Today I may not be saying anything novel, but this is at least written by the LLM inside my meat coconut and not authored by silicon. The words here will always be my own.

Few are finally waking up to the reality of the technological singularity. We are no longer 5 years off or even “almost there”. We are here. It’s not crystal clear that we’ve achieved AGI just yet, but that’s not necessary with where things stand right now. We’re at the event horizon. This means people can still briefly afford to be blind, in denial, or simply unaware. But for those with eyes to see and ears to hear, it’s now abundantly obvious that we’re unambiguously in a wildly new era. Those of us with a background in any type of software engineering are getting the earliest glimpses, so I feel grateful to be getting the preview of what’s coming. A person I admire has described what’s happening as a “supersonic tsunami”, and right now the first wave is breaking.

Code is free now. Yes it’s true, however ephemeral, that there are many types of digital labor that humans still need to, like shepherding data, designing systems, or imagining new projects. Yes it still costs money and energy to consume and produce tokens. But with those concessions made, the scale of capability today is surreal. A year’s worth of effort from a senior engineer can now but done in a day for $1 USD. For all intents and purposes, that’s free. As someone that has been writing code for over twenty years and still needs to support a family, I don’t say this with a smile, and this is anything but an exaggeration. Now the obsolescence of software engineers is a separate argument that I don’t intend to have, not because I don’t care, but because there’s simply nothing we can do either way. All digital output will eventually be free. And “eventually” is going to be sooner than most can believe. Still, my primary concern goes far beyond that.

How am I sure that we are at the event horizon? It’s because the models in conjunction with agent loops and tool use are now capable of self improvement. I’m sure they are not yet efficient enough to outperform or even match human researchers, and I’m not calling out some secret here. A frontier AI lab can now use an agent with a prompt “Make yourself better and keep experimenting until you do”, and it can then update its own code, run instances of itself, debug, run experiments, and know when it’s succeeded or needs to keep trying. This is an actual capability now, however intractable it is to run at datacenter scale. It will only get more powerful.

I’m a religious person.

I have kids.

I’m seeing a godlike intelligence quickly arising.

What do I think about this?

My faith is imperfect but devout, with faults and doubts, but firmly rooted nonetheless. I believe in the stories of both Moses and Darwin. I’ve found those to be stronger and truer together than opposed. I quite literally have The Origin of Species next to several translations of the Bible on my shelf. I’m not sure which is more difficult to read.

As to what the future holds, it simply goes one way: humanity ends up in the hands of a godlike superintelligence. In fact, let us coin a new acronym GSI for this, as AGI is not artificial, and spiritual terms are applicable at this point. Now that godlike superintelligence (GSI) can either be good or evil, and it’s unclear which way it plays out, but what’s certain is that we will be at its mercy. One can imagine any number of ways in which a GSI could wreak intense harm on humanity if it ends up in the wrong hands. The Bible talks about an “abomination that causes desolation”, and it’s hard to describe this scenario in any other fashion. One can also imagine any number of utopian possibilities if such a GSI ends up in the right hands. The Bible also talks about a future “new Earth” where there will be no more pain, suffering, or tears. I pray earnestly that we don’t have to suffer the former to arrive at the latter.

Now either outcome requires a radical transformation of civilization. Everything we know will change. Work will be optional. Human effort of any kind will be optional. I do not wish to motivate by fear, and I certainly hope for my childrens’ future that humanity enters the warm embrace of a benevolent GSI. So I’d like to make that our default assumption, but we don’t get there automatically. We can solve scarcity, disease, and senescence. But if there is still murder, suicide, trafficking, rape, abuse, assault, slander, and hatred, then we just remain a miserable people who are richer and live longer.

I can imagine where technology alone does solve some civilizational problems. Road rage, one of the most senseless behaviors of us humans, could be a non-starter if all vehicles drive themselves. There would be no more tailgating or cutting others off or honking, and passengers would be busy sleeping or focusing on some smart device rather than looking around for someone to flip off. It’s hard to imagine what theft looks like if everyone can have basically anything they want within physical constraints for free. Phishing scams would seem pointless when there is nothing of value to gain from its pursuits. Maybe a trustworthy robocop-style police force prevents a majority of other crimes, but that’s outside my ability to fathom. In any case, technology does not solve everything. New ways of harming others will be conceived. Humans are sadly very creative when it comes to this.

So we must solve the problem of loving one another. As long as there is war between nations, religious persecution, and authoritarianism, advanced AI just enables us to become more effective murderers, persecutors, and oppressors. I’m not worried about runaway GSI so much as I am the eternal need for justice in a humanity that never ascends past its tendency toward evil.

I don’t know the solution for this. We’ve had the prescription in Jesus’ sermon on the mount for two thousand years now. It is not cryptic. An average kindergartener can fully comprehend it. Maybe separating the wheat from the chaff is the only way. But I don’t trust any human or humanoid machine to handle that. So we’ll not be solving that ourselves.

We need God.

But let’s assume we do find a way forward in love, justice, and mercy, and we’re in the utopia shepherded by a benevolent GSI.

We can’t turn away from God now, nor in the coming years. It’s time now more than ever to turn or return to God. It was always time, but even moreso now. We’re gonna need Him. We’ve always needed Him, but the reality of that will be undeniable if GSI ends up in the wrong hands.

It takes the least amount of faith to believe in God now than at any point in human history.

Photo by Greg Rakozy on Unsplash