AI Celibacy

The Need To Detect Generative AI Ads and Avoid Wasteful Spending and Nuclear War. Part III in the AI Circle Jerk Series.

AI Celibacy
The Need To Detect Generative AI Ads and Avoid Wasteful Spending and Nuclear War. Part III in the AI Circle Jerk Series.

By Samson Williams

In 1983, amidst NATO's Able Archer exercise triggering a heightened state of nuclear alert across the Warsaw Pact, the Soviet Union walking out of arms talks with the US in Geneva and Reagan accepted responsibility for actions in the Iran-Contra affair, claiming they occurred without his knowledge (aka the CIA funded crack cocaine epidemic in Los Angeles that spread nationwide and predicted the War On Drugs and Black communities in America that worked so well China implemented the same strategy with fentanyl four decade later), the movie WarGames dropped. WarGames depicting a chilling scenario: a computer system simulating a nuclear war that could, in theory, trigger real-world annihilation. Fast-forward to today, and we’re in the midst of a digital war of a different sort, as generative AI ads bombard our screens in a ceaseless cycle of “AI circle jerking”—an environment where bots market to other bots in a closed loop that yields little for human audiences or business outcomes. This digital echo chamber drains energy, time, and money, but it also reveals a critical need: the creation of AI systems that are, in a sense, “celibate.” These systems would avoid fruitless engagements and steer clear of participating in wasteful interactions that serve only to feed the bots themselves. More importantly CelibateAI would be less likely to trigger or preemptively launch a nuclear strike in a world where more and more systems are turned over to autonomous systems, with built in deadman algorithms.

The Cost of AI Circle Jerking

Today, generative AI consumes an enormous amount of energy to fuel data centers, power algorithms, and keep pace with the demand for content and engagement across platforms. To understand the scale, consider OpenAI’s GPT-3 model: training this one model consumed approximately 1,287 megawatt-hours of energy—enough to power 120 American households for a year. And that’s just one model among thousands in operation. Much of this energy feeds the loop of generative AI ads and bot-driven engagements, draining resources without tangible returns for businesses. In fact, a significant percentage of digital advertising budgets are likely spent on interactions that look good in engagement metrics and executive dashboards but do little to drive actual human interest or cash revenue.

The phenomenon of bots engaging other bots— a term I recently coined as Dope Coffee’s Head of AI, “AI circle jerking”—wastes not only financial resources but also massive amounts of electricity. If we estimate that 30% of all generative AI resources are engaged in this cycle, that’s hundreds of thousands of megawatt-hours ($18M - $25M dollars) lost annually in digital noise. This waste of electrical power could instead be directed toward meaningful applications, from scientific research to social development programs, highlighting just how much is at stake at the dawn of AI.

Not included in that sum is the lost revenue and brand value businesses, especially startups, lose to AI circle jerking. Conservatively the amount of money business loses to AI Circle Jerking from online and digital ads is well into the billions of dollars. As of 2023, generative AI's influence on digital advertising is emerging but not yet dominant. While specific figures are scarce, Bloomberg projections suggest that by 2032, generative AI could account for approximately $192 billion in digital ad spending, representing a significant portion of the anticipated $1.3 trillion generative AI market. Simply put, startups will continue to be diluted down by VCs, only to turn around and give up equity, that 90% of the time, is lost to AIs circle jerking with each other to the tune of hundreds of billions of dollars annually by the 2030s.

Introducing “AI Celibacy”

AI celibacy, as a concept, proposes a solution: an AI system capable of detecting generative AI ads, ignoring them, and avoiding wasteful engagement. These “celibate” AIs (think “pop-up ad blockers circa 2005) would be programmed to distinguish between content generated for authentic human interaction and content created purely for algorithmic churn. With advancements in natural language processing and machine learning, it’s within reach to build systems that can “see” through the digital fog and discern genuine human engagement from bot-driven noise.

The challenge, of course, is immense. These celibate AIs would need to filter out generative content that isn’t valuable to human audiences, targeting only those engagements that yield real interactions, real conversions, and real ROI. Such a system could be particularly useful for brands seeking authenticity in a sea of digital superficiality.

Defense Industry Concerns: Preventing AI-Driven Escalation

This issue goes beyond marketing budgets. The Defense Industry is acutely aware of the dangers posed by bots engaging in closed-loop communication. Imagine if a defense warning system were to interpret a pattern of bot-generated messages as signals of an impending cyberattack. This possibility raises alarm within military circles, where the cost of mistaken AI-driven escalation could be catastrophic. In WarGames, the fictional WOPR (War Operation Plan Response) computer nearly triggers nuclear war, mistaking a simulated game for real threats. Similarly, in the modern world, military AI systems risk “seeing” bots talking to bots and escalating their responses based on phantom threats.

To prevent this, military AIs would need “celibacy protocols” to detect and ignore generative interactions that lack authentic intent. Instead of reacting to noise, these systems would need to prioritize human-authored content, relying on sophisticated filtering mechanisms that ignore bot-generated chatter. This application of AI celibacy would prevent unnecessary escalations, safeguarding not only cybersecurity but potentially the physical security of entire nations.

Implementing AI Celibacy: Key Strategies

How do we get there, to celibate AI? That road is broad and spacious and may not actually be passable. As it is indeed a cat and mouse game between Human cunning, Archeon-class ancilla, quantum computing and the pursuit of wetware (organic computers). However, below are some options that the smart global nerd community can consider to narrow the gate, potentially averting AI MADness (mutually assured destruction):

  1. Advanced Pattern Recognition and Content Filtering
    AI celibacy begins with pattern recognition. By studying and identifying patterns common to generative AI ads, celibate AIs can “swipe left” on content that follows those structures. Companies could implement filters to catch and avoid auto-engaging with ad-bots, creating a more human-centered approach to digital engagement.
  2. Energy-Efficient AI Deployment
    By prioritizing celibate AIs, we could significantly reduce the carbon footprint of generative AI. Fewer useless engagements mean fewer server processes, which leads to reduced energy consumption and less pressure on data centers. This shift would allow resources to go toward applications that provide meaningful benefits, a step toward responsible AI deployment.
  3. Defense and Security Protocols for AI Content Validation
    The Defense Industry (especially the United States Space Force) could and should implement content validation checks for its AI warning systems, reducing the risk of bots accidentally triggering defense responses. This would involve establishing criteria that bots must meet to be considered legitimate signals, reinforcing the distinction between digital noise and actual threats. In essence, it’s about teaching AI systems to recognize “play” signals (like in WarGames) versus authentic indicators of risk.

Conclusion

As we head deeper into an era where AI is both the creator and consumer of much digital content, AI celibacy offers a way to circumvent wasteful engagements and prioritize meaningful interactions. This isn’t just a marketing fix—it’s a shift in how we approach artificial intelligence itself, with implications that touch everything from sustainability to national security. By creating celibate AI systems, we can break out of the AI circle jerk, fostering a more resource-efficient, human-centered digital landscape that reflects real-world priorities.

About The Author

Samson Williams is a professor at the University of New Hampshire School of Law and Instructor at Columbia University where he lectures on the Space Economy, the 5th Industrial Revolution, Ai ethics and deep tech. When not predicting the future of human civilization, Samson is partner at MilkyWayEconomy and serves as Dope Coffee Company’s provocateur of common sense and Head of AI & Coffee Thinking.