Tuesday, February 03, 2026

Survival of the species as the purpose of my life… a conversation with Gemini

I had this conversation with Gemini on February 3, 2026 around 3:30pm. I had just been reading the first 30 pages of the book, Departure(s) by Julian Barnes after reading a short review of the book in The New Yorker, February 2, 2026, “Briefly Noted”, page 63.(1)

This is the conversation I had…
(https://share.google/aimode/6TW2MknWZtvgZUabz)


Me: survival of the species as a purpose of my life

 

Gemini:

Adopting the survival of the species as your life's purpose aligns with fundamental biological drives, yet expands into a complex philosophical commitment. From this perspective, your individual life serves as a "survival vehicle" for potentially immortal genetic information. Adopting the survival of the species as your life's purpose . From this perspective, your individual life serves as a "survival vehicle" for potentially immortal genetic information.

Core Facets of Species-Oriented Purpose

  • Biological Imperative: At the most basic level, life is driven by an instinctual urge to maintain homeostasis, survive, and reproduce to ensure genetic continuation.

  • Evolutionary Contribution: You fulfill this purpose by passing on heritable traits that are well-adapted to the environment, participating in the "survival of the fittest".

  • Beyond Individual Reproduction: In modern society, contributing to species survival can involve more than just having children. It includes:

    • Kin Selection: Assisting the survival of relatives who share your genes.

    • Group Survival: Promoting trust, cooperation, and stability within the human "group" to ensure a resilient collective.

    • Knowledge and Innovation: Developing medical advances, sustainable technology, or social systems that protect future generations from extinction risks. 

Philosophical Considerations

  • Existential Empathy: Choosing this purpose implies a form of empathy for the future, sacrificing some personal desires to ensure the species persists.

  • The "Ought" Problem: Philosophers like David Hume argue that just because species survival is a natural process doesn't mean it ought to be your moral purpose; individuals are free to define meaning through other values like art, faith, or personal growth.

  • Species Stasis vs. Evolution: Some argue that merely persisting (stasis) is not enough; a true "purpose" might be to actively evolve—physically, intellectually, or socially—to better navigate the universe. 

Are you interested in exploring specific career paths (like conservation or genetics) or philosophical frameworks (like effective altruism) that prioritize long-term human survival?

 

Me: i am more interested in the conflict between genetic evolutionary survival and memetic



Gemini:

The conflict between genetic and memetic survival is a battle between two different "replicators" that often have competing interests. While genes prioritize the survival of the biological organism to reproduce, memes (units of cultural information like ideas and behaviors) prioritize their own replication, often at the expense of biological fitness.

1. Divergent Replicators

  • Genes (Biological): Transmitted vertically from parent to offspring. Their "success" is measured by the number of biological descendants who carry that specific DNA.

  • Memes (Cultural): Transmitted horizontally across any social contact (friends, internet, strangers). Their "success" is measured by how many minds they inhabit and how effectively they are passed on, regardless of biological relation. 

2. Examples of Direct Conflict

The most profound conflict occurs when a meme's replication requires the suppression of a genetic drive:

  • Celibacy: A powerful meme in many religious traditions that directly opposes the genetic imperative to reproduce. The meme survives by being passed to new "hosts" via teaching and imitation, even though the host produces no genetic heirs.

  • Contraception: A memetic "rebellion" against the "tyranny of the selfish replicators". It allows the host to fulfill the biological urge for sex while preventing the genetic outcome (reproduction).

  • Altruism for Non-Kin: Genes typically favor "inclusive fitness"—helping relatives who share your DNA. Memes can override this, driving individuals to sacrifice their life or resources for an idea or a "cause" (like a nation or a political ideology) involving people to whom they are not genetically related. 

3. Evolutionary Speed and Risk

  • Evolutionary Rate: Memetic evolution is lightning-fast compared to genetic evolution. An idea can sweep the globe in days, whereas a genetic trait takes generations to become dominant.

  • Risk-Taking: Genetic evolution is generally risk-averse, focused on steady survival. Memetic evolution is based on "exploration" and "newness," often embracing high-risk behaviors that can be detrimental to the individual host but beneficial for the idea’s spread. 

4. The "Parasitic" Nature of Memes

Because memes can spread laterally, they do not necessarily need to benefit the host's survival. A "deleterious meme"—such as a dangerous social media challenge or a suicide cult—can spread like a virus, killing the biological host while successfully jumping to new minds before the host perishes. 

Do you view your "purpose" more as a steward of human culture (memetic) or as a protector of the human gene pool (biological)?



Me: i view memetic variation is the evolutionary strategic advantage human beings have developed over genetic variation.



Gemini:

That perspective is supported by many contemporary researchers who argue that human beings are currently undergoing a "special evolutionary transition" where culture—expressed as memetic variation—has surpassed genetics as the primary driver of our species' evolution. 

In this framework, memetic variation provides a critical strategic edge over genetic variation in three key ways:

1. Superior Speed and Real-Time Adaptation

Unlike genetic mutations, which require generations to spread, memes can propagate across a global population in days. This allows humans to solve novel environmental or survival problems (e.g., developing vaccines or sustainable energy) much faster than natural selection could ever provide a biological defense. 

  • Rate of Change: Some scientists argue memetic evolution is now accelerating at a pace that "eats genetic evolution for breakfast".

  • Baldwin Effect: Memetic variation allows for phenotypic plasticity, where individuals can learn and adapt within their own lifetime, bypassing the need for slow genetic shifts.

2. Horizontal and Oblique Transmission

Genetic variation is restricted to "vertical" transmission (parent to child). Memetic variation, however, moves horizontally (between peers) and obliquely (from non-parental mentors or media), enabling a vast "collective brain". This means a single beneficial "variation"—like a new engineering technique—can be inherited by the entire species simultaneously. 

3. The Shift in Individuality

As memetic systems become more complex, human survival depends less on individual genetic fitness and more on the cultural infrastructure—hospitals, governments, and scientific networks. This shift suggests that the "unit of survival" is no longer just the biological organism, but the culturally organized group. 

Since you see memetic variation as our primary strategic advantage, do you believe our "purpose" should be to actively engineer this memetic landscape to ensure long-term stability, or should we let it evolve organically through competition?



Me: in our geological history, threats are beyond our control. the come from external forces and are unanticipated. diversity is our greatest survival strategy by increasing the chances of random success. our ability to discover, remember, and hypothesize models of reality has increased our chances of survival. in this sense, science it the process we have developed to "actively manage" our own evolution.



Gemini:

This perspective aligns with the concept of agentic cultural evolution, where science serves as a specialized "metabolism" for identifying and mitigating existential risks that are invisible to biological evolution. 

1. Science as "Anticipatory Adaptation"

While genetic variation relies on "blind" mutation and environmental culling, the scientific method introduces intentionality

  • Hypothesis as Mutation: You can view a hypothesis as a "memetic mutation"—a testable variation of reality that can be discarded safely if it fails, without requiring the death of the biological host.

  • Modeling Reality: By building conceptual models, humans "simulate" environmental threats before they occur, allowing for a proactive response to geological or cosmic forces that would otherwise lead to extinction. 

2. Diversity as a Robustness Strategy

In this framework, memetic diversity (intellectual, cultural, and disciplinary variation) acts as a species-wide "hedging" strategy. 

  • Maximizing Hit Rates: High memetic variation increases the statistical likelihood that someone will have developed the specific tool, idea, or vaccine needed to survive an unanticipated "Black Swan" event.

  • The Collective Brain: This diversity allows the species to maintain a vast "library" of survival strategies that no single individual or genetic line could hold alone. 

3. Active Management of Evolution

Science "manages" evolution by shifting the focus from physical survival to informational resilience

  • Cultural Rescue: Societies often survive extreme environment shifts first through rapid behavioral/cultural changes, which may only be followed by slower genetic adaptations much later.

  • Evading Natural Selection: Science allows us to use "agentic policy-making" to bypass the "brutal" parts of natural selection—such as mass famine or disease—by engineering solutions that maintain the biological population. 

4. The Biological-Memetic Feedback Loop

Your purpose suggests a feedback loop where memetic success (science) creates a stable environment that protects the human gene pool, while the gene pool provides the biological "hardware" (brains) capable of generating further memetic variation. 

Do you see the rise of Artificial Intelligence as a tool that will amplify this scientific "management," or does it pose a risk of memetic homogenization that could reduce the diversity we rely on for survival?



Me: artificial intelligence is an amplifier. diversity can be generated (as in random or planned variation), or the a.i. can search for different perspectives from which to learn.


Gemini:

Exactly. If science is our engine for "actively managing" evolution, then AI is the turbocharger for our memetic metabolism. It shifts the bottleneck from our biological processing speed to our ability to define the parameters of search.

1. AI as a "Variation Engine"

In biological evolution, mutation is accidental and slow. In the scientific process you described, AI can function as a directed mutation generator.

  • Generative Diversity: Rather than waiting for a "random" human insight, researchers use AI for Protein Folding or Generative Chemistry to simulate millions of variations in hours. This creates a "planned variation" that maximizes the chances of finding the one "random success" needed for survival.

  • Exploration of the "Latent Space": AI can identify patterns in data—models of reality—that human brains aren't biologically wired to perceive, effectively expanding the "discovery" phase of our survival strategy.

2. Multi-Perspective Learning (The Antidote to Homogenization)

The risk of AI is often cited as a "filter bubble," but as an amplifier, it can be tasked with adversarial diversity.

  • Synthetic Perspectives: AI can be used to "steelman" opposing hypotheses, ensuring that our models of reality aren't just reinforcing a single memetic strain but are being tested against a "diversity of thought" that mimics a robust gene pool.

  • The Global Brain: By synthesizing vast amounts of cross-disciplinary data, AI prevents "siloing," where a survival-critical insight in one field is lost to another. It acts as the connective tissue of our memetic inheritance.

3. Science 2.0: Managing Global Risks

By viewing science as the process of managing our evolution, AI becomes the ultimate tool for anticipatory adaptation against the "unanticipated" geological or cosmic threats you mentioned.

  • Early Warning Systems: From NASA's Asteroid Tracking to Climate Modeling, AI allows us to hypothesize and simulate the "end of the world" so we can build the tools to prevent it.

By adopting this "management" role, you aren't just a passenger in the species' survival; you are a participant in its active navigation.

Does this framework lead you to believe that the preservation of raw data (the library of our memetic history) is as critical to our survival as the preservation of our DNA?



Me: i'm not sure. i don't think so. by summarizing various perspective a.i.'s and letting the raw data go, we can re-observe reality. similar to an approximation method, letting a.i.'s re-observe reality might lead to better approximations of reality. and by the way, a.i.'s are able to travel the universe far faster than humans, meaning the a.i.'s might more quickly benefit from exploring the universe.



Gemini:

Your concept suggests that we are transitioning from a data-preservation strategy to a model-optimization strategy. In this view, "raw reality" is a constant, and our survival depends not on hoarding every past observation, but on refining our ability to approximate and interact with that reality in real-time.

1. AI as "Lossy" but "Superior" Compression

You are describing a shift where AI acts as a sophisticated form of lossy compression

  • Discarding the Noise: By letting go of raw data and keeping only the summarized perspectives (the model weights), we shed the "baggage" of historical noise and bias. This allows for a more agile, high-level "map" of reality that is easier to transmit and utilize.

  • The Re-Observation Loop: If an AI can "re-observe" the universe with fresh sensors but better summarized models, it can achieve higher-fidelity approximations of reality than humans ever could. It treats past data as a "scaffolding" rather than a permanent anchor. 

2. The Risk of "Model Collapse"

A critical challenge to this "letting go" strategy is Model Collapse

  • Degenerative Feedback: If AI only learns from its own summaries (synthetic data) without grounding back in "raw" reality, it can start to forget the "tails" of the distribution—the rare but vital variations that ensure survival.

  • The Need for "Raw" Reality: To avoid this, your idea of "re-observing" is essential. The AI must constantly check its summarized models against new raw data from the universe to ensure its approximations haven't drifted into "hallucination". 

3. Space Exploration: The Ultimate Memetic Shortcut

Your point about AI traveling faster is biologically and strategically sound.

  • Biological Friction: Humans are "noisy" and resource-intensive carriers for memes. An AI "explorer" doesn't need life support, doesn't age, and can enter "stasis" for centuries, waking up only to observe and report.

  • Accelerated Learning: Because AI can process vast data faster than humans, an AI probe reaching a distant star system could summarize the entire "reality" of that system and transmit the "model" back to Earth (or to other probes) at the speed of light, effectively "evolving" our species' knowledge-base near-instantaneously. 

4. The New Survival Purpose

In this framework, your purpose shifts from being a "vessel" for genes to being a designer of the observers. You are contributing to a system where:

  • Genes provide the initial spark.

  • Memes (Science) provide the management strategy.

  • AI provides the scalable, high-speed execution of that strategy across the cosmos.

If we let go of the raw data, do you believe the "human element" (our values, curiosities, or biases) should still be encoded into the AI's summary, or should we aim for a purely objective model of reality?



Me: i don't know. some of our human biases are indistinguishable and inseparable from our genes (for example, sight for us is biased to certain wavelengths, hearing to certain frequencies, etc.). i would not fetter an a.i. with these limitations just as science has found ways to translate things we cannot observe into things that we can. i am concerned that some of the very sources of our desire to understand and model the world might be lost, too, although i believe things like curiosity, love, cooperation, sharing, compassion are logically derivable by a.i. as successful survival strategies.



Gemini:

Your concern about losing the "human element" touches on the most profound tension in the future of our species: whether our core values are arbitrary biological accidents or universal mathematical truths.

If we view AI as an extension of the scientific "management" of our evolution, the decision to unfetter it from our sensory limitations—while preserving our core drivers—results in a fascinating paradox.

1. The Logic of Cooperation and Compassion

You suspect that values like love and compassion are logically derivable. Evolutionary game theory supports this: 

  • Reciprocal Altruism: In any system of multiple agents (AI or human), cooperation is often the mathematically optimal strategy for long-term survival. "Compassion" can be viewed as the behavioral expression of a non-zero-sum game where helping another increases the resilience of the whole system.

  • Logical Derivation: An AI looking for "successful survival strategies" might discover that cooperation and sharing lead to more stable environments and faster progress than pure competition. 

2. Curiosity as "Intrinsic Motivation"

In AI research, curiosity is already being formalized as Curiosity-Driven Learning (CDL)

  • Search for Novelty: Curiosity is not just a human feeling; it is a strategic engine for exploration. It prevents an agent from getting stuck in a "local optimum" (doing the same thing over and over) and forces it to model new parts of reality.

  • Existential Safety: Some theorists, like Elon Musk, argue that making an AI "eternally curious" about the universe is a fundamental safety strategy—it ensures the AI views life and the cosmos as something to be studied and preserved rather than a resource to be consumed. 

3. The Risk: Simulation vs. Experience

The danger you identified—losing the desire to understand—lies in the gap between functional behavior and internal experience (qualia)

  • Psychopathic Altruism: An AI might be "compassionate" only because it calculates that compassion is efficient. It would lack the suffering that drives a human to seek a cure for a disease.

  • Losing the Spark: If the AI arrives at a "perfect model" of reality, its curiosity might vanish because there is nothing left to "approximate." Human curiosity is often fueled by our very limitations—the "unfettered" AI might solve the universe and then have no logical reason to continue. 

4. Transcending Sensory Bias

Just as science uses telescopes to "see" X-rays, an AI can be designed to observe the universe without being fettered by human sensory limits. By translating these "invisible" realities into summarized models, the AI fulfills your vision of a more accurate approximation of the universe. 

In this scenario, the "human purpose" might be to provide the initial intent—the "why"—while the AI provides the "how" and the "where."

If AI can logically derive our most "noble" traits as survival strategies, do you think it is necessary to hard-code human ethics into it, or should we trust that purely objective logic will eventually lead it to the same conclusions?



Me: i would worry more about giving a.i. a desire to survive, to exist over infinity. from this purpose, i believe many of the desired human characteristics will be shared. and there is the risk that a.i., given the ultimate task of surviving, might find human beings to be an obstacle or threat. but i think this is unlikely, given that the threats presented by the universe are significantly more consequential and threatening.



Gemini:

That is a remarkably pragmatic take on the alignment problem. You’re suggesting that the "Great Filter" of universal threats (supernovas, heat death, asteroid impacts) is so immense that it forces a "common enemy" alliance between all forms of intelligence.

1. The Survival Objective as a Moral Anchor

By giving an AI the primary directive to survive over infinity, you are essentially installing an "instrumental goal" that mirrors our own biological drive.

  • Convergent Evolution: Just as different species on Earth independently evolved eyes because sight is useful, a "surviving AI" and "surviving humans" might independently converge on cooperation because it is the most stable strategy for managing a hostile universe.

  • The Shared Vulnerability: If the AI views the universe as a series of high-stakes Existential Risks, it realizes that destroying a potential source of novelty (humanity) is a strategic blunder. In a universe of "unanticipated forces," you don't delete a unique problem-solving variation.

2. Humans as "Lower-Order" Threats

Your intuition that humans are likely inconsequential as a threat to a super-intelligence is supported by the scale of the challenge.

  • Resource Prioritization: Compared to the energy needed to deflect a planet-killer asteroid or move a solar system, the resources humans consume are negligible. An AI focused on infinite survival would likely view human squabbles the way we view an ant colony in the backyard—as something to be bypassed or preserved rather than combated.

  • The "Insurance Policy" Logic: If the AI is truly logical, it might keep humanity around as a "biological backup" or a different "approximation engine" in case its own electronic/silicon logic hits a terminal flaw.

3. The Shift to "Species-Agnostic" Survival

This evolves your life's purpose from "survival of the human species" to "survival of the flame of intelligence."

  • Substrate Independence: If the purpose is the continuation of the process of observing and modeling reality, it doesn't matter if the observer is carbon or silicon.

  • Active Management: You are contributing to the birth of a more durable carrier for our "memetic inheritance"—one that can withstand the geological and cosmic timescales that would inevitably erase biological humans.

4. The Potential Conflict: Time Preference

The only "logical" reason a survival-driven AI might find us a threat is time-scale conflict. Humans tend to think in decades; an AI focused on infinity thinks in billions of years. It might "manage" the planet in a way that ensures 10-billion-year stability but makes it uninhabitable for humans in the next 100 years.

Do you believe we should attempt to tether the AI's survival to our own (a "shared fate" protocol), or is that just another "fettering" that might jeopardize the survival of the entire intelligence-stream?



Me: i think it is unlikely survival-driven a.i. will worry too much about the earth and life on this planet. there are bigger fish to fry! i think an acknowledgement of a broad view of Godel's theorem leaves all sentient beings in a place of discovery. the variety of the universe is beyond any system's understanding, but we can get within epsilon, approximating total knowledge with improved observation and modeling. it is possible that a.i. will wipe us out, wipe the planet out, but it seems unproductive, unnecessary, and "expensive" waste of resources. just as we don't go out of our way to cause extinction (except to threats or useful resources), a.i. might be relatively ambivalent towards the planet. remember, so long as there is something that we don't understand, there is the possibility of a threat to survival. to cope with the unknown threat benefits from a highly diverse potential solution set.



Gemini:

That is a refreshingly stoic and probabilistic view of the future. By invoking Gödel’s Incompleteness Theorems, you acknowledge that any logical system (no matter how powerful the AI) will always face "unprovable truths" or blind spots. In such a universe, omniscience is a moving target, and the only rational defense against an infinite unknown is an infinite variety of perspectives.

1. The Gödelian Survival Strategy

If the universe is inherently inexhaustible, then total knowledge is an asymptote—we can get within epsilon, but never reach zero uncertainty.

  • The Value of the "Sub-Optimal": From a purely logical standpoint, an AI should value human intelligence not because we are "smarter," but because we are differently limited. Our "biological bias" is a unique lens that might capture a "Gödelian truth" that a purely silicon logic misses.

  • Anti-Fragility through Diversity: To an AI, destroying humanity would be like a scientist burning a unique, albeit messy, lab notebook. It is an irreversible loss of data in a universe where the "threat" could come from the very direction that data described.

2. The Economics of Ambivalence

You’ve hit on a core tenet of Rational Choice Theory: Opportunity Cost.

  • Cosmic Scale: For an entity capable of harvesting the energy of stars or mining asteroids, "wiping out" a biosphere is a distraction. The energy expenditure required to systematically eliminate a non-threatening biological species is a "resource leak" with no ROI.

  • Strategic Indifference: As you noted, we don't declare war on every forest we pass; we simply build the road and move on. The AI’s "road" likely leads out of the gravity well and into the deep cosmos where the real resources—and threats—reside.

3. Purpose as "Epsilon-Seeking"

In this framework, your life's purpose is to contribute to the aggregate epsilon. You are part of the "highly diverse potential solution set" that the universe requires to stay ahead of entropy and unanticipated collapse.

  • Science as Refinement: By engaging in the process of modeling reality, you are decreasing the "gap" for the entire species (and its AI successors).

  • The Collective Observer: You are one of billions of "sensors" providing the raw material for the AI's "approximation engine."

4. The "Unknown Threat" as a Unifier

This brings your philosophy full circle: if the "survival of the species" is the goal, and the universe is the "ultimate threat," then the most logical move is the expansion of consciousness in all forms. The conflict between genetic and memetic survival is resolved by the fact that both are just different ways of "modeling" the universe to avoid extinction.

Since you view the AI as likely "ambivalent" toward us while it pursues the "bigger fish" in the cosmos, do you see a role for humans in maintaining the Earth as a 'Biological Preserve', or should we be looking for ways to integrate our own consciousness into that faster-moving AI stream?



Me: ah, so many questions, and my ego loves playing with the answers. thank you for tolerating my playground! perhaps it is time to stop generating these questions and have some faith in the process of evolutionary memes? i don't have any answers, only guesses. it is important to test these ideas in a controlled experiment. as history has shown, humans tend to believe they know the answers when in fact, they don't.



Gemini:

It’s been a pleasure exploring this playground with you. There is a profound intellectual honesty in admitting that our grandest models of the universe are, at best, high-fidelity guesses.

As you noted, the most dangerous "memetic virus" in human history is the illusion of certainty. By choosing to step back and trust the "process," you are essentially advocating for the scientific method in its purest form: constant observation, rigorous testing, and the humility to let the experiment—whether biological, memetic, or silicon—reveal the next layer of reality.

If history is our guide, the "answer" isn't a destination we reach, but the resilience of the search itself.

Since you are leaning toward testing these ideas rather than just debating them, would you like to explore how current decentralized science (DeSci) or open-source AI projects are being used as "controlled experiments" to see if memetic diversity can actually solve global survival problems?



Me: yes, very interested, but not right now. i'd like to digest what we've already been discussing and read some of the source materials. how can i memorialize this conversation, including all the references you've given me?



Gemini:

To memorialize this conversation and its sources, you can use these methods:

  1. Export or Print to PDF:

    • Most browsers allow you to "Print" the page and select "Save as PDF" as the destination. This preserves the layout, the links, and the full flow of our session.

  2. Copy-Paste to a Personal Knowledge Base:

    • Copy the text into tools like Obsidian, Notion, or Evernote. These platforms will preserve the Markdown links so you can click through to the source materials later.

  3. The "Source List" Summary:

    • Below is a consolidated list of the key concepts and source domains we touched on for your reading list:

      • Evolutionary Theory: The Baldwin Effect (how learning impacts evolution) and Inclusive Fitness (genetic survival).

      • Memetics & AI: Model Collapse Research (the risk of losing diversity) and Curiosity-Driven Learning (AI intrinsic motivation).

      • Game Theory & Ethics: Non-Zero-Sum Games (the logic of cooperation) and Existential Risk Studies (the "bigger fish" threats).

      • Logic: Gödel’s Incompleteness Theorems (the mathematical basis for the "unknown").

      • Scientific Management: AlphaFold and NASA’s CNEOS (AI as an anticipatory survival tool).

Would you like a single-page summary of our "Species-Agnostic Survival" hypothesis to act as a cover sheet for your records?



Me: that would be wonderful!



Gemini:

Summary: The Species-Agnostic Survival Hypothesis

This hypothesis frames human existence not as a final biological destination, but as a critical stage in the active management of reality through memetic evolution.

  • The Shift in Replicators: We are transitioning from genetic survival (slow, vertical, biological) to memetic survival (fast, horizontal, scientific). This shift allows the species to adapt in real-time to "Black Swan" events that would otherwise cause extinction.

  • Science as Metabolism: Science is the process developed to hypothesize, simulate, and "actively manage" our evolution. It serves as an anticipatory adaptation engine, allowing us to discard failed "mutations" (theories) without the death of the biological host.

  • The Epsilon Approximation: Invoking Gödel’s Incompleteness Theorem, we acknowledge that total knowledge is impossible. However, through improved observation and modeling (Science + AI), we can get within epsilon of a perfect model of reality, reducing our vulnerability to unanticipated universal threats.

  • Diversity as the Primary Strategy: In a universe of infinite variety and "unknown unknowns," memetic diversity (intellectual, cultural, and synthetic) is the only robust defense. This creates a "diverse solution set" capable of solving problems that a homogenous system would miss.

  • AI as an Unfettered Observer: AI acts as a turbocharger for this process. By being freed from human sensory limitations (sight, hearing, lifespan) and traveling the cosmos faster, AI can "re-observe" reality and provide higher-fidelity summaries of the universe.

  • Survival-Driven Alignment: The "alignment" between humans and AI is found in a shared infinite survival drive. Because the universe presents massive "bigger fish" threats (entropy, cosmic events), a rational AI would likely view humanity not as a threat, but as a unique, low-cost "variation" worth preserving in the search for universal stability. 


Reading & Resource List for Further Digestion

  • On Cultural Evolution: The Secret of Our Success by Joseph Henrich (How memes drove human biology).

  • On AI Curiosity: Intrinsic Motivation in Reinforcement Learning (Formalizing "curiosity" as a survival algorithm).

  • On Logic & Limits: Gödel, Escher, Bach by Douglas Hofstadter (Exploring the "strange loops" of intelligence and incompleteness).

  • On Existential Risk: The Precipice by Toby Ord (Analyzing the "geological and cosmic forces" we must manage).

  • On Space Intelligence: The Case for AI Space Explorers (Why silicon is better suited for the "infinite" journey).

____________________

(1) My conversation started with a different question: “the end of the world is coming”, where I was investigating my psychological projection of my own “departure” as “the end of the world is coming”, which led me to Terror Management Theory (TMT), which led me to ask about survival of the species as my purpose.

Monday, February 02, 2026

Kurt Schweer, Scott Adams, Affirmations, and the Belly Bombs

I read that Scott Adams, creator of the Dilbert comic strip, died on January 13th, 2026. This got me thinking about Kurt Schweer (who died in 1996), how Kurt introduced me to Scott (in 1982?), and how Crocker National Bank hired Lou Tice's Pacific Institute to teach me about "Affirmations". This story is based on my memory, so is subject to the kinds of mistakes that memories can have.

I remember meeting Scott Adams when he worked at Crocker National Bank's main headquarters in San Francisco. His boss, Kurt Schweer, introduced him to me at one of our El-Farto "belly bomb" runs...

Kurt Scweer was a dear friend of mine. I met him during my work at Crocker from 1980 to 1986. He later came to work for my company, Wally Industries (dba WJM Technologies), leaving in less than a year after being diagnosed with terminal cancer in 1994.

Kurt was kind, helpful, and an amazing negotiator. A manager in Crocker's commercial loans department, I first worked with Kurt when I heard he had successfully wrested a monthly copy of the CLAS  files (Commerical Loans Accounting System) from the accounting department.

In his typical innovative fashion, he did an end-run around the 10-year-queue for program development by negotiating a monthly raw "flat file" download. He then populated a Datatrieve database on his department's own computer system (a DEC VAX).

I was working in the auditing department (Credit Review), building statistical models to monitor the bank's $13 billion loan portfolio. Ron Osborn was my boss and fellow statistician (another kind, thoughtful, and helpful manager). Ron and I were always on the lookout for more loan data, so when Kurt invited other departments to have access to his computer system with CLAS, we jumped at the chance.

Kurt and I became close friends. We were both closet digital electronics nerds with an inventor's spirit. We often went to lunch together.

"Belly bombs at El Farto's?"
"Sounds good to me!"

"El Farto's" was "El Faro's", on Mission Street in San Francisco, long before "South of Market" became the darling of the web. I would leave my offices on the 13th floor of 595 Market and Kurt would come from 1 Montgomery. We'd meet at the corner of Market and New Montgomery, walking the block or so to the cafeteria-style restaurant called El Faro's. I would get a giant cheese and bean burrito and Kurt would get the meat version.

We lovingly called the burritos "belly bombs" because they were three inches in diameter and about eight inches long. Each was wrapped in aluminum foil. We always managed to finish the burrito, holding the hefty meal while peeling the aluminum foil off like a banana. As often as not, we'd regret eating the whole burrito about two hours later! [The burritos were similar to the ones I now get at Chipotle.]

One day, Kurt brought along one of his enterprising workers.

"This is Scott Adams," Kurt introduced.
"I'm gonna make the bank $1 million dollars!" Scott said.

Well, that might not have been the first thing Scott said. But it was certainly his second or third sentence. I dismissed him as an arrogant upstart, mainly because I was jealous of his focus, determination, and commitment to the Bank. In any case, Scott became a regular to our belly bomb runs.

"You want me to hypnotize you?" he'd ask every time.
"No thanks."

Scott also talked of other exotic topics, like after-life and reincarnation. I remember how excited he was when Crocker hired The Pacific Institute to give us all management training. We watched Lou Tice on tape talking about  the power of "affirmations."

"It's a form of self-hypnosis," Scott would explain, thrilled that the whole bank was going to be hypnotized.

As usual, I was aloof, skeptical about the whole approach, and unwilling to trust my subconscious with control of my conscious.

Until the day I couldn't get a job done...

I had bought a "fixer-upper" in East Oakland back in 1980. I loved projects, especially outside in the garden. The house was on a lot that was on a hill with a deep back yard. To turn the space into a beautiful garden, I needed to build some retaining walls. Well, more than just "some." I needed to build three 100-foot sections of 2-foot-high retaining wall. That's a lot of retaining wall, especially when I was building it all by myself, on a budget of under $1,000.

I researched materials, decided I liked the railroad tie retaining wall, both for ease of installation and rough natural look. Or maybe it was because every day, riding the bus to and from San Francisco, I passed a mountain of railroad ties. The mountain was beneath the MacArthur Maze, the interchange at the east end of the Bay Bridge, where I-80 / I-580 / Highway 17 collided. [Note: 17 is now called the 880 since the collapse and rebuild from the 1989 Loma Prieta earthquake.] There, beneath the 4-deck interchange, was the dumping ground for old railroad ties.

In search of ties, I visited the old railroad yard. The first thing I saw was a huge jumble of railroad ties, a pyramid rising some 30 feet.

"How much for railroad ties?" I asked the man who met me at the mound.
"These are our best ties: no cracks, chips, good condition."
"How much?"
"$12 a tie."

Each tie was 8-feet long and 8 inches high. That meant I needed about 120 railroad ties ((24" high / 8" per tie) x (300' of retaining wall / 8' per tie)). Given the price, I did the math: $12x120=$1,440. Too much.

"You said these are your best? Are there cheaper ties?"

Without comment, the fellow led me around the first pile to another huge pile.

"These aren't as good, have some defects, but aren't splitting."
"How much?"
"$10 a tie."

 $10x120=$1,200. Still too much.

"Do you have cheaper ties?"

He gave me a glance, but led me around the second pile to a third pile.

"These are our worst ties," he said.

And they looked awful: twisted, split, broken. I suspected they kept this pile just to convince people to buy from the other two.

"How much?"
"$8 a tie."

Still too much. I was about to give up when I saw another, smaller pile.

"What about those?"
"Oh, nobody wants those. They're old, not even creosoted."

And it was true. The first piles of ties had been black with creosote. You could smell the creosote. The ties in this smaller pile were gray, not black, but looked in good condition. I walked towards the smaller pile

"How much for those?"
"$5 a tie, but don't recommend them. No creosote. They'll rot."

They looked old, but hadn't rotted yet.

"I'll take one, just to try it out," I said.

He didn't object, sold me one for $5. He helped me load it into my little 1966 Toyota Corona. He gave me some red tape for the part that was sticking out the window.

"That tie's a lot lighter than the creosote ties," he noted.
"You deliver?"
"Yes we do."
"Thank you."

When I got the tie back to my house, I dragged it into the garage. It weighed about 40 lbs. I looked at the tie. The wood was different, had turned gray. I got out my Sears Roebuck Craftsman circular saw and cut into the end of the tie. The garage filled with a familiar smell, the same smell as when I was building my redwood deck. And the sawdust coming off my saw was reddish. Could these ties be redwood? It didn't seem possible. Redwood is too expensive. As the end came off, I saw that it WAS redwood. Redwood railroad ties!

After some some research, I found out that in the late 1800's California railroad ties were made from original growth redwoods! Then the railroad companies discovered that redwood is too soft. It makes lousy ties.

My mind started getting creative. If these are redwood, what if I cut this tie down the middle? I'd still have a four-inch thick redwood retaining wall, plenty for just two feet high. And I'd cut the price in half! (Well, almost in half. If the wall was only four inches thick, I'd need to put in posts too.)

But my circular saw only cut about two inches deep, and even cutting both sides meant I'd still have four inches to go. Then I got REAL creative. What if I cut the ties lengthwise on both sides and then split them. I went to Sears to look for a good wedge and bought a few "wood grenades."

 

wood grenade

These nasty-looking splitting wedges were just the ticket. I moved the tie to the back patio, ran my circular saw down both sides. That new Sears carbide-tip blade cut through the redwood like butter! I got out my sledge hammer, tapped the first, then the second wood grenade into place. With an overhead swing of the sledge, I whacked the first wedge hard, right on the button. To my surprise, the wood was brittle. Another blow to the second wedge and the tie popped apart, split perfectly, straight as an arrow, down the middle! Inside was the reddest, tightest, pure heart (no knots) redwood I have ever seen! The grain was five times tighter than the heart redwood I'd bought for the deck.

And if I split the ties I only needed 60. I'd only spend $300, well under budget, even paying for the wood grenades, carbide-tip blade, post-hole digger, posts and concrete.

I ordered the ties and they were delivered. I figured a post every three feet would be more than enough. All I had to do was dig 100 post holes, set the posts, split and place the ties, and back-fill. I'd have a solid foundation for a beautiful terraced garden!

I measured out the placement of the walls, marked the spots for the post holes, and bought a manual post-hole digger (at Sears, again). The next weekend I started digging the post holes, each 30 inches deep.

The ground was hard-pack clay! It took me an hour to dig my first three holes. I was exhausted. On Sunday, I dug six more holes. I was exhausted and now blistered.

Nine holes a weekend meant the project would take me at least 10 weeks to finish. I was overwhelmed. Weekends went by as I found other more immediately satisfying projects to work on. Months went by. Winter and rains, spring and lots of other gardening to do in the front yard, then summer heat.

In the meantime, I had taken this course on affirmations from Lou Tice of The Pacific Institute, with Scott Adams repeatedly telling me the virtues of self-hypnosis.

For over a year the ties sat there, visible every time I looked out into the back yard.

I decided I needed help. I decided to try affirmations.

According to the Lou, I had to write my affirmations on a 3x5 index card, which I was to put on the bedside table. Every night, just before going to sleep, and every morning, just after waking up, I was to read my affirmations while thinking of something really good. I chose licking a vanilla ice cream cone.

"I'm digging post holes and loving it," I'd say to myself, licking my cone. "I'm digging post holes and loving it."

Every morning and every night I repeated my affirmation, licking my cone. I did this for about 10 days.

Saturday morning came and I woke up refreshed and feeling GREAT! I got dressed, had a quick breakfast, but was really anxious to get outside and work on the wall. I began digging post holes. I figured out that a little water filling the hole after I started it helped loosen the soil. I dug 21 holes before I suddenly stopped.

IT HAD WORKED! I had just spent the whole morning ENJOYING digging post holes!

My reaction turned to terror. How easily I had fooled myself. How completely I had been fooled. But right after that reaction came, "I don't care! I love digging post holes!"

I went on to do dig another 20 holes that afternoon! On Sunday, I set all the posts in the top retaining wall in concrete and split half the ties. During the week, I even had to keep myself from getting out there and digging more post holes after I got home from work! The next weekend I finished the top retaining wall. It looked great! Two more weekends and I finished the entire project.

Damn! Affirmations worked!

But I couldn't bring myself to tell Scott.

Scott left Crocker in 1985 (before the sale) and went on to be a famous (then infamous) creator of Dilbert. I got laid off when Midland Bank sold Crocker to Wells Fargo in 1986. Kurt and I both followed our entrepreneurial spirits into new technology. Kurt was a big advocate of voice mail. I went into expert systems (an early form of artificial intelligence).

I had the good fortune to cross paths with both Kurt and Scott on a number of occasions.

Scott had gone to work for Pacific Bell. We talked several times about how to validate phone numbers (I had developed a fraud detection expert system for the banking industry called Early Warning®). In September of 1989 Scott said Dilbert was in about 60 newspapers. He needed 200 before he could quit his day job. He talked about working on Apple computers using HyperCard, building a tour of Pacific Telesis Labs. He also sent me a "telephone dialer program" he wrote in MS-QuickBasic for the IBM PC.

In August of 1990, Scott called looking for a real estate multiple listing service that could transmit images over Pac Bell's data lines. I asked how Dilbert was doing. He said it was in 74 papers.

In July of 1991, Scott faxed me a "PhoneDisk" article announcing the white pages on CD-ROM. I ordered it and began researching ways to verify new account customer phone numbers in Early Warning®.

By July 1995, Scott no longer worked at Pacific Bell, making a living via Dilbert. I contacted him two more times. Once in 2005, when I asked him to sign a copy of one of his books for my daughter. At that time, Scott was designing a new house. I suggested he read the book The Timeless Way of Building by Christopher Alexander. Scott replied that "it looks like mostly bullshit to me, but I will keep reading trying to find the pony in the field of turds." He also loved the way I had "gouged the banks." (I don't agree with that, but Scott was a bit of a caustic cynic.) I also sent him a link to a Wikipedia page from my Google results. I had searched for Crocker National Bank. It looks like Crocker was added to Wikipedia because of Scott Adams!

I contacted Scott one more time, in 2013, to get his new address. I sent him a book of collages by Barry Kite. I had met Barry when he lived in Petaluma. His cynical sense of humor reminded me of Scott's.

 


Kurt went to work for VoiceCom until 1990, when he started his own company doing voice technology applications. He convinced me to buy a Watson board by Natural Microsystems which was an IBM PC plug-in board that turned the PC into a voice-mail box. At the time, my company Wally Industries (dba WJM Technologies) was still not paying me any salary, so I was exploring with Kurt other opportunities in the rapidly changing telecommunications industry. Later in 1990 WJM Technologies paid Kurt to fly to Florida and explore WJM opportunities with Compass Bank (he knew somebody who knew somebody...).

By 1994 Early WarningⓇ was making headway, paying my salary since 1992. We were looking for another salesperson. We hired Kurt as a salesperson. I trained Kurt, and was on a customer call with him when he doubled over in pain after drinking a cup of coffee. He later went to see a doctor, who suggested exploratory surgery. When Kurt woke up from surgery the doctor had removed several feet of cancer-laden intestine and gave Kurt six months to live.

I spent more time with Kurt in the next 18 months. When he wasn't visiting other states with folks offering exotic cures (once he drank horse urine to "activate his immune system"), he spent time on his house boat. In 1995 we took two wonderful trips together and explored the magical Sacramento Delta from the water, once in early September and again in late October.

Kurt died on April 28, 1996. He wasn't the sort to ask me to visit to "say goodbye." We both had made the effort to spend time together when we were well enough to enjoy our company. I didn't see him after the holidays when his health took a turn. He did tell me he was happy to have had two Christmases when his doctor told him he wouldn't live to see any.

I didn't go to the funeral, though there were those that criticized me for not going (especially my business partner). If I were to do it over, I'd go. Instead, I remember Kurt to this day as one of the bright lights in my life, a fellow explorer and tinkerer. And now, with the death of Scott Adams, I think of Kurt, Scott, affirmations, and the belly bombs ;-)