When the best answer is that we don’t know

To Aubrey de Grey who dared to put a number on our uncertainty concerning the prospect of comprehensive biological rejuvenation

In the early 90s as a high school student interested in the natural sciences, I went to listen to a talk by Ede Teller, the controversial ‘father of the hydrogen bomb’. Much to my surprise, Teller, in his early 80s  & sitting on the top of a table at the ELTE University in Budapest, answered one question the following way: ‘Erre tudom a pontos választ. Nem tudom.’ which translates as: ‘I know the exact answer to this question. I don’t know.’ This honest bon mot captures an epistemological puzzle: sometimes acknowledging well informed uncertainty is the adequate form of a valid answer. How come?

One problem with seasoned experts in science and technology is that exactly what makes them experts in the first place is what limits them acknowledging when sometimes the exact answer just cannot be provided. But when the scientific and technological question has a potentially long running impact on human society and Planet Earth, acknowledging the lack of a definite scientific position becomes even more challenging. In these cases, not just individual professional credit is at stake but pre-scientific moral integrity and post-scientific political responsibility. 

An emerging prime example for such a problem is scientists, technologists commenting on the possible limits of what we can achieve in terms of human longevity with ever more advanced biomedical technology. How far can we extend healthy life expectancy? Can we possibly break the maximum human lifespan barrier? If yes, when are we going to hit the next roadblock, if ever? Please note that these questions are usually take the form of what philosophers call modal questions, asking about the practical possibility, the feasibility of some science intensive technological scenario.

What are the components of this particular challenge in terms of science?

On one hand, we have now an almost comprehensive list of the major hallmarks of the aging process (or perhaps use the plural: ‘agings’) thanks to breakthrough research accumulated in the last decades[1]. These insights are mostly based on studies in model organisms, where every such hallmark process was manifested during normal aging, its experimental aggravation accelerated, while its experimental amelioration delayed some agings phenotype measure and increased healthy lifespan. For C.elegans, the nematode worm, there’s 10-fold increase in lifespan. For Drosophila, it’s 2-fold. For a mouse it’s only 20-30% increase in median lifespan, not in terms of maximum longevity [2]. As for humans, the evaluation of this technological possibility is largely informed by growing demographic data on the malleability of the default aging process happening inside us in the wild. This so-called shifting mortality scenario informs us that ‘old-age survival follows an advancing front, like a travelling wave’[3]. Resonating personally, old people always seemed as pioneers to me, operating at the expanding wavefront of the living.

The current pace of biogerontological and related relevant research, be it regenerative medicine, genetic studies or senotherapetuic trials aiming to minimise cellular senescence, is breathtaking, every other week we have a major paper published, and a sensationalist title propagated into mainstream media and given its due hype. 

On the other hand, the scientific details of these processes are just being hashed out and it is not even clear that the list is exhaustive or there’s a chance for a so far unknown hallmark process to emerge later. Another important point is the lack of a quantitative model or simulation able to predict the aggregated human lifespan effect of different sequences of consecutive and hypothetical interventions counteracting and even reset the ongoing hallmark processes of biological aging. No wonder: computationally handling ~40 trillion human cells, comprising the average human body (plus our holobiontic permanent guests, the human microbiome) and their continuous mitotic trajectories; well that scientific paradigm is just not here yet [although I’ve just recently tried to make a scientific case concerning those mitotic trajectories, please see. Hoping for a single miracle drug is naive and indeed unlikely, but now data has started to pour in from polypill approaches and different versions of synergy, the beneficent ways of combination treatments on lifespan are being uncovered [4].

In terms of the positions taken concerning these complex and challenging longevity questions, there are different camps amongst scientists, split according to overemphasising one hand or the other. Much have to do with the particular branch of science those specialised experts are familiar with and something have to do with the affiliation of those scientists making these claims. For instance demographers, biostatisticians looking into centenarian, supercentenarian survival data might be in favour of further malleability but classical geneticists, genomicists might see a strong limit on human lifespan looking into the same data [5]. This also hints at another potential issue: while scientists have the aura of expertness around them, the public, even the journalists can easily miss the fact that Scientist Z is  overstepping their disciplinary limits, and communicates a modal opinion confidently to the public, using interpretation of data not of their particular subfield of specialisation. Also don’t expect that scientists working in the nascent longevity industry, for instance on the clearance of senescent cells, will acknowledge the full potential of longevity science, their job is hard, their bosses expect them to oversell the immediate healthspan benefits of the interventions of their own choosing, at the same time they feel the urge to seriously alleviate the fears of the public concerning the long term potential of their endeavour. To use a mathematical analogy, for them to prove the upper limits of this technology combined, it is like to show the convergence of a series, to give an algorithm to show that there is a serious upper bound.

In science many many advances seemed highly unlikely, practically impossible earlier, just to emerge as normal science and technology as usual at some point, think of  cancer immunotherapy or AlphaGo recently.

I think in the light of these factors, the exact answer currently to the question of how malleable is human longevity if one factors in advanced biomedical technology is that we just genuinely don’t know. But we might be able and should  try to quantify how much we don’t know. On a methodological level this kind of honesty is not unlike to Bayesian methods focusing on quantifying uncertainty, while the other camp has run into serious issues by trying to assign an arbitrary, binary threshold for scientific truth via p-values. This issue is now being openly questioned as a legitimate and default scientific practice [6]. 

But when it comes to human longevity, we are both deeply affected individually and politically, as can be inferred from our expressed opinions, and this is an extra sharp double-edged sword.

Individually we are facing our deepest, darkest fears, in terms of temporality and our fate. When I talk with people (a lot) about the potential of living much longer lives aided by science, I see that their first reaction of potentially living up to say 150 year old is practically equating that with ‘immortality’. Here lies the biggest binary trap of how we usually think on the options here, and I call this the Immortality Trap. It is a trap as it forces thinking into acknowledging only two options, the closed cap mortal and the impossible immortal. Even some supporters and the majority of opposers of human longevity attempts are using this polarising and completely useless rhetoric. As a result the framing of the potential in mainstream media is full with titles of living ‘forever’, ‘eternal’, ‘perpetually’, infinitely’. The way out conceptually is to acknowledge a possible third scenario which I dubbed as Open Lifespan. ‘Open’ in this term is short for open-ended, indefinitely long lifespan. It is definitely mortal though but it might not be essentially bounded. To use the toolkit of philosophers, whose expertise lies in conceptual analysis, Open Lifespan is a possible world that is accessible from this actual world as its upper limit. Equally importantly, indefinite means uncertain as well, so there’s our honest answer. As opposed to this, biomedical immortality is an utterly impossible and unimaginable/unimaginative concept and we should really get rid of using this term in the public related to humans, contra and pro.

This last remark leads to political worries and confusions based on the sharply different and categorical looking ‘scientific’ opinions formed in terms of the potential of human longevity. If the position is that only couple of decades of healthspan extension is feasible then it’s hard to make the claim of regulating out a longer stretch reached by increasing use of biomedical technology. The bread and butter of the power of healthy longevity science will be randomised clinical trials eventually, and the case for that is just under democratic and open discussion right now [7]. 

One big question is whether longer sustainability for individual human lives will trigger better sustainability practices, more pre-emptive responsibility for the planet, a scenario I call ecolongevity. Premature, hardliner scientific positions on human longevity will limit the scope of honest discussion and in depth investigation of possible political scenarios.

In political philosophy it was suggested, that texts of certain pre-modern thinkers, belonging to an oppressed religious or political minority say, can only be interpreted with separating their esoteric, secret, concealed content by reading between the lines and their exoteric, polished content complying with the majority and frequently expressing opposing views compared to the esoteric reading. In science there’s a long tradition of big scientists working outwards and inwards, perhaps the most high profile case being of Isaac Newton [8], but also his contemporary competitor, G.W. Leibniz [9], whose work on gravity went to centre stage but his work on transmutation of metals landed in the attic in a locked up box.  

When it comes to the potential limits of science intensive human longevity we must collapse and exclude this double accounting in our communications towards non-scientists without the fear of loosing scientific credits. An exoteric answer is one that is telling you that all we are aiming and can possibly achieve is 2 decades of healthy lifespan extension, so we can stop at 93 and die healthy. This seems only a little more established answer scientifically, than the other, exoteric, publicly promoted and practically impossible answer of completely defying death with technology. This latter drive is so deeply rooted in human history and culture, think of Gilgamesh’s quest for immortality or the same motivation behind the tomb treasures and terra-cotta army of China’s first emperor, Qin Shi Huang Di, that it conceptually conflates different technological scenarios for many.

I think playing this double game is politically dangerous and adjustments are needed to counteract both extremities, the one that constantly underplays and the other that exaggerates the opportunity ad absurdum [10].

The honest, collapsed, plain answer is that we genuinely don’t know but we are honestly working on to figure out just how much we don’t realise the limits here. Let’s not conceal this cognitive fact behind the veil of expertness.

When it comes to rigorous conceptual analysis and innovation, sometimes excellent scientists turn out to be mediocre philosophers not being able to get the job done. Science needs the toolset of humanities if it wants to inform policy and yield compromises without being compromised itself.

The question of the possible limits of human longevity is cracked wide open, thanks to science, and the genie is out of bottle. How far it’s gonna go, we genuinely don’t know. Can we put it back in the bottle? No. Can we estimate how far it’s spread? Certainly. Can we be sure about it? Uncertainly.  


[1] López-Otín C, Blasco MA, Partridge L, Serrano M, Kroemer G. The hallmarks of aging. Cell. 2013 Jun 6;153(6):1194-217. PMID: 23746838

[2] This interview below with Judith Campisi was the initial trigger behind writing this piece and am using her quick recap on lifespan fold difference results

[3] Zuo W, Jiang S, Guo Z, Feldman MW, Tuljapurkar S. Advancing front of old-age human survival. Proc Natl Acad Sci U S A. 2018 Oct 30;115(44):11209-11214. PMID: 30327342

[4] Admasu TD, Chaithanya Batchu K, Barardo D, Ng LF, Lam VYM, Xiao L, Cazenave-Gassiot A, Wenk MR, Tolwinski NS, Gruber J. Drug Synergy Slows Aging and Improves Healthspan through IGF and SREBP Lipid Signaling. Dev Cell. 2018 Oct 8;47(1):67-79.e5. PMID: 30269951.

[5] This pun is intended to the methodologically super-controversial Evidence for a limit to human lifespan in Nature from 2016, none of the authors are hardcore demographers, statisticians. 

[6] Here am referring to the relatively recent call to ditch p-values in Nature.

[7] I have in mind the super careful and lengthy framing of the TAME metformin trial by Barzilai et all, here’s a good insider piece on that by Lisa Harris.

[8] A good piece of what historians of science currently think about Newton’s transmutation work: briefly it was integrative part of his scientific practice: Dmitri Levitin.

[9] John Whipple: Leibniz and the Art of Exoteric Writing DECEMBER 2015, VOLUME 15, NO. 35, PP. 1-24f.

[10] I think of the transhumanist immortalists who have serious philosopher backings in Oxford for instance Nick Bostrom and Anders Sandberg at Future of Humanity Institute.