What’s a Platonic Biosingularity?

On finding biological free lunches

Author’s note: a very informal draft.

I.

A “singularity” is when technology goes asymptotic. I.e. making AIs that can make smarter AIs, who make smarter AIs, etc. “100 years of change in two weeks” sort of stuff; things get weird very quickly.

A “biosingularity” is the same process but with biology — making smarter humans who make smarter humans. People like Elon Musk, Vitalik Buterin, and Anatoly Karlin believe this could be safer because we’d be part of the process, and could leverage the “bio alignment” tech we’ve evolved over millennia.

But biosingularities could still get really weird really quickly. There’s a third option I’m calling a “platonic biosingularity” — the basic idea is to try to amplify the essence of existing organisms by extrapolating them with zero damage.

II.

Plato talks about the “forms” of things — the platonic form of a cat is the cattiest cat you’ve ever seen. The platonic form of a buffalo is the highest, most distilled version of a buffalo. Etc. These forms tend to be beautiful.

Whether Plato’s forms exist “for real” is an interesting question. I dismissed the possibility at first — “ha ha ancient philosophy man, clever argument but we have science and evolution now” —

But after learning more about physics, I find e.g. Tegmarkian mathematical platonism to be plausible.

Regardless, the concept of “technological platonic extrapolation” is not about whether the forms exist metaphysically, but rather whether we can distill and extrapolate organisms into a healthier, more quintessential state. Whether the Greek demigod version of Bryan Johnson exists for him to turn into.

III.

There’s an eternal debate in biology over tradeoffs vs damage. Some believe that adaptation is so contextual that any changes involve deep tradeoffs. Even aging, they argue, is planned obsolescence — evolution ensuring we don’t get locked into a genetically stale geritrocracy. In this view there are no free lunches.

Others believe that some things (such as aging) can be straightforwardly understood as damage. Many genetic mutations are straight-up harmful, and could be fixed with no downsides. There can be free lunches, if we can find reservoirs of damage and fix them.

The “platonic biosingularity” frame suggests humans have a lot of invisible damage weighing them down, and if we fixed all such damage we might see startlingly large effects. And it would be ‘clean’ in a way that other singularities wouldn’t be — humans would just get straight-up better. More healthy, more happy. A reversion to the platonic mean.

Whether or not this frame makes sense, and where these hidden reservoirs of damage are, are among the most important questions of the century. My research points to three reservoirs of biological damage that are endemic in humans: genetic load, pathogens, and trauma.

IV.

1. Genetic load: we accumulate genetic typos because cell division is noisy. Empirically, it looks like we accrue about ~2 additional major mutations per generation, and cleanse ~2 per generation (mostly through the womb spontaneously aborting high-mutation embryos), with a steady-state absolute level of ~100 major mutations (MacArthur) and ~1000 minor mutations (Hsu).

The biological optimist says: what if we just fixed all of these with something like CRISPR or whole-chromosome synthesis?

Hsu has a simple statistical model that suggests we could get +15 standard deviations of improvement if we fixed all minor errors in a genome; this doesn’t count the biggest mutations, since they’re rare enough they’re difficult to evaluate statistically.

+15SD is a lot. Even +5SD would be huge.

A particularly compelling element of this is that fixing genetic typos wouldn’t homogenize people; it would simply make people healthier, happier, smarter, “more themselves”. If you like your eye color, you can keep your eye color. But with fewer missense variants that lower your biochemical fitness.

V.

2. Pathogens and parasites: Riva Tez has a fantastic essay on the invisible damage common pathogens and parasites do. We talk about “long covid” but we should talk about “long everything”: https://www.hardtowrite.com/pathogens/

Health often inflects downward after an infection. It makes unfortunate sense — pathogens’ evolutionary niche involves fucking up their hosts’ ability to defend themselves. Sometimes they’re going to get some good hits in.

A consistent pattern in population genetics is the bottleneck-expansion — if a population gets very small, it may no longer be large enough to support the parasites which feed off it it, and some of these parasites may vanish. The extra vitality from the reduction in parasite load then enables a rapid population expansion.

I hope humans don’t go through such a population bottleneck. But we do have things like CRISPR and vaccines; if we could use these to destroy, say, the worst 5 viral offenders for permanent damage, humanity’s health trajectory would magically improve overnight.

I recall reading one study (which sadly I can’t seem to find) which autopsied nuns’ brains. They found them to be particularly healthy well into their 90s — in particular, they found very little of the scarring and plaques which accumulates from HSV-1 (the virus that causes cold sores, spread by kissing). Approximately 67% of all people in the world have HSV-1. What would life be if 0% of people had it?

And how many other infections cast a shadow on health? What would the world look like if we could destroy HSV-1, EBV, Covid, & CMV? I expect we would age much more slowly, and be significantly smarter, healthier, more beautiful by default.

Smallpox killed ~500 million people over the last hundred years of its existence. Since 1980, it’s killed zero people. Eliminating pathogens is possible, at least with 1970 social technology.

This is also why I believe covid was a massive crime against humanity. I believe it was created in a lab, and that we should not be funding the creation of such things in labs, and we should work to change the conditions which allowed this to happen.

VI.

3. Trauma. I think everyone’s walking around with a lot of technical debt in their nervous system.

An aspiration of 1970s psychedelic research was to find a penicillin for the soul; it’s a beautiful metaphor, and might even be possible. I hope to say more on this later.

VII.

Ultimately, the widespread presence of damage in the average human is a huge whitepill, because it implies clean upside.

If we fix this damage, the world gets better;

If we do this in a systematic way, we could get a platonic biosingularity.

Eventually, we’d run out of damage to fix and the “platonic vector” would fizzle out. Platonic humanity would have a ceiling. But I expect the ceiling is very high and we’d be in a good position to cross that bridge when we came to it.