Aza Raskin from the Earth Species Project is trying to translate whale language to English, by modelling whale language using an LLM, and rotating the whale LLM's embedding space to fit an English LLM's embedding space. It sounds very advanced but, as far as I know, they haven't translated anything yet. I'm not sure if they tried orcas in particular. Project CETI has also been working on sperm whales for a while but made no headlines.
That said, it does seem they all tried to understand whale language instead of teaching whales human language like your idea. There is an honest chance you'll succeed when they haven't.
If sperm whales actually are "superintelligent" after getting enough education, the benefits would be a million-fold greater than the costs.
They would be far easier to control/align than ASI because they might have human-like values to begin with, get smarter gradually, be as bribe-able as humans, and work in human-like timescales.
In conclusion, it feels very worthwhile :)
Edit:
Reasons orcas might be smarter:
- Their brains are bigger, as you said.
- They evolved a very long time with a large brain, and have algorithms better adapted to a larger brain. e.g. if you genetically engineered a chimp to have a human sized brain, it probably won't be as smart as humans because it didn't evolve long enough with a large brain.
Reasons orcas might be dumber:
- They evolved a very long time with a large brain. Paradoxically, this reason can make them dumber too. Their brains may "overfit their environment," relying more on a ton of small heuristics fine tuned for their ancestral environment, and relying less on general intelligence.
- The parts of intelligence required for tool use, engineering, and inventing, are more useful for prehistoric humans than orcas.
- Attempts to communicate haven't succeeded. This doesn't prove much because the simplest explanation is that communication is very hard, in fact many human languages can't be decoded. There's a lot of circumstantial evidence they do have language (as you mentioned in your posts).
Thanks. Yep I agree with you, some elaboration:
(This comment assumes you at least read the basic summary of my project (or watched the intro video).)
I know of Earth Species Project (ESP) and CETI (though I only read 2 publications of ESP and none of CETI).
I don't expect them to succeed in something equivalent to decoding orca language to an extent that we could communicate with them almost as richly as they communicate among each other. (Though like, if long-range sperm whales signals are a lot simpler they might be easier to decode.)
From what I've seen, they are mostly trying to throw AI at stuff and hoping somehow they will understand stuff, without having a clear plan how to actually decode it. The AI stuff might look advanced but it's sorta obvious things to try and I think it's unlikely to work very well, though still glad they are trying this.
If you look at orca vocalizations, it looks complex and alien. The patterns we can currently recognize there look very different from what we'd be able to see in an unknown human language. The embedding mapping might be useful if we had to decode a human language, and maybe we still learn some useful stuff from it, but for orca language we don't even know what their analog of words and sentences are and maybe their language works even somewhat differently (though I'd guess if they are smarter than humans there's probably going to be something like words and sentences - but they might be encoded differently in the signals than in human languages).
Though definitely plausible that AI can help significantly with decoding animal languages, but I think it also needs forming deep understanding of some things and I think it's likely too hard for ESP to succeed anytime soon, though like possible a supergenius could do it in a few years, but it would be really impressive.
My approach may fail, especially if orcas aren't at least roughly human-level smart, but it has the advantage that we can show orcas precise context of what some words and sentences mean, whereas we basically have almost no context data on recordings of orca vocalizations, so it's easier for them to see what some signals mean than for humans to infer what orca vocalizations mean. (Even if we had a lot of video datasets with vocalizations (which we don't), it's still a lot less context information about what they are talking about, than if they could show us images to indicate what they would talk about.) Of course humans have more research experience and better tools for decoding signals, but it doesn't look to me like anyone is currently remotely close, and my approach is much quicker to try and might have at least a decent chance. (I mean it nonzero worked with bottlenose dolphins (in terms of grammar better than with great apes), though I'd be a lot more ambitious.)
Of course, the language I create will also be alien for orcas, but I think if they are good enough at abstract pattern recognition they might still be able to learn it.
though I'd guess if they are smarter than humans there's probably going to be something like words and sentences
the highest form of language might be "neuralese", directly sharing your latent pre-verbal cognition. (idk how much intelligence that requires though. actually, i'd guess it more requires a particular structure which is ready to receive it, and not intelligence per se. e.g. the brain already receives neuralese from other parts of the brain. so the real question is how hard it is to evolve neuralese-emitting/-receiving structures.) also, in this framing, human language is a discrete-ized form of neuralese (standardized into words before emitting); maybe orca language would be 'less discrete' (less 'word'-based) or discrete-ized at smaller intervals (more specific 'words').
(warning: armchair evolutionary biology)
Another consideration for orca intelligence; they dodge the fermi paradox by not having arms.
Assume the main driver of genetic selection for intelligence is the social arms-race. As soon as a species gets intelligent enough (see humans) from this arms-race they start using their intelligence for manipulating the environment, and start civilization. But orcas mostly lack the external organs for manipulating the enviroment, so they can keep social-arms-racing-boosting-intelligence way past the point of "criticality".
This should be checkable, IE how long have orcas (or orca-forefathers) been socially-arms-racing? I tried asking claude to no avail, and I lack the domain knowledge to quickly look it up myself. Perhaps one could also check genetic change over time, perhaps social arms race is something you can see in this data? Do we know what this looks like in humans and orcas?
1
1Very interesting write-up! When you say that orcas could be more intelligent than humans, do you mean something similar to them having a higher IQ or g factor? I think this is quite plausible.
My thinking has been very much influenced by Joseph Henrich's The Secret of Our Success, which you mentioned. For example, looking at the behavior of feral (human) children, it seems quite obvious to me now that all the things that humans can do better than other animals are all things that humans imitate from an existing cultural “reservoir” so to speak and that an individual human has virtually no hope of inventing within their lifetime, such as language, music, engineering principles, etc.
Gene-culture coevolution has resulted in a human culture and a human body that are adapted to each other. For example, the human digestive system is quite short because we've been cooking food for a long time, humans have muscles that are very weak compared to those of our evolutionary cousins because we've learned to make do with tools (weapons) instead and we have relatively protracted childhoods to absorb all of the culture required to survive and reproduce. If we tried to “uplift” orcas, the fact that human culture has co-evolved with the human body and not with the orca body would likely be an issue in trying to get them to learn it (a bit like trying to get software built for x86 to run on an ARM processor). Still, I think progress in LLM scaling shows that neural networks (artificial or biological) are able to absorb a significant chunk of human culture, as long as you have the right training method. I've made a similar point here.
There is nothing in principle that stops a chimpanzee from being able to read and write English, for example. It’s just that we haven’t figured out the methods to configure their brains into that state, because they don’t have a strong tendency to imitate, which human children do have, which makes training them much easier.
Follow up to: Could orcas be smarter than humans?
(For speed of writing, I mostly don't cite references. Feel free to ask me in the comments for references for some claims.)
This post summarizes my current most important considerations on whether orcas might be more intelligent than humans.
Evolutionary considerations
What caused humans to become so smart?
(Note: AFAIK there's no scientific consensus here and my opinions might be nonstandard and I don't provide sufficient explanation here for why I hold those. Feel free to ask more in the comments.)
My guess for the primary driver of what caused humans to become intelligent is the cultural intelligence hypothesis: Humans who were smarter were better at learning and mastering culturally transmitted techniques and thereby better at surviving and reproducing.
The book "the secret of our success" has a lot of useful anecdotes that show the vast breath and complexity of techniques used by hunter gatherer societies. What opened up the possibility for many complex culturally transmitted techniques was the ability of humans to better craft and use tools. Thus the cultural intelligence hypothesis also explains why humans are the most intelligent (land) animal and the animals with the best interface for crafting and using tools.
Though it's possible that other factors, e.g. social dynamics as described by the Marchiavellian Intelligence Hypothesis, also played a role.
Is it evolutionarily plausible that orcas became smarter?
Orcas have culturally transmitted techniques too (e.g. beach hunting, making waves to wash seals off ice shells, faking retreat tactics, using bait to catch birds, ...), but not (as far as we can tell) close to the sophistication of human techniques which were opened up by tool use.
I think it's fair to say that being slightly more intelligent probably resulted in a significantly larger increase in genetic fitness for humans than for orcas.
However, intelligence also has its costs: Most notably, many adaptations which increase intelligence route through the brain consuming more metabolic energy, though there are also other costs like increased childbirth mortality (in humans) or decreased maximum dive durations (in whales).
Orcas have about 50 times the daily caloric intake of humans, so they have a lot more metabolic energy with which they could power a brain that consumes more energy (and can thereby do more computation). Thus, the costs of increasing intelligence is a lot lower in orcas.
So overall it seems like:
Though it's plausible that (very roughly speaking) past the level of intelligence needed to master all the cultural orca techniques (imagine IQ80 or sth) it's not very reproductively beneficial for orcas to be smarter for learning cultural techniques. However, even though I don't think it's the primary driver of human intelligence evolution, it's plausible to me that some social dynamics caused selection pressures for intelligence that caused orcas to become significantly smarter. (I think this is more plausible in orcas than in humans because intelligence is less costly for orcas so there's lower group-level selection pressure against intelligence.)
Overall, from my evolutionary priors (aka if I hadn't observed humans evolving to be smart) it seems roughly similarly likely that orcas develop human-level+ intelligence as that humans do. If one is allowed to consider that elephants aren't smarter than humans, then perhaps a bit higher priors for humans evolving intelligence.[1]
Behavioral evidence
Anectdotes on orca intelligence:
Two more anecdotes showing orcas have high dexterity:
Also, some orca populations hunt whales much bigger than themselves, like calfs of humpback, sperm, or blue whales. Often by separating them from their mother and drowning them.
Evidence from wild orcas
(Leaving aside language complexity, which is discussed below,) I think what we observe from wild orcas, while not legibly as impressive as humans, would still be pretty compatible with orcas being smarter than humans (since it's find sth we don't observe, but what we probably would expect to see if they were as smart as us).
(Orcas do sometimes get stuck in fishing gear, but less so than other cetaceans. Hard to tell whether humans in orca bodies would get stuck more or less.)
(I guess if they were smarter in abstract reasoning than the current smartest humans, maybe I'd expect to see something different, though hard to say what. So I think they are currently not quite super smart, but it's still plausible that they have the potential to be superhumanly smart, and that they are currently only not at all trained in abstract reasoning.)
Evidence from orcas in captivity
I mostly know of a couple of sublte considerations and pieces of evidence here, and don't share them in detail but just give some overview.
I think overall the observations are very weak evidence against orcas being as smart as humans, and nontrivial evidence against them being extremely smart. (E.g. if they were very extremely smart they maybe could've found a way to teach trainers some simple protolanguage for better communicating.)
I'm not sure here, but e.g. it doesn't seem like orcas learn tricks significantly faster than bottlenose dolphins, but maybe the bottleneck is just communication ability for what you want the animals to do. (EDIT: Actually orcas seem to often learn tricks a bit slower than bottlenose dolphins, though orcas are also often a lot less motivated to participate.) Still, I'd sorta have expected something more impressive, so some counterevidence.
Thoughts on orca languages
I have quite some difficulty to relatively quickly estimate the complexity of orca language. I could talk a bunch about subtleties and open questions, but overall it's like "it could be anything from a lot less complex to a significantly more sophisticated than human language". I'd say it's slight evidence against full human-level language complexity. (Feel free to ask for more detail in the comments. Btw, there are features of orca vocalizations which are probably relevant and which are not visible in the spectrogram.)
Very few facts:
Orca language is definitely learned; different populations have different languages and dialects.
It takes about 1.5 years after birth[2] for orca calfs to fully learn the calls of their pod (though it's possible that there's more complexity in the whistles, and also there are more subclusters of calls which are being classified as the same calltype).
Louis Herman's research on teaching bottlenose dolphins language understanding
In the 80s, Louis Herman et al taught bottlenose dolphins to execute actions defined through language instructions. The experiments used proper blinding and the results seem trustworthy. Results include:
AFAIK, this is the most impressive demonstration of grammatical ability in animals to date. (Aka more impressive than great apes in this dimension. (Not sure about parrots though, though I haven't yet heard of convincing grammar demonstrations as opposed to it just being speech repetition.))
In terms of evolutionary distance and superficial brain-impressiveness, orcas are to bottlenose dolphins roughly as humans are to chimps, except that the difference between orcas and bottlenose dolphins is even a big bigger than between humans and chimps, so this is sorta promising.
Neuroscientific considerations
Orca brain facts
(Warning: "facts" is somewhat exaggerated for the number of cortical neurons. Different studies for measuring neural densities sometimes end up having pretty different results even for the same species. But since it was measured through the optical fractionator method, the results hopefully aren't too far off.)
Orcas have about 43 billion cortical neurons - humans have about 21 billion. The orca cortex has 6 times the area of the human cortex, though the neuron density is about 3 times lower.
Interspecies correlations between cortical neurons and behavioral signs of intelligence
(Thanks to LuanAdemi and Davanchama for much help with this part.)
I've tried to estimate the intelligence of a few species based on their behavior and assigned each species a totally subjective intelligence score, and a friend of mine did the same, and I roughly integrated the estimates together to what seems like a reasonable guess. Though of course the intelligence scores are very debateable. Here are the results plotted together with the species' numbers of cortical neurons[3]:
As can be seen, the correlation is pretty strong, especially within mammals (whereas the birds are a bit smarter than I'd estimate from cortical neuron count). (Though if I had included humans they would be an outlier to the top. The difference between humans and bottlenose dolphins seems much bigger than between bottlenose dolphins and chimps, even though the logarithmic difference in cortical neuron count is similar.)
(Also worth noting that average cortical neural firing rates don't need to be the same across species. Higher neuron densities might correlate with quicker firing and thus more actual computation happening. That birds seem to be an intelligent outlier above is some evidence for this, though it could also be that the learning algorithms of a bird's pallium is just a bit more efficient than that of the mammalian cortex or so.)
How much does scale vs other adaptations matter?
A key question is "how much does intelligence depend on scale vs other adaptations?".
Here are some rough abilities that seem useful for intelligence that seem like they might probably come in some way from non-upscaling adaptations (rather than just arising as side-effect of upscaling):
(Some of those might already exist to some extent in non-human land mammals too though.)
It's also conceivable that humans got more adaptations that e.g. increased the efficiency of synapsogenisis or improved the learning algorithms somewhat, though personally I'd not expect that a few million years of strong selection for intelligence in humans were able to produce very significant improvements here.
We should expect humans to have more of those non-scaling intelligence improving mutations: Orcas are much bigger than humans, so the fraction of the metabolic cost the brain consumes is smaller than in humans. Thus it took more selection pressure for humans to evolve having 21billion neurons than for orcas to have 43billion.[1] Thus humans might have other intelligence-increasing mutations that orcas didn't evolve yet.
The question is how important such mutations are in contrast to scaling up? And in so far as they matter, were they hard to evolve or easy to evolve once the brain was large enough to make use of metacognitive abilities?
My uncertain guess is that, within mammalian brains, scaling matters a lot more for individual intelligence, and that most of the subtleties of intelligence (e.g. abstract pattern recognition or the ability to learn language) don't require hard-to-evolve adaptations. (Though better social learning was probably crucial for humans developing advanced cultural techniques. Also, it's not like I think scale alone determines the full cognitive skill profile: I think there are other adaptations that can trade off different cognitive abilities, as possibly unrealistic example e.g. between memory precision and context generalization.)
Overall guess
Having read the above, you might want to try to think for yourself how likely you think it is that orcas are as smart or smarter than humans, before getting contaminated with my guess. (Feel free to post your guess in the comments.)
Orca intelligence is very likely going to be shaped in a somewhat different way than human intelligence. Though to badly quantify my estimates on how smart average orcas might be (in some rough "potential for abstract reasoning and learning" sense):
I'd say 45% that average orcas are >=-2std relative to humans, and 20% that they are >=6std.[4]
Aside: Update on my project
Follow up to: Orca communication project
I'm currently trying to convince a facility with captive orcas to allow me to do my experiment there, but the chances are mediocre. Else I'll try to see whether I can do the experiments with wild orcas, though it might be harder to get much interaction time and it requires getting a permit for doing the experiment with wild orcas, which might also be hard to get.
I'm now no longer searching for collaborators for doing the relevant technical language research work (though still reach out if interested)[5]. However, I'm looking for:
Those 2 roles can be filled by the same person. If you might be interested in filling one or both of those roles, please message me so we can have a chat (and let me know roughly how much money you'd want).
In case you're wondering, no this isn't a hindsight prediction from me having observed orca's large brains. Orcas are the largest animal engaging in collaborative hunting. Sperm whales would also be roughly similarly likely to develop intelligence on my evolutionary priors - they have even more metabolic energy though they are less social than orcas.
Note that orcas have about 17 months gestation period.
For asian elephants we actually don't have measurements, so I took estimated values from wikipedia, though hopefully the estimates aren't too bad since we have measurements for african elephants. Also measurements can be faulty.
Though again, it's about potential for if they got similar education or so. I'd relatively strongly expect very smart humans to win against current orcas in abstract reasoning tests, even if orcas have higher potential.
A smart friend of my tried to do the research but it seems like I'm just unusually good and fast at this research and it didn't seem like I could be sped up significantly, so I'm planning to do the technical research myself and find good ways to delegate the other work to other competent people.