Epistemic status: I’ve been mulling this over for a few months but have done no in-depth research on the topic. I think it’s plausibly true that the case I make below is solid but it’s outlandish enough that, on priors, I discount it significantly. It’s also outlandish enough that I’ve chosen to write this anonymously (at least for now).
Explanations of UAPs
The Pentagon has now endorsed the idea that UAPs (Unidentified Aerial Phenomena) exist (see this report by the Office of the Director of National Intelligence). Several prominent individuals have expressed support for the idea that UAP are real objects and cannot be explained by current technology. All are uncertain as to what they are. Below are some potential explanations, most supported by the DNI report.
- Ultra-advanced Chinese or Russian technology: Russia or China have advanced systems (seemingly much more than 10 years ahead of the US) and they have had them for more than 10 years (Pentagon evidence of UFOs goes back to 2004). (DNI Report)
- This contrasts with the fact that, to date, all military systems deployed by Russia or China we know of have lagged behind the US and they have not leveraged this ultra-advanced technology at all.
- Ultra-advanced US technology: The US military could have secret projects which are ultra-advanced but are so secret that only very few are allowed to know. UAP reports are an attempt to dissuade adversaries from recognizing they have this technology. (DNI Report)
- I would also be surprising that the US has not yet leveraged such technology and that it has been kept secret for so long.
- Airborne clutter: “These objects include birds, balloons, recreational unmanned aerial vehicles (UAV), or airborne debris like plastic bags that muddle a scene and affect an operator’s ability to identify true targets, such as enemy aircraft.” (DNI Report)
- Natural Atmospheric Phenomena: “Natural atmospheric phenomena includes ice crystals, moisture, and thermal fluctuations that may register on some infrared and radar systems.“ (DNI Report)
- Spoofing by adversaries: China or Russia could have somehow hacked into the systems or otherwise tampered with their inputs so as to make sensors detect movement which was not actually happening. This could explain some of the more unrealistic maneuvers which UAPs have been documented as performing. (DNI Report)
- This level of spoofing would be extremely advanced and is of unclear purpose.
- Extraterrestrial craft: There is some alien presence on earth which possesses advanced flight capabilities.
- Notably not endorsed by the DNI report but widely speculated on by people like John Podesta, former CIA Director John Brennan, and other fairly credible people.
- Edit: A friend mentioned that this could be a PSYOP by an intelligence agency.
- This seems plausible but very unlikely to me given this would constitute an incredibly complicated project and there is no clear reason why any intelligence agency would go to the trouble of executing it.
The report explicitly states that only one of the 144 sightings could be conclusively attributed to any of the categories listed above (excluding extraterrestrial crafts and PSYOP) and argues that we need more analytic resources or scientific breakthroughs to understand what is happening. While the report does not name extraterrestrial crafts as a possibility it seems there is a realistic possibility that some sightings are indeed extraterrestrial crafts.
Indeed, a 5% chance of these being extraterrestrial crafts seems to be a very conservative estimate with another 5% allocated to advanced earth technology also seeming a reasonable lower bound. Either of these possibilities could be game-changers for any number of questions longtermists care about.
Why we should care about UAPs
While the (probably) most likely explanations of airborne clutter, instrument failure, or natural atmospheric phenomena would have no effect on longtermist thinking, any case in which ultra-advanced flying craft exist should rock our world models significantly.
- Implications for the great filter hypothesis: The great filter hypothesis theorizes that the reason we don’t see advanced life in the universe is because there is some point at which the probability of survival drops to near-zero (or maybe zero). This point could be behind or in front of us. If other advanced alien life exists this at least partially invalidates the great filter hypothesis. This should give us much more confidence that humanity will survive.
- Potential for transformative technology acquisition: If any nation-state shoots down a craft then the technology it could acquire could lead to rapid and discontinuous breakthroughs in any number of areas.
- Potential for contact: Contact with an advanced civilization could be incredibly useful for humans or have any number of important consequences.
- Potential threat: Alien civilizations might be looking for habitable planets to colonize. Earth is a habitable planet and is therefore at risk.
- Importance of non-universe-destroying existential risk: If we’re sure there are aliens in the universe then the extinction of the human race is less important as we’re not the only sentient life. So long as whatever kills us doesn’t kill all life in the universe (as a superintelligent AI plausibly could) then some longtermist arguments matter less.
Ultra-advanced earth-originating technology
- World Hegemony: If one state possesses ultra-advanced technology then dominance by one world power is much more likely in the near future. Given this hasn’t yet happened, perhaps the technology demonstrated by UAPs is not as militarily useful as it seems or still requires significant refinement to be useful.
- There are more secrets than we thought and actors are capable of keeping even extremely important developments secret.
- Other things?
A theory: Maybe they’re von Neumann probes
The most likely extraterrestrial explanation of UAPs in my opinion, and which I’ve never seen mentioned, is that we are seeing von Neumann probes (aka self-replicating probes) sent by distant civilizations. If von Neumann probes are both feasible and useful for advanced civilizations/Superintelligent AIs (which seems likely) and advanced civilizations/superintelligent AIs are out there (also seems likely) then there should be massive amounts of probes in the universe. If each civilization launches just a few which then replicate each time they find viable materials then the result should be that, after many thousands of years, there should be billions or trillions of these probes. The galaxy should be littered with von Neumann probes!  Indeed, from this perspective perhaps we should have a strong prior that von Neumann probes have visited Earth.
It would then make sense that, if they were to find a habitable planet, the probes would stick around and monitor things while reporting back to the home planet. Under this theory the probes might avoid contact with humanity while they wait for instructions from the home planet (which could take thousands of years given the distances involved). On the other hand, it's hard to say why a Superintelligent AI wouldn't delegate some actions to a self-replicating probe (unless maybe the alignment problem is impossible to solve?) so that counts as some evidence against the hypothesis.
It might also make sense for the probes to behave strangely. After many thousands (or millions) of replications the probes might degrade from their original state and start to malfunction in various ways. The UAPs we see today could be the entropy-riddled craft of very distant civilizations.
What could we do about all this?
It’s not at all clear and I haven’t thought much about it. Perhaps more research would yield some insight into what is happening. Perhaps we could create better estimates of the likelihood of different explanations. For instance, we could look into how long the US government managed to keep stealth technology secret as a case study. Perhaps efforts to shoot down a UAP would be worthwhile in order to advance technology. Perhaps we should try to show UAP how cooperative we are and think of humanity’s behavior as our application for joining the galactic UN.
Another interesting line of reasoning is to speculate as to why, if there are probes from multiple or even many alien civilizations, none choose to contact, destroy, or otherwise engage with us. Does this imply there is some theoretically dominant strategy for von Neumann probe usage? If so, what might this strategy be? Does it perhaps imply something about the difficulty of the alignment problem that no alien civilization chooses to delegate actions (e.g. contacting or destroying us) to an intelligent AI in their probes? Indeed if there are many existing advanced alien civilizations what does that imply about superintelligent AI? Could we know whether these civilizations are run by superintelligent AI?
Obviously all of this is highly speculative, but the evidence and the current state of the discourse still indicates that maybe not enough people are working on this with the appropriate rigor. Mull it over, share your thoughts if you feel like doing so publicly, or DM me to do so privately.
The new term for UFO ↩︎
David Norquist, the Deputy Secretary of Defense; former CIA Director John Brennan; President Clinton’s former chief of staff and Obama advisor John Podesta; Senator Harry Reid. See this New Yorker piece for information and Tyler Cowen’s comment on his Brennan interview. ↩︎
One plausible reason for alien craft on our planet could be to stop a superintelligent AI from being created. One could imagine something like a prime directive to not interfere with developing civilizations but with the caveat that they should not be allowed to create something potentially universe destroying. ↩︎
I have thought very little about the implications of this scenario but surely there are other important implications. ↩︎
This seems much more likely than manned craft being on planet earth because manned craft cannot be replicated so easily. ↩︎
This logic echoes Bostrom’s Simulation Argument in which he argues that if you buy the premises that building an ancestor simulation is possible and potentially useful to future civilizations and that humans are not almost certainly going to go extinct then it’s more likely than not that we live in a simulation. In this case, if you believe advanced species exist and that von Neumann probes would be useful to them it is more likely than not that we have von Neumann probes on Earth. ↩︎
This seems really important! If you have thoughts on whether this seems true please share. ↩︎
It does seem like they could have some kind of pre-recorded message or send a simple AI but I could imagine any number of reasons why they might not want to do this (e.g. might not want to help some civilizations, keep the element of surprise, prime directive, etc). Another question is why they don’t just kill us and leave the planet undefended for future colonization if that’s their objective. I don’t know why this doesn’t happen but perhaps while we’re at a non-threatening stage and they’re simply curious? ↩︎
E.g. Maybe some alien civilization just got the news that we’re starting to learn how to farm and have sent us a message already and soon we’ll get a message written in Sumerian. ↩︎
Instead of waiting and hiding, they could immediately begin converting the solar system into useful structures such as more probes, Dyson swarms, habitats, etc. Why would they need to await further instructions? With advanced technology they can have entire societies of superintelligent minds living on the probes, plus the home territory that sent the probes will presumably have anticipated various possibilities (and literally experienced many of them in the course of expansion) and based on those simulations + experiences given its probes instructions for how to handle the top billion or so most likely scenarios.
Preventing degradations due to entropy is a pretty solvable problem, I think. Just crank the Redundancy Dial and let the law of large numbers do the rest. Am I being naive here? I'd like to hear more on this. Sure, extra redundancy might slow things down a bit, but probably not much?
I think I agree (somewhat in contrast to what I said in the post) that I would expect vNM probes to take major action on finding a habitable planet. The fact that they are not taking action seems like the most damning critique of the theory.
It does seem like there are plenty of reasons they wouldn't do something noticeable though. E.g. curiosity about a non-threatening species, ethics, etc. Granted this seems extra unlikely if there are vNM probes from many different civilizations present.
As far as degradations I have no idea... Haven't though much about it. Seems like redundancy would be very helpful and that it's a maybe solvable problem but could also imagine it being hard to solve given e.g. lot's of radiation, many generations, etc.
And there would be vNM probes from many different civilizations present, unless somehow they could coordinate to block others from their territory... all while remaining in hiding...
Relatedly, even if they do await further instructions, why haven't we been terraformed yet? If awaiting further instructions takes only a thousand years, this theory still predicts with high confidence that either they haven't arrived yet or they have already terraformed us (and thus, updating on our observed lack of terraforming, they probably haven't arrived yet.) If it takes a billion years to get further instructions, then OK, but that's pretty darn implausible for reasons mentioned above.
Also they'd have to await further instructions not just in this star system but in the entire light-cone; they can't convert Alpha Centauri into a Dyson swarm either because then Earthly astonomers would notice.
I generally agree. Though for what it's worth it seems like it wouldn't be hard to block others from the territory while remaining in hiding. Presumably any blocking could take place by intercepting ships well before they reach earth. We would have a very hard time noticing something like that.
Apparently there's something like 7-8 million stars within 1000 light years so yeah, seems like there could have been some instructions sent. That said maybe life is super rare and only exists in 1 in a billion stars or 1 in a trillion. Then maybe any vNM probe would be coming from 10s of thousands of light years away or something.
Another possibility is that something about technological maturity tends to make species less interested in massive and aggressive expansion and/or species capable of making vNM probes tend to die off or become totally inward facing and/or there's some intergalactic governance system preventing certain actions by vNM probes.
Overall I agree with it lowering probability for vNM probes on earth but I don't think this reasoning invalidates the theory
The Future Strategist podcast episode "Cochran on UFOs" brings up mass hysteria as a possible explanation, but without discussing it in much detail.
Yup that seems pretty plausible! I think I'd put that as a part of the Airborne clutter/natural atmospheric phenomena clusters. Though I guess mass hysteria adds the important addition of better explaining non equipment-mediated human sightings.
Thanks for the interesting post!
One disagreement I have:
"Indeed, from this perspective perhaps we should have a strong prior that von Neumann probes have visited Earth."
I don't think this should be strong prior.
The probability of von Neumann probes visiting earth seems heavily constrained by the age of the universe, the time taken for intelligent / superintelligent life to emerge elsewhere, the distance of this life from us, the accelerating expansion of the universe, the speed of replication of von Neumann probes (including the abundance of viable resources) and the speed of travel of von Neumann probes.
I think I agree with you actually. I think what I wanted to get across was that we shouldn't have a strong prior against there being extraterrestrial craft on earth. To me the term strong prior feels pretty ambiguous so I just kind of threw it in there without thinking.
My intuition is that right now almost everyone has a strong prior against extraterrestrial craft being on earth in that they would assign a prior probability of like 0.01% or something. I think maybe our prior should be more like 5% that they're on earth. This would dramatically change what probabilities look like for the extraterrestrial craft hypothesis after updating on new evidence (e.g. Pentagon report stating that they think UAPs are actual objects behaving in ways unexplainable by modern science).
Actually, I think these estimates are extremely high. I don't think this post engages enough with mundane explanations (as in this Salon post)).
Agree these estimates are high, but disagree with Salon. While there are plenty of mundane potential explanations, I think suggesting that the 144 minus 1 sightings investigated in the Pentagon report are in fact explained is misleading. My starting assumption is that the Pentagon would have been able to diagnose something like a FLIR glare filter.
I get that there's a sensationalism to the UFO angle, but I suspect in this community, we might be more susceptible to to letting cultural taboos around "seeing things in the sky" lead us to a really unscientific skepticism on this topic, which was definitely the status quo prior to the release of the report.
My only other point is that the mundane explanations are also super important to explore! Many of them have major national security implications.
Have you seen the 2020 documentary The Phenomenon? Compelling film on this topic. Clearly there are a lot of sightings concentrated around nuclear facilities and bodies of water, which is disturbing under any number of explanations.
I thought about this some more and thought maybe investigating UFOs could be important in that it is part of the larger goal of the search for extraterrestrial intelligence. The search for extraterrestrial intelligence could hold at least several opportunities/implications for us.
They could provide us with knowledge and technology that gives us the push past the point where survival is extremely improbable. Or, alternatively, maybe we would have found the knowledge and built the technology eventually without their help. If this were true, then obtaining the knowledge and technology through them would bring improved living conditions for billions of humans sooner than we would have brought it on ourselves and thus we have more time to possibly come up with more breakthroughs that would enable us to live longer.
We could partner with them. We could perhaps form some kind of trade agreement. Or perhaps they would be willing to help us for altruistic reasons. Maybe they have asteroid deflection technology, climate control technologies, solar flare protection technology, and other technologies that they would use to help us. Even if they didn't have much interest in partnering with us, if they are visiting us, depending on their intentions, we could make their stay more worthwhile which might warrant something in return from them.
Maybe, as suggested by Robin Hanson in James Miller's podcast (https://soundcloud.com/user-519115521), we are here because of panspermia.Then it is possible that aliens developed from the same seed we started as but on a different planet. In that case, however different from us they ended up because of their different environment and upbringing, we would need to realize they are essentially the same as us. Maybe we share the same common seed with many alien species in the universe. Maybe some would be much younger than our species and some would be much older. I suppose some could be only a few hundred years younger than us. It could probably be ruled out that there are any that developed radio around the same time as us or we would have found each other. Maybe some species are similar to us, some are very different, and some are radically different. We might be appalled by some of the species' customs. If they all come from the same seed as we do, finding all these aliens would involve a coming to terms with the fact that, no matter how shocked we are by them, we all have a common ancestor/seed.
As suggested by Robin Hanson, they might be concerned that we would be appalled by their customs and they would be appalled by our customs. For that reason they would choose not to know anything about us and not let us know anything about them. Conflict might erupt because of one side being offended by the other side.
Aliens might have the technology of time travel or otherwise have some kind of prescience and are observing us to ensure some fate doesn't happen to us. Maybe their extremely long existence or extremely sharp prescience has taught them that certain technologies lead to inevitable doom. It is possible that if we learned what they knew about us we would get our hands on a forbidden fruit and spell doom for ourselves and/or the entire universe. In that case, trying to learn more about their intentions for visiting us could have negative consequences.
It seems like it might be worthwhile investigating UFOs/UAFs for a large umbrella purpose of ensuring all technology, information, knowledge about the universe, etc. is democratized and accessible to everyone and not monopolized for nefarious purposes.
It might also be worthwhile to study them to safeguard ourselves from government's psychological operations. It seems that the sky has the potential to have a huge influence on a huge number of people.
Given that people can conflate a spacefaring extraterrestrial craft with a plastic bag in the sky, studying UFOs/UAFs could benefit us by reducing our capacity to miss the astronomical significance of objects right before our eyes. This benefit could be similar to the benefit that is gained from learning how to spot disinformation and misinformation.
The Great Filter Hypothesis
Regarding the great filter hypothesis, maybe I'm wrong, but wouldn't the discovery of just one extraterrestrial civilization with universe colonization technology increase the probability of a species surviving past a certain point only by a single speck? The discovery would tell us there was at least one species in the entire universe that survived long enough to develop such advanced technology.
If it is incredibly unlikely for us to survive for much longer, communication with them might lead to them sharing their technology and thus providing us with the means to survive longer than we otherwise would have.
It is also possible that the species survived in a region of the universe incredibly far away that had an environment (maybe less risk of asteroid impacts and other astronomical events, etc.) with greater odds for longterm survival than our region of the universe. If their survival was due in large part to various characteristics of their region of the universe, then obtaining this knowledge through communication with them would be important to us.
If there were an extraterrestrial civilization with such advanced technology, it would be useful to communicate with them to discover whether there are more civilizations like them and then update our estimate of the probability of longterm survival in the universe. If there were a significant number of other civilizations like them, that might end up revealing that surviving long enough to develop such advanced technology is common in the universe and thus tell us that our moment in history is not that special.
A.I. Series by Vaughn Heppner
Regarding what you said about our own future not being as important when we take into account all sentient life in the universe, it reminded me of the A.I. series by Vaughn Heppner that I was listening to on Audible a few months ago. I still need to read/listen to the last book of the series. In it, several species across the universe band together to fight against an A.I. civilization that aims to eradicate all biological life in the universe. Several of the species find it easy to bond with each other. One individual is the last of their own species and becomes afflicted with loneliness and depression. However, they are able to make friendships with other humans.
I was reminded of the A.I. series again by your discussion of the probes. In the series, the A.I. civilization's domination of the universe was not perfectly coordinated. The A.I. civilization sent one huge ship to destroy a species in a region of the universe. If the ship was defeated, then another three would be sent, then nine, and so on. The number of ships sent after each unsuccessful attempt would be three times the number of ships sent the last time. For all the intelligence they had, the limitation on how fast information could travel across the universe seemed to dampen how effective their domination of the universe could be.
I sometimes wonder whether extraterrestrial civilizations send probes into the universe like we do. Even if a civilization is advanced enough to send members of their own species to travel extremely long distances, maybe the detrimental health effects or the time it takes for such travel makes sending probes more economical. Or, perhaps there are no such drawbacks for them and they do travel to a few incredibly distant regions of the universe then send probes to many other incredibly distant regions to maximize the number of regions of the universe that they explore (maybe they don't have enough members of their species who are space explorers to explore all the regions of space they want to explore).