Living a long time forces decisions that could otherwise be ignored. This is true of individuals as well as societies, but let’s think in terms of the individual human being. Getting older creates survival scenarios as simple as ensuring safety and nutrition for the elderly. But let’s extend lifetimes to centuries and beyond. In this thought experiment, we create a society of people so long-lived that their personal planning takes in events like a possible asteroid strike in 200 years. A person who could live for a billion years has to think in terms of surviving a dying Sun engulfing his or her planet.

If we assume a kind of immortality, the individual and the society merge in terms of their key concerns. It’s hard to imagine biological beings living for lifetimes like these, but as we’ve often considered in these pages, non-biological machine intelligence, constantly upgrading and improving itself, should be able to pull it off. Because we know of no extraterrestrial civilizations, we can only speculate, but the speculation is a good way to override our anthropomorphic prejudices. And it’s safe to say that any being or society will try to preserve itself in the event of local catastrophe.

Thus the needed insurance of interstellar migration, which philosopher Clément Vidal (Center Leo Apostel, Vrije Universiteit Brussel, Belgium) thinks might involve journeys vast enough to reach places where stars are as plentiful as in the globular clusters surrounding the Milky Way’s hub. Or galactic center, where resources abound and the need for frequent migrations is thus eased. If a culture like this spreads life through its own variant of panspermia, so much the better. I leave the motivations of a machine culture spreading biological life to science fiction authors, but believe me, there’s a cool plot in there somewhere. Someone should run with it.

Vidal’s idea of a stellar engine seizes on an unusual astronomical object, the so-called spider pulsar. We looked briefly at these last time, but let’s dig deeper. We imagine a civilization (Vidal calls them ‘stellivores’) that uses a low-mass star in its home system as a source of fuel, gradually consuming the star’s energies by accretion. So far so good, as we know that accretion is a well established phenomenon. It occurs, for example, in the formation of a Type 1a supernova, where a white dwarf reaches the Chandrasekhar limit (1.4 solar masses or so), drawing in material from a companion red giant through the process. Soon we have runaway nuclear fusion and a bright new object for astronomers to study.

A spider pulsar, a one millisecond pulsar with a very low-mass companion star, can interact not just through accretion but in some cases through evaporation. In reading Vidal’s new paper, I’ve been puzzled by this process, which actually can alternate with accretion in some instances. Things get complicated and quite interesting. Vidal explained in an email that evaporation becomes the primary process when a neutron star has a strong wind that actually quenches accretion and causes the move toward evaporation. Adds Vidal:

The astrophysical reason this would happen is after a long accretion journey, the companion star would get lighter and lighter, the orbit would shrink, up to the point where it is exposed to the strong pulsar wind and radiation. Then the dynamics would switch from accretion to evaporation. A subclass of (redback) spider pulsars, transitional millisecond pulsars, have their accretion that starts and stops abruptly. This is a fascinating phenomenology that has been studied intensively.

And indeed it has, as witness, for example, Baglio et al., “Matter ejections behind the highs and lows of the transitional millisecond pulsar PSR J1023+0038” (citation below), where I read:

Transitional millisecond pulsars are an emerging class of sources that link low-mass X-ray binaries to millisecond radio pulsars in binary systems. These pulsars alternate between a radio pulsar state and an active low-luminosity X-ray disc state. During the active state, these sources exhibit two distinct emission modes (high and low) that alternate unpredictably, abruptly, and incessantly. X-ray to optical pulsations are observed only during the high mode. The root cause of this puzzling behaviour remains elusive.

But back to spider pulsar terminology. You probably noticed the reference to ‘redback’ pulsars. Astronomers divide spider pulsars into black widows, where the companion star is in the range of 0.01 to 0.1 stellar masses, and redbacks, where the companion star mass is between 0.1 and 0.7 stellar masses. Again, the phenomenon Vidal homes in on is evaporation, and it is at the core of the concept of using such a system as an engine. Suppose we look at a transition millisecond pulsar – one of those switching between emission modes – and consider it within the possibility that it is an engine.

Asymmetric heating could be our technosignature as a millisecond pulsar adjusts the heat of its companion star by moving between accretion and evaporation modes to perform, for example, steering maneuvers outside the orbital plane. Asymmetric heating in varying accretion and evaporation phases can compensate for increases in orbital separation. In Vidal’s view, the goal of the binary stellar engine might be to capture a new star whose energies can now be used to supplant the depleted companion and supply the needs of the engine’s creators. Thus the civilization travels to a new star. We imagine a billion-year culture in constant journey mode in search of energy, with the entire galaxy in range.

When we find objects in our data that show transitional millisecond pulsars in configurations that are suggestive, how do we distinguish between technological activity and natural phenomena? As with all technosignatures, it’s not an easy call. As Vidal notes in his email: “Now the game is to make predictions starting with (1) natural, astrophysical hypotheses and models, and (2) artificial, intelligent, “spider stellar engine” hypotheses and models, and see which predictions turn out to be correct.”

Image: This is Figure 1 from the paper. Caption: Four steering configurations of the binary stellar engine. Figures (a-c) are top views, face on, while figure (d) is side on. In situation (a) the stellar engine is accelerating or cruising; in situation (b), assuming the system velocity is towards the top, the thrust creates a force towards the left; in situation (c), assuming the system velocity is towards the top, the thrust creates a decelerating force. Situation (d) changes the orbital plane by asymmetric heating of the companion, which creates a lifting force in relation to the orbital plane. Note that the pulsar size and orbital separation are not to scale.

The mechanics of acceleration, steering and deceleration are intricate and fully described in the paper, which also includes information on other types of stellar engines going back to the original Shkadov concept – that chart is fascinating. In addition the paper analyzes candidate stellar engines that can be investigated for possible signs of intelligent control. But notice what kind of civilization we may be talking about here. After all, the payload of a spider engine is the millisecond pulsar, the propellant the low-mass companion star. We are thus considering postbiological intelligence on the neutron star itself. I want to quote Vidal’s email on this subject:

…[W]hat really matters is not the hardware of life, but what it does, its software, its functions. Recently we’ve seen with large language models that some form of intelligence can run on computer hardware. Of course, we could change this hardware multiple times and still have the same ChatGPT answering our questions. Life on Earth started with biochemical reactions, now we use semiconductor technology to process data, and we might see optical or quantum hardware taking over in the near future. If a civilization is on a billion-year long track to optimize its hardware and makes the most of the computational capacity of matter in the universe, it would likely continue to improve its hardware using more and more compact, high energy solutions such as nuclear reactions or subnuclear reactions (i.e. neutron star stuff). This is the level I imagine those stellivores are at. So, no planet, no individual traveler. Rather an integrated organism organized around high energy (sub)-nuclear reactions. It might still have some sub-organizations like species, nations, etc. but these would be at extremely small scales and impossible for us to detect.

As with all Vidal’s work, this paper is intricate and deeply researched. About spider engine maneuvering I have had the time only to cover the basics, and encourage interested readers to go to the paper, and also to mine the background laid out in Vidal’s magisterial 2014 title The Beginning and the End: The Meaning of Life in a Cosmological Perspective (Springer). The search for technosignatures demands moving far beyond the assumptions ingrained in our perspective as a species in technological infancy. No one works this turf better and with more elegance than Clément Vidal.

The paper is Vidal, “The Spider Stellar Engine: a Fully Steerable Extraterrestrial Design?” Journal of the British Interplanetary Society Vol. 77 (2024), 156-166 (full text). The Baglio paper is “Matter ejections behind the highs and lows of the transitional millisecond pulsar PSR J1023+0038,” Astronomy & Astrophysics Vol. 677, A30 (September 2023). Abstract. See also Papitto et al., “Transitional Millisecond Pulsars,” in Millisecond Pulsars, edited by Sudip Bhattacharyya, Alessandro Papitto, and Dipankar Bhattacharya, 157–200. Astrophysics and Space Science Library. Cham: Springer International Publishing.