#futurology

I recently spent a few hours viewing a huge trove of videos on the YouTube channel Fantasy Universe. These are all generated by some unidentified person or group using artificial intelligence. Sometimes the music sounds like it also was generated by AI, sometimes it sounds more like stock music.

In the following when I say transparent I mean that whatever it is, is obviously and completely understandable, like one step in a rigorous mathematical proof, or one axiom in a well-known formal system; and when I saw opaque, I mean that while the surface may be comprehensible, we do not not easily understand what created the surface or what is really going on under it.

Futritionist is a place for curious forward-thinkers to explore the nexus of nutrition and futurology with an off-duty clinical dietitian.

If you love food and nutrition, futurism, science, and culture, you’re in the right place.

If you don’t care about any of those things but you’re trying to look busy at work, you’re also in the right place.

Why is this nutrition/futurism connection important? Because nutrition is the foundation of every way that we interact with the world. 

Whether your vision of the future is a high-tech, space-exploring, non-scarcity paradise or a dystopian hellscape (or a little of column A, a little of column B), future-humans will still (probably) need to eat. And even if you envision a future where nutrition no longer matters, our nutritional science game will have to be on point to get us there.

Futritionist explores topics like menu fatigue on long term space missions, endocrine changes in zero-G, culinary lessons from HI-SEAS, and more. Although these subjects are heavily speculative in nature, articles here are based on science and referenced where appropriate.

*** Not medical advice *** Not really non-medical advice, either.

A future in which:

Standard issue human beings exist.

AIs have created human bodies with AI brains.

Some humans have direct neural interfaces to AIs.

AIs have created robot bodies with AI brains.

Humans have created robot bodies with human brains (i.e. a direct neural interface to a human's brain).

A scientist at Columbia University has posted a preprint of research on using a new technique of genome editing, called base editing, to change the genes in human embryos without the damaging side effects that are too easily caused by using the CRISPR-Cas9 toolkit.

A tool like base editing would be used to cure genetic defects in human embryos. But the tool could also be used to add or modify genes that go beyond repairing defects, i.e. to add desirable traits to the embryo.

Is this eugenics and, if so, what is wrong with eugenics in this form?

Repairing a severe disease-causing mutation in an embryo is not wrong, even though it is heritable genome editing and requires exceptional caution.

Editing embryos for enhancement is eugenic in a potentially troubling sense: it uses reproductive technology to shape the genetic composition of future persons according to contested ideals of superiority, under conditions likely to intensify inequality, social coercion, biological risk, and the commodification of children.

One obvious issue with enhancement is that it must be done without the informed consent of the person that the embryo will become. Of course, if the person happened to have extraordinary beauty, intelligence, and athleticism purely by natural accident, that person would simply feel very fortunate. And of course, logically speaking, the question of consent applies to removing defects just as much as it does to adding enhancements. I personally see this as a gray area, and that moral objections fade as the unintended side effects of enhancements become improbable.

A more serious objection is that enhancement seems likely to further privilege the already privileged. This is in keeping with other ways in which the increasingly technology-enhanced lives of the middle class also futher privilege the already privileged, e.g. with greater access to high-quality education, or to artificial intelligence.

In a number of recent posts here, I have been exploring the nature and risks of artificial intelligence. This has been based partly on theoretical understanding, and partly on personal experience using AI tools.

Here I simply wish to state, for the record, the following thesis.

There is no such thing as infinite intelligence. Just as no signal can travel faster than light, and in fact for that very reason, physics places absolute limits on computation, and computation is the basis for all intelligence -- even if intelligence has, as I believe it does, a non-computational aspect.

These physical limits ultimately derive from the single fact that doing more than a certain amount of computation in a single volume of space would collapse that volume into a black hole, from which no accessible information could be gained.

Given unlimited time and resources, an evolving intelligence would approach the physical limits on computation. Two or more such competing or warring intelligences could not out-compute each other. If they warred, they could win only by denying their enemy physical resources, just as in any other war. If they were at peace, that union would not constitute a greater intelligence, but simply a greater polity.

The physical limits of computation based on Earth are fantastically larger than our presently active compute. Still, there are limits, and the speed with which we are approaching them appears itself to be increasing.

Paradoxically, such physical limits may indicate a path by which humanity, with the assistance of artificial intelligence at the limit, could defend itself against any adversary, even one possessed of artificial intelligence at the limit.

A possible path.

A few days ago, the New York Times published an article on artificial-intelligence powered computer worms invented at the University of Toronto that can invent new attacks on new vulnerabilities as it spreads. See the original paper here. I can do no better here than quote in full its abstract:

A computer worm is malware that spreads on a network by replicating itself from one machine to another. Traditional worms, like WannaCry, exploited predetermined vulnerabilities, and their spread can be halted by patching those vulnerabilities. Here we show that artificial intelligence (AI) agents enable a fundamentally new threat: a worm that generates tailored attack strategies to each target it encounters. The worm parasitically uses compromised machines to run open-weight large language models (LLMs) to sustain its reasoning, or extend its reach for further attacks. Deployed on a network of machines spanning Linux, Windows, and IoT (Internet of Things) devices, the worm propagated by exploiting common, real-world corporate network vulnerabilities. Since the worm is powered by stolen compute, the attacker’s marginal cost per new infection is zero. This creates a destabilizing economic asymmetry between attackers and defenders. Moreover, because the worm requires no commercial AI platform, centralized safety controls, such as service refusals or rate limiting, are structurally irrelevant. Our results demonstrate that self-sustaining AI driven cyber-threats are no longer theoretical. We must prepare for autonomous generative adversaries: malware systems that propagate without human operators and are defined not by fixed exploit code, but by the capacity to reason about targets, adapt to observations, and synthesize attack logic in real time.