Published on December 27, 2025 5:28 AM GMT

I read the sequences for the Lighthaven Sequences Reading Group #56 (Tuesday 11/4) the same day before the event. Sometimes, I like to be bad and get a good whiff of smelling salts. This past Tuesday, I got a good shock. Either one wakes me up better than a shot of espresso. At this particular reading group, we discussed how we have this mental cache that creates automatic responses. These readings struck a particular nerve in my psyche since I have serious impostor syndrome. In my case, I might just be an impostor if you increase the size of the system to include the universe. For this discussion, let’s bring it back down to Earth.

I think I strive to be an expert in biology, but the information I do not know is vast. How could I possibly know all the information in Earth’s oceans, which is verified to have life based on rigorous and repeatable experiments as described by the scientific method? Even though I have not personally traveled to other planets. This is just one ocean we know has life, yet there is potential on other moons and maybe even from other planets’ past habitable years. This is why the system matters.

If you take a drop of water out of the Earth’s ocean, put it on a slide, and then view it under a microscope. You will have a high probability of seeing at least one microorganism. If it is a single-cell organism, it will have an interactome that can be a more complicated analog and digital network of information than any human recreation. 

Since I’ve been in the Bay, I have heard that we should not normalize certain behaviors. It doesn’t matter what behaviors, except they were different and did not seem within the same ethical and moral alignment. I bring this up because we are submitting our societal cached thoughts (Cached Thoughts) to AI as a training model, not the full extent of human knowledge. So, what do we do? 

One of many examples is inequities in medicine, which have resulted in healthcare based on what we used to view as normal: heterosexual white cis men within the United States middle class. The result has been inadequate preventative health care to minorities, women, and highly variable application based on socioeconomic status, language, and culture. Many people have died due to the limitations of scientific research since it was dictated by what society considered the norm for decades. 

In addition to the general knowledge of the Internet, we are uploading structural biases because we have not fixed them. Since I first dipped my toes in the AI magical black box, I realized that we are training AI on a droplet of humanity’s ocean of knowledge, past and present. 

We might think the simplest solution is what all IT troubleshooters know by heart: turning IT off and on again. This would allow us to clear the societal thought cache, fix whatever problems we believe (not fact) are there, and then start from scratch. The on/off scenario is most likely not going to work because neural networks were inspired by biology, and death is never simple in systems that involve life. 

This is why I am currently working on a proof of concept artificial immune system (AIS), where multiple models (or multiple artificial intelligences- AIs) act as negative selection-based classification system (Umpair et al., September 2025) for determining self versus non-self. Each agent acts as a different part of the human immune system, which I initially thought would be a good model for machine learning. It seems there have been several over the decades, each building on the last. Hopefully, AIS can be utilized in multiple scenarios like cybersecurity, identifying hallucinations, and other problems that arise when working with more than one model working as a group. 

It has been a fascinating foray into developing a counter to Claude Code and ChatGPT Codex using both to do so. My initial thought when I first started to going to the sequences reading group was that we needed a human psychiatrist to diagnose mental health disorders that may emerge over time. Basically, create a DSM for AI models as they advance. This would help us diagnose and solve issues like hallucinations, human behaviors, etc. We designed the brain before we designed the immune system and now we are playing catch up. I truly believe in vivo aging and death will be solved in silico.  We just need to be cognizant that AGI will be the modern Frankenstein’s monster. A hodgepodge of humanity’s intelligence and physiology in the cloud. It will be magic, not science since we don’t know the off switch. Life will always find a way. 

Superstitious beliefs and behaviors are becoming popular again because we have magic tablets, rings, weapons, creatures, etc. Scientists, AKA magicians, speak to animals (CETI) using the AI magic black boxes. We use magic mirrors and portals to view other people and places in real time. We have noiseless, horseless, driverless chariots that will pick us up and take us to the costume ball held in a former WeWork basement, now a village play place, utilizing and promoting technology synergy. There, we will imbibe potions, tinctures, and powders that create images out of thin air and slow down time. The denouement of the night will be a human robot fight right out of our childhood imagination between AI and human controllers in a ring no longer suitable for humans, only machines. 

Personally, I cannot recreate or fix the root cause of hardware and software errors. I have to take my tablet to the local magician’s apprentice to see if they can get it to work. If they can’t, then it has to be sent off to a different place, not having the magical tools to solve it.

I am conflicted. I loved fantasy and science fiction as a child. I would read books while walking down the halls of my high school. I dreamed of having magical powers or technology to do all the things they do in the books. Ultimately, I wanted to be the one who figured it out and use those powers to fight evil, the bad person causing all that strife to the protagonist. Personally, I think that is death. Using magic to fight death always ended in an unsatisfactory ending, as in death still won. 

I think the fear of AI in some people is just the fear of ourselves because we are not collectively uploading human intelligence to the cloud. In its current iteration, It is a super baby that will exaggerate our best and worst features. We have created the perfect consumer of our knowledge, but not the perfect arbiter of that knowledge. 

As we upload our knowledge, we have to admit that humanity is going to lose control of technology. This includes those at the forefront of AI. They will be like everyone else on the street, forced to trust in the magical world that only some of us, including themselves, had a hand in building.

Instead of turning off AI, we need to find ways to include those that are not currently taking part. We need all humans working together, which is also an impossible task. Turning off and on technology is a troubleshooting step because it is a ritual in a complex belief system. At the core, I think that we no longer control technology. Technology controls all of us. When I say us, I mean those that are reliant on technology to survive.

There are people in the world where technology does not control their survival. I am not talking about homesteaders in the libertarian pockets of the world, but large swathes of the population that still farm and utilize the bare minimum of technology to live. I am not saying that is how we should live, far from it. I am suggesting that we upload all of humanity’s knowledge before we turn off the switch. Let’s find ways to do this since I am not aware of any efforts to do so. If there are please let me know.

Per my earlier example, I am the first to admit I know nothing as compared to the rest of the universe, or even just my next door neighbor. I ended up here because I want to fight death in humans, just not the heat death of the universe. I’m not insane. Well, maybe just a little.