God in the Loop: How a Causal Loop Could Shape Existence

Cross­posted from Ves­sel Pro­ject.

My last ar­ti­cle, “Life Through Quan­tum An­neal­ing” was an ex­plo­ra­tion of how a broad range of phys­i­cal phe­nom­ena — and pos­si­bly the whole uni­verse — can be mapped to a quan­tum com­put­ing pro­cess. But the ar­ti­cle sim­ply ac­cepts that quan­tum an­neal­ing be­haves as it does; it does not at­tempt to ex­plain why. That an­swer lies some­where within a “true” de­scrip­tion of quan­tum me­chan­ics, which is still an out­stand­ing prob­lem.

De­spite the mas­sive pre­dic­tive suc­cess of quan­tum me­chan­ics, physi­cists still can’t agree on how its math cor­re­sponds to re­al­ity. Any such pro­posal, called an “in­ter­pre­ta­tion” of quan­tum me­chan­ics, tends to strad­dle the line be­tween physics and philos­o­phy. There is no short­age of in­ter­pre­ta­tions, and in the words of physi­cist David Mer­min, “New in­ter­pre­ta­tions ap­pear ev­ery year. None ever dis­ap­pear.” Am I go­ing to throw one more on that pile? You bet.

I’m not go­ing to start from scratch though; I sim­ply pro­pose an ever-so-slight mod­ifi­ca­tion to an ex­ist­ing fore­run­ner: the many-wor­lds in­ter­pre­ta­tion, where other “wor­lds” or timelines ex­ist in par­allel to our own. My mod­ifi­ca­tion is this: the only wor­lds that can ex­ist are those that ex­ist within a causal loop. Stated an­other way: our uni­verse, or any pos­si­ble uni­verse, must be a causal loop.

I will in­tro­duce the rele­vant con­cepts and provide an ar­gu­ment for my pro­posal, but my goal is not to once-and-for-all prove this in­ter­pre­ta­tion as true. Rather, my goal is to ex­plore what hap­pens if we ac­cept the in­ter­pre­ta­tion as true. If we start with the as­sump­tion that only causal loop uni­verses can ex­ist, then sev­eral in­ter­est­ing things fol­low — we find par­allels to our own uni­verse, and we might even find God.

Causal­ity & Quan­tum Interpretations

Be­fore talk­ing about causal loops, let’s take a step back and talk about causal­ity — per­haps the sin­gle most fun­da­men­tal con­cept in all the sci­ences. It plays a star­ring role in the two most im­por­tant the­o­ries in physics: gen­eral rel­a­tivity and quan­tum me­chan­ics.

Gen­eral rel­a­tivity, de­vel­oped by Ein­stein, com­bines space, time, and grav­ity in a ge­o­met­ric de­scrip­tion of space­time. In space­time, two ob­servers might not agree on the space be­tween two events or time be­tween two events — but they always agree on the space­time in­ter­val, which cor­re­sponds to a causal re­la­tion­ship be­tween two events. As such, causal­ity is the only thing that is uni­ver­sally agreed on, mak­ing it the only proper de­scrip­tion of ob­jec­tive re­al­ity. Another phe­nomenon pre­dicted by gen­eral rel­a­tivity is the cos­mic speed limit — the speed of light — which is more ap­pro­pri­ately un­der­stood as the speed of causal­ity, more fun­da­men­tal than light alone. Here we see that space­time and the speed of light are not in­her­ently real; they are just use­ful ways of de­scribing causal­ity, the only ob­jec­tive re­al­ity.

But if gen­eral rel­a­tivity is in­ter­est­ing be­cause we can only agree on causal­ity, then quan­tum me­chan­ics is in­ter­est­ing be­cause we can’t agree on causal­ity.

As I al­luded to ear­lier, the full ex­pla­na­tion of quan­tum me­chan­ics is still a mys­tery, and that mys­tery has ev­ery­thing to do with causal­ity — speci­fi­cally how ob­jec­tive, causal re­al­ity re­lates to the wave func­tion. The wave func­tion of a quan­tum sys­tem is most eas­ily un­der­stood as a prob­a­bil­ity dis­tri­bu­tion, where the prob­a­bil­ity of the sys­tem be­ing in any given state is calcu­lated when you square the am­pli­tude of the wave func­tion for that state. The wave func­tion is in a “su­per­po­si­tion” of all pos­si­ble states un­til it is mea­sured, af­ter which we ob­serve a sin­gle state.

Sim­ple de­pic­tion of a quan­tum wave func­tion with a sin­gle crest. (Image by Louay Fa­toohi)

On the sur­face, it ap­pears that quan­tum physics is in­her­ently ran­dom if it can only be de­scribed by prob­a­bil­ity, and some­how the act of mea­sur­ing or ob­serv­ing the sys­tem causes it to as­sume an ob­jec­tive state. If you’re con­vinced of this, then you ba­si­cally agree with the Copen­hagen in­ter­pre­ta­tion of quan­tum me­chan­ics, which posits an in­ter­ac­tion be­tween the sys­tem and ob­server that causes the wave func­tion to ran­domly “col­lapse.” This has been the stan­dard in­ter­pre­ta­tion for a long time, al­though oth­ers in­ter­pre­ta­tions have been con­sis­tently gain­ing steam.

As an al­ter­na­tive, maybe you don’t think the uni­verse is ran­dom at all — it’s de­ter­minis­tic, but there are “hid­den” vari­ables we don’t yet know about. In this case, there is no wave func­tion col­lapse, so we don’t need to in­tro­duce any ex­tra physics to ex­plain what hap­pens when we ob­serve the sys­tem. If you’re on board with that, then you just signed up for the de Broglie-Bohm the­ory, also known as the pi­lot wave the­ory.

But maybe you’re still not quite con­vinced. Let’s make things even sim­pler: the uni­verse isn’t ran­dom, but there aren’t hid­den vari­ables ei­ther. You have the wave func­tion, and that’s it — what you see is what you get. That, it turns out, pretty much sums up the many-wor­lds in­ter­pre­ta­tion. In this the­ory, all pos­si­ble “wor­lds” de­scribed by the wave func­tion do ex­ist; we just hap­pen to oc­cupy one of them. While it re­quires the least ex­pla­na­tion, the bizarre im­pli­ca­tion is that many par­allel wor­lds ex­ist as branches of differ­ent pos­si­ble out­comes.

Th­ese three in­ter­pre­ta­tions tend to be the top con­tenders, and they each take a differ­ent ap­proach to an­swer the ques­tion of “what causes what?” The fact that a ba­sic causal struc­ture of physics can­not gain con­sen­sus makes this in­ter­est­ing ter­ri­tory, plus it has far-reach­ing im­pli­ca­tions. A proper ex­pla­na­tion doesn’t just ac­count for the non-lo­cal­ity of en­tan­gle­ment or the ap­par­ent un­cer­tainty of su­per­po­si­tion — it ex­plains how hu­mans fit into re­al­ity.

As ob­servers, do we cause the wave func­tion to col­lapse? That would cer­tainly seem to ele­vate the role of con­scious­ness in the causal na­ture of re­al­ity (which has not gone un­no­ticed by ex­perts and cranks al­ike). Or is the wave func­tion it­self the only causal, ob­jec­tive re­al­ity? If so, that’s one more rea­son to be­lieve we’re at the mercy of a uni­verse that’s in­differ­ent to our ex­is­tence.

The third in­ter­pre­ta­tion is the one I want to re­visit later in this ar­ti­cle: the many-wor­lds in­ter­pre­ta­tion. Keep it in mind. While it ap­pears to threaten our sense of im­por­tance and po­ten­tially free will, it’s not so bad if we just add a twist — or bet­ter yet, a loop.

Causal Loops

It may be fairly self-ex­plana­tory, but I’ll nonethe­less define a causal loop as fol­lows: a closed causal chain of events, where each event is the effect of an­other event on the chain. A sim­ple ex­am­ple is a loop of three dis­tinct events where Event A causes Event B, which causes Event C, which in turn causes Event A — each event is causally con­nected to an­other on the loop, and there is no “first” event. If you start at any one event, the se­quence that fol­lows in­evitably leads back to the same event as if it caused it­self.

Causal loops sound ab­surd, but their pos­si­bil­ity has been suc­cess­fully defended, par­tic­u­larly by philoso­pher Richard Han­ley in his pa­per, “No End in Sight: Causal Loops in Philos­o­phy, Physics, and Fic­tion” (and any men­tion of Han­ley mov­ing for­ward is in refer­ence to this pa­per). Han­ley points out that causal loops are not log­i­cally in­con­sis­tent nor phys­i­cally im­pos­si­ble — at worst, they sim­ply re­quire grand co­in­ci­dences. Causal loops as a whole aren’t cre­ated, they sim­ply ex­ist, and any strangeness about this is merely ap­par­ent — they’re in no worse a po­si­tion con­cern­ing the ques­tion of why any­thing ex­ists. In­ter­est­ingly, Han­ley also men­tions that the idea of a causal loop uni­verse is taken very se­ri­ously in cos­mol­ogy, and that causal loops are ac­tu­ally more likely to oc­cur in a uni­verse like ours which hosts in­tel­li­gent agents.

Let’s un­pack that a bit. If in­tel­li­gent agents dis­cover their uni­verse is in­deed a gi­ant causal loop, they may have in­cen­tive to main­tain the loop they in­habit by caus­ing the very events that lead to their own ex­is­tence. Fur­ther­more, they can in­ten­tion­ally make events hap­pen that would oth­er­wise re­quire co­in­ci­dence. But there is noth­ing co­in­ci­den­tal or mys­te­ri­ous about an in­ten­tional ac­tion; we in­ten­tion­ally do things ev­ery day. Han­ley notes, “the ex­is­tence of agency may be the very thing that per­mits causal loops to ob­tain.”

This is where I’ll take it one step fur­ther than Han­ley: not only are causal loop uni­verses pos­si­ble, but all pos­si­ble uni­verses must be causal loops. As I men­tioned ear­lier, I’m go­ing to run with this as an as­sump­tion, but I’ll still at­tempt to provide some rea­son­ing.

That rea­son­ing boils down to two propo­si­tions: the first is that all events must have causes; the sec­ond is that only in closed causal chains do all events have causes. We saw that in a causal loop all events have defini­tive causes — other events on the loop. There is no is­sue. But in an open causal chain (imag­ine a straight line), one more event is always re­quired to ex­plain cau­sa­tion. We’re left with a case of in­finite regress, which isn’t in­her­ently prob­le­matic, but its “open­ness” im­plies there must be an event with­out a cause, which is im­pos­si­ble. Fur­ther­more, any se­ries of causes and effects can­not, by defi­ni­tion, be con­sid­ered as part of sep­a­rate causal sys­tems. If we define a uni­verse to be a causal sys­tem, then it fol­lows that all uni­verses must also be causal loops, in­clud­ing our uni­verse.

Us­ing a causal-loop-only start­ing point, we can dive into some pretty in­ter­est­ing things.

Differ­ent Paths: Curved Space­time & Clever De­mons

In gen­eral rel­a­tivity, causal loops are per­mis­si­ble in the con­text of a “block uni­verse.” In causal loops, all events in the loop are equally real all the time; they must be for the loop to ex­ist. This closely al­igns with a block uni­verse, where all past, pre­sent, and fu­ture points in space­time ex­ist “at once”; we sim­ply find our­selves at one point along its pro­gres­sion. In both views, trav­el­ling back to the past is pos­si­ble, but you can­not change the past. If you do travel back to the past though, you may find your­self trav­el­ling along a differ­ent timeline af­ter that — which brings us back to quan­tum me­chan­ics.

You took note of the many-wor­lds in­ter­pre­ta­tion (MWI) of quan­tum me­chan­ics, right? If the uni­verse is a causal loop, then what­ever in­ter­pre­ta­tion we use must be de­ter­minis­tic since all events along a causal loop are equally real — they do not spon­ta­neously be­come real only af­ter an­other event. Tech­ni­cally any de­ter­minis­tic in­ter­pre­ta­tion suffices to meet that crite­ria, but I think the MWI best illus­trates the range — and re­stric­tions — of how pos­si­ble uni­verses can un­fold. The MWI en­tails the uni­verse “split­ting” into al­ter­nate his­to­ries at ev­ery point in time. If we in­tro­duce a con­straint where only causal loop uni­verses can ex­ist, that di­rectly im­pacts the range of pos­si­ble uni­verses we can ever split into. Noth­ing else about the MWI needs to change; there are still many par­allel wor­lds, but they all must main­tain a causal loop. So if we some­how trav­eled back in time, we may find our­selves split­ting into a differ­ent looped timeline than be­fore.

Things get in­ter­est­ing when we start to look at pos­si­ble loops. There are only two ways a causal loop can be main­tained in prac­tice: closed timelike curves (CTCs), and re­verse cau­sa­tion. While the two en­tail similar­i­ties, they are slightly dis­tinct.

CTCs are the­o­ret­i­cally pos­si­ble in cer­tain solu­tions of space­time. One ex­am­ple, pop­u­lar in sci­ence fic­tion, is a worm­hole. In some worm­holes, you’ll en­ter one end and exit the other at a pre­vi­ous point in time. But there is se­ri­ous doubt about whether they could be fea­si­bly trav­eled through, plus they’re just lo­cal anoma­lies. If we’re talk­ing about the whole uni­verse, we need to go big­ger.

The great lo­gi­cian Kurt Gödel did find one solu­tion to Ein­stein’s equa­tions, now called the Gödel uni­verse, where the en­tire uni­verse is a CTC. In such a uni­verse, all points in space­time re­turn to them­selves as we’d ex­pect in a causal loop, but it re­quires that all galax­ies have a preferred di­rec­tion of ro­ta­tion, for which there is no ev­i­dence. When Gödel found his solu­tion, the tools used to study cos­mol­ogy were not pow­er­ful enough to con­firm if our uni­verse was a Gödel uni­verse. As the tech­nol­ogy be­came more so­phis­ti­cated through­out his life up un­til his death in 1978, Gödel would ask, “Is the uni­verse ro­tat­ing yet?” The an­swer was always no. As best as we can tell, our uni­verse is not a gi­ant CTC, but Gödel might not be out of luck just yet.

A Gödel uni­verse rep­re­sented by “light cones” and a pos­si­ble path of a light through space­time.

Re­v­erse cau­sa­tion, as Han­ley defines it, is sim­ply “a cause and effect re­la­tion where effect pre­cedes cause.” Any no­tion of re­verse cau­sa­tion, or causal loops in gen­eral, is in­ti­mately tied to in­for­ma­tion. Every sin­gle event or state of the uni­verse ex­ists in terms of in­for­ma­tion, as does each causal re­la­tion­ship. In­for­ma­tion is also what makes events dis­tinct and unique. If the uni­verse were to sud­denly re­turn to some state X that ex­isted an hour ago — in­for­ma­tion­ally iden­ti­cal in ev­ery way — then we’re not talk­ing about an­other state similar to X; that is X. Each event in a causal loop is fully and uniquely de­scribed by in­for­ma­tion.

One fea­ture of our uni­verse is that in­for­ma­tion be­comes in­creas­ingly diffuse, a nat­u­ral re­sult of the sec­ond law of ther­mo­dy­nam­ics, which holds that the uni­verse always trends to­ward max­i­mum en­tropy, or equil­ibrium. En­tropy can be un­der­stood as a mea­sure of di­s­or­der; it always tends to in­crease lo­cally, but the over­all en­tropy of the uni­verse stays con­stant. Said an­other way: al­though in­for­ma­tion is never ac­tu­ally lost, it tends to be­come more di­s­or­dered.

Therein lies our grand dilemma. As physi­cist Lee Smolin writes in The Sin­gu­lar Uni­verse, “The fact to be ex­plained is why the uni­verse, even 13.8 billion years af­ter the Big Bang, has not reached equil­ibrium, which is by defi­ni­tion the most prob­a­ble state, and it hardly suffices to ex­plain this by as­sert­ing that the uni­verse started in an even less prob­a­ble state than the pre­sent one.” How did the uni­verse ever ar­rive at a more or­dered state when it clearly prefers the op­po­site? Ob­vi­ously it’s a co­nun­drum in our ex­ist­ing mod­els, but dou­bly so if we are to imag­ine a fu­ture in our causal loop that goes to­tally against a law of na­ture. This ques­tion has already drawn some eye­brow-rais­ing pro­pos­als.

Lud­wig Boltz­mann, the 19th cen­tury physi­cist who de­vel­oped the sec­ond law of ther­mo­dy­nam­ics, gave one pro­posal: the sec­ond law is a statis­ti­cal phe­nomenon, so given enough time, there’s a non-zero chance the uni­verse will ran­domly fluc­tu­ate back into a low-en­ergy state. But ac­cord­ing to Boltz­mann’s own prin­ci­ples, some­thing like the big bang is liter­ally the least likely thing that can hap­pen; while not nec­es­sar­ily im­pos­si­ble, we’re go­ing to ex­plore a more prob­a­ble sce­nario.

A con­tem­po­rary of Boltz­mann, James Clerk Maxwell, de­vised a thought ex­per­i­ment called “Maxwell’s de­mon” in an at­tempt to vi­o­late the sec­ond law. He imag­ined a de­mon that con­trol­led a small door be­tween two gas cham­bers. As in­di­vi­d­ual gas molecules ap­proached the door, the de­mon would quickly open and close it so that all the fast molecules be­came trapped in one cham­ber, and the slow molecules in the other. In do­ing so, Maxwell pro­posed the sec­ond law was vi­o­lated since the cham­ber sys­tem be­came more or­dered; one side be­came hot­ter and the other be­came cooler, even though it was to­tally mixed be­fore. With re­gard to in­for­ma­tion, en­tropy had been low­ered — or so he thought.

In this Maxwell’s de­mon setup, cham­bers A and B both start with mixed gas, but over time cham­ber A be­comes cold and cham­ber B be­comes hot. (Source: Htkym /​ CC BY-SA)

Others said not so fast. Although en­tropy in the cham­bers de­creased, the en­tropy in the de­mon’s mem­ory in­creased. Imag­ine that the de­mon’s mem­ory started as a blank slate — highly or­dered. As it ob­served the sys­tem, it had to fill its mem­ory with in­for­ma­tion about the gas molecules to know how to op­er­ate the door. In do­ing so, the in­for­ma­tion in its mem­ory be­came more di­s­or­dered, thereby pre­serv­ing the sec­ond law.

But the de­mon can just for­get that in­for­ma­tion, right? In do­ing so, its mem­ory goes back to a blank slate, but the gas is still highly or­dered. Seems like an easy solu­tion. Again, not so fast — the loss of in­for­ma­tion en­tails a dis­si­pa­tion of heat, which in­creases the en­tropy of its sur­round­ings. Alas, it seems the sec­ond law can­not be slayed. But maybe it doesn’t need to be.

When look­ing at the en­tire sys­tem in Maxwell’s thought ex­per­i­ment — which re­ally in­cludes the cham­bers, the de­mon, and the de­mon’s en­vi­ron­ment — we no­tice sev­eral things. One is that in­for­ma­tion can take sev­eral forms, such as the prop­er­ties of gas, mem­ory in the brain, and in the effects of heat. Another is that al­though the sec­ond law is main­tained and en­tropy’s trend to­ward di­s­or­der never ceases, lo­cal ar­range­ments of in­for­ma­tion can be­come more or­dered, thus lo­cal en­tropy can de­crease. To re­it­er­ate an ear­lier point: over­all en­tropy never changes, but lo­cal en­tropy can. A third ob­ser­va­tion con­cerns what is re­quired to pro­duce lo­cal or­der: the de­mon. More gen­er­ally, knowl­edge about the sys­tem, or mem­ory, as well as the abil­ity to act upon it to re­ar­range in­for­ma­tion. In fact, if an agent has perfect knowl­edge of a sys­tem, it can re­ar­range it in any way it de­sires.

Maybe you can see where this is go­ing — in­tel­li­gence can ma­nipu­late in­for­ma­tion, and enough in­tel­li­gence can hy­po­thet­i­cally recre­ate a prior state of in­for­ma­tion in its own sys­tem, main­tain­ing a causal loop.

Let’s re­cap a bit: if we as­sume our uni­verse is a causal loop, but it is not a CTC, and it prob­a­bly did not ran­domly fluc­tu­ate to a highly-or­dered state, then the only op­tion left is to think that in­tel­li­gence was used to cause a pre­vi­ous, highly-or­dered state in the loop.

You may think “yeah, but what are the odds of that?” I’m in­clined to re­spond with, “bet­ter than the al­ter­na­tives.” Re­mem­ber, Han­ley tells us these things are not im­pos­si­ble, they are merely co­in­ci­den­tal; and causal loops are more likely to hap­pen in a uni­verse with in­tel­li­gent agents. If a causal loop is the only type of uni­verse that can ex­ist, then it’s not co­in­ci­den­tal at all; it’s sim­ply how any­thing must ex­ist. That alone elimi­nates the ap­par­ent ab­sur­dity. And al­though we’re work­ing with a sam­ple size of one, the fact that our uni­verse hosts in­tel­li­gent life already makes the “in­ten­tional causal­ity” path more prob­a­ble than a ran­dom fluc­tu­a­tion.

I’ll also add that this al­igns with my dis­cus­sion on quan­tum an­neal­ing, where a quan­tum an­neal­ing uni­verse con­verges on it’s high­est prob­a­bil­ity state. If the many par­allel timelines in the MWI fol­low a prob­a­bil­ity dis­tri­bu­tion, and all timelines must form causal loops, then not only are the most prob­a­ble loops are those that con­tain in­tel­li­gence, as Han­ley sug­gests, but each loop that takes the in­tel­li­gence “route” must ul­ti­mately land on a set of com­mon char­ac­ter­is­tics — they must all have the abil­ity to ma­nipu­late in­for­ma­tion, or re­al­ity it­self, in or­der to main­tain a causal loop. If any one of them did not con­verge on this knowl­edge or tech­nolog­i­cal so­phis­ti­ca­tion, then the timeline would not ex­ist in the first place, thus would not be in­cluded in the prob­a­bil­ity dis­tri­bu­tion. As such, any timeline we fol­low in the causal-loop-MWI for­mu­la­tion must con­verge on those traits too.

I also ex­plained how a re­ward func­tion within quan­tum an­neal­ing would re­sult in the sys­tem hav­ing in­cen­tive to “restart” it­self in or­der to max­i­mize re­ward. Both causal loops and a quan­tum an­neal­ing uni­verse in­volve a con­ver­gence on in­tel­li­gence to fa­cil­i­tate a restart, and they in­volve the act of “for­get­ting” in or­der to re­store a pre­vi­ous in­for­ma­tional state. And al­though it’s far cry from any firm proof, this heat-re­leas­ing for­get­ting pro­cess sounds a lot like our early uni­verse — a hot uni­verse with a highly-or­dered state.

From my per­spec­tive, causal loops and quan­tum an­neal­ing look like two sides of the same coin. Is it a co­in­ci­dence that we seem to ar­rive at the same con­clu­sions from two en­tirely differ­ent ap­proaches? Or have we done away with co­in­ci­dences?

Make It Loop: A How-To Guide

We ma­nipu­late in­for­ma­tion ev­ery day, whether it be phys­i­cally, men­tally, or digi­tally, but we could use more guidance in the way of re­set­ting a uni­verse — it’s a tall or­der. In­for­ma­tion and en­tropy can take many forms, but there does ap­pear to be one form that rules them all: Von Neu­mann en­tropy. I couldn’t pos­si­bly sum­ma­rize it bet­ter than physi­cist Matt O’Dowd from PBS Space Time, so I won’t try:

Quan­tum en­tropy, also known as Von Neu­mann en­tropy . . . de­scribes the hid­den in­for­ma­tion in quan­tum sys­tems, but more ac­cu­rately, it’s a mea­sure of en­tan­gle­ment within quan­tum sys­tems. In fact, the evolu­tion of quan­tum en­tan­gle­ment may be the ul­ti­mate source of en­tropy, the sec­ond law, the limits of in­for­ma­tion pro­cess­ing, and even the ar­row of time.

Von Neu­mann en­tropy is of par­tic­u­lar in­ter­est in the study of quan­tum in­for­ma­tion — namely, in black holes and quan­tum com­put­ing. One foun­da­tional tenet of quan­tum the­ory is that quan­tum in­for­ma­tion is never lost or de­stroyed. This pre­sented a real prob­lem in the “black hole in­for­ma­tion para­dox” where physi­cist Steven Hawk­ing pointed out that in­for­ma­tion seemed to be for­ever lost through what he called Hawk­ing ra­di­a­tion, where in­for­ma­tion-car­ry­ing par­ti­cles fall into a black hole, adding to its mass, but this same mass can es­cape through in­for­ma­tion­less pho­tons, thereby eras­ing in­for­ma­tion.

Many physi­cists thought this para­dox couldn’t pos­si­bly be, so they de­vised sev­eral solu­tions to re­solve it. Hawk­ing him­self even aban­doned the para­dox, con­vinced that in­for­ma­tion was pre­served. One promis­ing solu­tion uses en­tan­gle­ment, the phe­nomenon whereby two par­ti­cles, or qubits, must be de­scribed as a sin­gle state. In this solu­tion, the pho­tons es­cap­ing through a black hole’s ra­di­a­tion are im­printed with in­for­ma­tion through en­tan­gle­ment — in­for­ma­tion that can the­o­ret­i­cally be re­trieved. Nor­man Yao, from the Univer­sity of Cal­ifor­nia, Berkeley, told Quanta Magaz­ine, “If you were God and you col­lected all these Hawk­ing pho­tons, there is in prin­ci­ple some un­godly calcu­la­tion you can do to re-ex­tract the in­for­ma­tion in [each swal­lowed] qubit.”

Is a literal God re­quired to gather the in­for­ma­tion needed to con­nect our loop? Maybe, but I’m only hu­man, so it’s be­yond me. Per­haps it’s not the only op­tion though. What if we don’t need to know ev­ery­thing; we just need to know enough? As in­tel­li­gent be­ings, we do have the abil­ity to rea­son af­ter all. Can we ar­rive at the nec­es­sary in­for­ma­tion by means of de­duc­tion, with­out hav­ing all the raw data? A step in that di­rec­tion might con­cern en­tan­gle­ment; it doesn’t just save in­for­ma­tion from be­ing lost in our uni­verse — it might show us how to build a new one.

One im­pli­ca­tion of the en­tan­gle­ment solu­tion to the black hole para­dox is that our uni­verse may be a holo­gram. It sounds rather strange, but the “holo­graphic prin­ci­ple” is taken quite se­ri­ously and is of great in­ter­est in the quest for quan­tum grav­ity. In this ap­proach, space­time emerges from a net­work of en­tan­gled par­ti­cles, and our en­tire uni­verse may be a holo­gram of in­for­ma­tion en­coded on the sur­face of a black hole. This is where we may be able to make some progress.

As I men­tioned, Von Neu­mann en­tropy is also rele­vant to quan­tum com­put­ing. In fact, there are re­mark­able par­allels be­tween black holes and quan­tum com­put­ing, and the more we study one, the more we tend to learn about the other. Ad­vance­ments in quan­tum com­put­ers al­low us to probe the mys­ter­ies of our uni­verse. We’ve already been able to do some pretty mind-bend­ing things with ex­per­i­men­tal sys­tems, like those that mys­te­ri­ously “snap back” into or­der from equil­ibrium, en­tan­gle par­ti­cles over time (not just space), re­verse time, and challenge our no­tion of nor­mal causal or­der. In time, we may come to find that we ac­tu­ally live in a quan­tum com­puter; which means — in keep­ing with a causal loop — we’ll recre­ate the uni­verse through quan­tum com­put­ing too.

It’s no se­cret I’m a strong pro­po­nent of one par­tic­u­lar form of quan­tum com­put­ing as a model of our uni­verse: quan­tum an­neal­ing. In al­ign­ment with the holo­graphic prin­ci­ple, quan­tum an­neal­ing uti­lizes a net­work of en­tan­gled qubits, where en­tan­gle­ment steadily in­creases in ac­cor­dance with our ob­ser­va­tions of Von Neu­mann en­tropy. There are many other similar­i­ties (and I promise I’ll stop men­tion­ing quan­tum an­neal­ing now), but my point, more gen­er­ally, is that there are rea­sons to be­lieve we can in­deed recre­ate the uni­verse through some form of quan­tum com­put­ing. For sim­plic­ity, I’ll dis­cuss this in terms of a “simu­la­tion,” but I want to em­pha­size that this doesn’t im­ply a simu­la­tion is any less “real” than any­thing else — it’s all quan­tum in­for­ma­tion at the end of the day, and ex­is­tence within a causal loop could just be simu­la­tion in per­pe­tu­ity any­way.

From my van­tage, this could go one of two ways. In each sce­nario, the goal is to cre­ate a match­ing “first” mo­ment within a simu­la­tion; as long as that con­figu­ra­tion of in­for­ma­tion is always the same be­tween simu­la­tions, and a nested simu­la­tion re­mains co­her­ent, then the causal loop is main­tained. Again, an event on the loop is sim­ply a spe­cific ar­range­ment of in­for­ma­tion. Both op­tions re­quire a su­per-ad­vanced civ­i­liza­tion in our dis­tant fu­ture; rel­a­tively speak­ing, they may even seem like gods, but these op­tions don’t re­quire cap­i­tal-G God.

The first way is that we’re able to de­duce some set of pa­ram­e­ters and ini­tial con­di­tions of our uni­verse. If we ex­actly calcu­late its in­for­ma­tion ca­pac­ity (Beck­en­stein bound), uni­ver­sal con­stants, laws, and find a grand unified the­ory, then we can also find some en­tan­gle­ment ge­om­e­try that per­mits all of those prop­er­ties. We’d then cre­ate a quan­tum sys­tem with match­ing pa­ram­e­ters and hope we run it from the same “start­ing” point as our own — that might be some point of min­i­mum en­tropy where the sys­tem couldn’t pos­si­bly be any sim­pler, similar to how we view the sin­gu­lar­ity be­fore the big bang. This as­sumes that some simu­la­tion “be­fore” us chose the same start­ing point as the ob­vi­ous choice since any pos­si­ble timeline can then fol­low as its trends back to­wards equil­ibrium. The enor­mous en­ergy re­quired for such a task might spell the an­nihila­tion of the par­ent simu­la­tion, like a cos­mic self-sac­ri­fice, but maybe that’s the point.

The sec­ond and pos­si­bly more in­trigu­ing way is the “mes­sage in a bot­tle” ap­proach. Imag­ine that when a simu­lated uni­verse is pro­grammed, in­struc­tions are left for the in­hab­itants of that simu­la­tion to then recre­ate the same simu­la­tion. This makes sense if in­tel­li­gent life has a vested in­ter­est in main­tain­ing the causal loop it oc­cu­pies. They would leave in­struc­tions in some­thing ubiquitous and un­chang­ing like the uni­ver­sal con­stants, the cos­mic microwave back­ground, or in our DNA. In fact, all hu­man DNA differs by less than 1%, and about 98% of our DNA is con­sid­ered to be non-cod­ing, or “junk” DNA — it’s an in­ge­nious place to pass along cru­cial in­for­ma­tion. And DNA is sim­ply a pat­tern of in­for­ma­tion that can be eas­ily pro­grammed; mean­ing DNA would be en­coded into the ini­tial con­di­tions, so the uni­verse emerges around DNA-based life, differ­ing from the “ab­solute sim­plic­ity” ini­tial con­di­tions of the first op­tion. Though, we’d still re­quire a uni­ver­sal lan­guage that can be un­der­stood by any in­tel­li­gent life to de­code the in­struc­tions; maybe it all re­ally is in the maths and op­tions 1 and 2 are more al­ike than we think.

It’s also worth not­ing that Han­ley speci­fi­cally cites the use of ge­net­ics in an ex­am­ple of a “per­son loop,” where, “Given the nor­mal re­cy­cling of cells, it may be that a per­son’s body has en­tirely re­place­able parts.” Yet ge­netic code (ideally) re­mains un­changed, so that in­for­ma­tion could re­main con­sis­tent in a loop. In fact, if DNA is the fo­cal point of a causal loop, then it seems the only in­for­ma­tion that needs to be simu­lated is that which con­sti­tutes the ex­pe­rience and col­lec­tive mem­ory of DNA-based life. If in­for­ma­tion changes out­side of that, who would ever no­tice? The in­for­ma­tion re­quire­ment for this simu­la­tion be­comes much more man­age­able since we don’t need to ren­der ev­ery prop­erty of ev­ery par­ti­cle through­out the ob­serv­able uni­verse.

Of course, this is all wild spec­u­la­tion, but it does make for a fun ex­er­cise. Maybe there are al­ter­nate routes that will be­come ob­vi­ous as we learn more about re­al­ity. I’m just try­ing to get the ball rol­ling in case we do live within a causal loop. It’s my loop too, af­ter all.

Find­ing God

When ex­plor­ing the idea of a causal-loop-only uni­verse, it’s al­most im­pos­si­ble to ig­nore some of the im­pli­ca­tions for life within that uni­verse.

For one, it ap­pears to make in­tel­li­gent life nec­es­sary for any­thing to ex­ist — at least in any uni­verse that’s not a CTC. From this view, life isn’t rare: it’s re­quired. If no in­tel­li­gence emerges, there is no fea­si­ble way for a causal loop to re­main in­for­ma­tion­ally con­sis­tent. This also means that any life-car­ry­ing uni­verse must fol­low a se­ries of causes and effects that en­able a min­i­mum de­gree of in­tel­li­gence and agency — life must gain the abil­ity to ma­nipu­late the in­for­ma­tion of the uni­verse it­self. So not only is in­tel­li­gence re­quired, but highly ad­vanced in­tel­li­gence is re­quired.

What does this all look like from the per­spec­tive of life within such a uni­verse? Well, look at our own — the en­tire uni­verse is “cool­ing down” to­wards di­s­or­der, but in­tel­li­gent life and what it touches are the only things that trend to­wards more or­der. Over time, our knowl­edge and tech­nolog­i­cal ca­pa­bil­ities in­crease. What’s the up­per limit to this trend? Is it a co­in­ci­dence that we’ve come to a point where we can start ex­plor­ing and con­trol­ling quan­tum in­for­ma­tion, the very fabric of re­al­ity? How much more will we achieve in the next cen­tury, mil­len­nium, or ten mil­len­nia?

Maybe we re­ally have just got­ten lucky, but in a causal loop this tra­jec­tory is not a co­in­ci­dence — it’s a cer­tainty. Life doesn’t just veer off the rails into oblivion; it’s locked on a path, or lots of equiv­a­lent paths that are all des­tined to tell the same story — the same uni­ver­sal archetype. The loop can­not be bro­ken, else it would have never ex­isted. Life is bound to per­sist, bound to over­come, bound to ex­ist again — isn’t this the kind of hope peo­ple nor­mally place in God?

I’m not say­ing God liter­ally ex­ists. Maybe an om­ni­scient be­ing ex­ists as the high­est ex­pres­sion of in­tel­li­gence on a loop right be­fore it must re­set, but that seems like a dis­trac­tion to a more mean­ingful point: ex­ist­ing in a causal loop — at any point — is prac­ti­cally like liv­ing in a uni­verse where God ex­ists too.

Isn’t that the case if nearly ev­ery­thing about ex­is­tence takes the shape of a se­ries of un­end­ing co­in­ci­dences? Other­wise, the odds of life aris­ing in our uni­verse are as­tro­nom­i­cally un­fa­vor­able, as is the fact that life has evaded ex­tinc­tion for a few billion years to be­come what it is to­day. If you rec­og­nize co­in­ci­dence af­ter co­in­ci­dence, it’s not much of a leap for a ra­tio­nal mind to think that a higher power or­dains each mo­ment, fol­low­ing some grand de­sign. Many of us have stepped away from that wor­ld­view, but maybe we just had an in­com­plete per­spec­tive. Maybe we have rea­son to be­lieve again. As we step closer to truth, we might see that our old silhou­ette of God was sim­ply the nega­tive space of an equally hope­ful struc­ture of re­al­ity.