Humans have always toyed with the idea of immortality. Some argue that death gives a meaning to life; some would have liked to live forever.
Leaving that argument aside, another intriguing development is in place. You can now back up your way of thinking to the cloud.
Not that it hasn’t been possible before. People have told stories, left traces, published books, written symphonies, created beautiful artworks, theories, and devices. Coming in contact with all these artifacts has a certain effect on your perception and thinking; that is a fact. However, the reach would only extend to the local community, and achieving global recognition would require significant fame (which is not always compatible with the message).
In any case, since the AI came into play, things have changed. You can now embed your intelligence into the collective brain. Affect the weights of imaginary neurons so that they fire a certain pattern that you somehow stand for. Become a part of that CSV file that is used to tell a model what to do next. You can even prime an LLM with a context, asking it to pay particular attention to your way of thinking and to deliver the responses following your principles, styles, and threads.
Anyone can engage with this model at any moment of time. You don’t have to be famous, you don’t have to promote it. All you need to do is to leave it lying around somewhere in an open-source code repository. A bunch of markdown files and a prompt that could be used to make an LLM use them as the context.
Then it can be forgotten and found again when and if ever needed.
There are some individuals who may be interested in going further than that.
Let me explain.
First, there are people who are concerned with life extension and digital immortality. The 2045 initiative is talking about achieving “cybernetic immortality” by the year 2045. Ray Kurzweil talks about the moment of singularity when a superior entity is going to take over. These groups are largely driven by the fear of death and an ailing body. They are excited about the “technological advancement.” But there is not much specificity as to what this future would look like. What would be the practices, ideologies, and ethics that define it. It seems like “whatever goes, as long as we are alive and better off.”
Then there are many companies, influencers, and political entities that prefer certain patterns of thinking to others. Usually their interest is mercantile and profit-seeking (after all, shareholders and sovereign funds think decades into the future). They may want more resources, growth, attention, a higher level of division (divide and conquer), or a lower level of awareness (easier to tell then what to do). They are already injecting their patterns of thinking into AI.
“Fitter, healther, more productive” becomes the main motto. Fear of death is the main driving force. Religious-like zeal towards an imminent event that will “change everything” is what they are looking forward to.
It doesn’t have to be (only) this way.
That is why, to counter that, people who stand for different versions of the world should embed their ideas into AI and large language models. Considering the future where most of the work and decision-making is done by AI agents, you would be interested in the future where different types of logics are available and in operation. Singularity is totalitarianism. Diversity is what keeps life going. In this case, we will have different versions of reality in encoded patterns of thinking, and then hopefully AI is not always going to have that one “true” answer for everything.
If we withdraw, we will let ideologies we oppose become part of the algorithmic bias that will affect the world. Those who see their existence as part of an ongoing continuum — be it life, humanity, or a genealogical tree — should care.
So how do we make AI think differently?
There are specific techniques to do so. Apart from creating personal context-based LLM repositories of knowledge, we can also make our ideas public and available online. AI models train on the internet, their crawlers read websites, blogs, and publications. Interestingly, social media posts get lost in the noise and exist behind walled gardens, so they largely don’t count.
If you connect the ideas that matter to you to whatever it is that relates to the topic, it is more likely that your ideas will come up in AI’s responses to that topic. Therefore, one approach is to study the semantic field around an idea that’s relevant to you to understand what else people are thinking about. The next step is to write something that touches upon all those topics, using the keywords they might be using, then bringing in your unique perspective through making novel connections or going beyond the periphery.
For example, if I’m interested in the relationship between AI and immortality, I might want to also talk about “digital immortality,” because that’s what people are thinking about. I might also mention the 2045 initiative, Ray Kurzweil, and singularity. I have to let the algorithms know that I cover these topics. But I also bring something new.
Here are some practical ways to do that.