs to understand and attempt mitigation of major risks revolving around issues as for cyber security, data breaches, national security, surveillance, and job displacement. The potential advent of Artificial General Intelligence (AGI) by which AI can exceed human cognition and awareness comes at the new and added risk of compromised human agency. While guardrails may be proposed, they can not guarantee a never ending option for humans to override — or turn off — the AI machine.

Risk management to date has focused on strategies as for added regulation and standards, effective corporate management, human override capacity, pre-deployment testing and monitoring, private data protection, cyber security safeguards, and socio-economic (e.g., job/income loss) mitigation.

But current signs are that risk management measures actively considered or taken to date fall far short of the challenge now and ahead. Per recently publicized GatesNotes, no less an eminence than Bill Gates has warned that “the transition to the AI era will be one of the most turbulent times in human history,” an era for "which we are not preparing adequately.”

The Babel Solution

Whether planned or improvised, the Babel solution was divinely orchestrated, premised on three strategies:

Cessation of building the city including its signature and uniquely recognized edifice.
Confusing the language so that the populace “will not understand one another’s speech.”
Scattering the population including the technological talent geographically “over the face of all the earth.”

In other words, decentralize. No longer go big. Rather, go small. Go diverse and dispersed.

AI Application

How might this unique experience of Babel be applied going forward for the equally unique and challenging world of AI? Here are three ideas to consider:

Instead of ever-expanding and all-encompassing Large Language Models (LLMs), prioritize Small Language Models (SLMs) as a preferred means to better customize solutions, encourage market innovation and competition, and reduce threats of potential large scale infrastructure, security, social and economic dislocation.
Consider use of LLMs sparingly where breadth of scope is truly essential as for society-, nation, and/or world-wide record maintenance but subject to regulatory oversight and privacy protections including rigorous testing and model deployment protocols backed by adaptive governmental regulations with legislative authorization, reflecting general public and industry input.
Preclude AGI applications until and unless there is legislative, governmental and broad industry consensus as to testing, deployment and rigorous monitoring with fail-safe protocols for human oversight and overrides — as applicable for normalized AI operations and/or emergency situations.

From a spiritual perspective, the form(s) of AI applications implemented will prove pivotal. As God directs in Genesis 1, humans were created to “have dominion” as stewards. This directive for human stewardship can be applied not just to dominion over biologic life but also to non-living material attributes of this earth including human contrived technologies — now extended to include AI.

Action Agenda

Summing up, the risk of AI is that civilization as we know if is subjected to the very real possibility of another Babel-like experience. The potential benefits of AI have been cogently and widely articulated. But as with Babel, humanity may be again undone, whether by unbridled human arrogance or carelessness. We need to take the time this time coupled with the humility necessary to get the AI revolution done right.

What is suggested here is by no means a cure-all for the short- and long-term risks posed by AI, especially AGI. But taking action now to focus on SLMs can be a positive step in the right direction. Lessons learned can then be proactively and productively applied to the more pervasive and catastrophic risks posed by LLMs in concert with AGI. Let’s begin now!

This blog post has been prepared by Eric D. Hovee, an independent economic and development consultant for over 40 years. Mr. Hovee is also author of the recently published book “Conflict and Peace: At Home with Jesus.”