Currently working as an independent research contractor, primarily for the Qualia Research Institute. Previously worked as Chief of Staff at the Institute for Law & AI (formerly âLegal Priorities Projectâ) and as COO at the Center on Long-Term Risk (formerly âEffective Altruism Foundationâ). I also co-founded EA Munich in 2015. I have a PhD in Computational Science from TU Munich.
đś 10% Pledger
I get what you mean! Here are a few very quick thoughtsâhope they make sense!
Maybe one challenge around introspecting on binding is that one typically has to be in altered or ill states of consciousness to experience what itâs like for binding to break down. For instance, under some psychedelics, the color of an object can âbleed outâ of its boundaries.
In normal waking consciousness, we just take binding for granted. We have no problem recognizing that the eyes, nose, mouth, etc. of the person in front of us are âgluedâ together into a unified gestalt that we can make sense of (and e.g. recognize as our friend). Some people with integrative agnosia canât do that.
As David Pearce says, no binding = no mind. Without binding, there would only be âqualia dustâ floating around. It makes sense that evolution would recruit bound experiences to make sense of the external world. Ethically, we should care about bound experiences.
Binding is also the reason why you canât do your taxes on a high dose of LSD (though maybe altered states of consciousness can be used for other types of computation).
Happy to chat about this more in person one of these days! I just moved to London and plan to visit Oxford every now and then. âşď¸
Thanks for sharing your thoughts!
Iâm unsure whether we can in principle ascertain whether a digital mind is conscious.
I used to believe this too, but Iâve actually become quite optimistic that this is a tractable question. If, as David Pearce argues, the laws of physics donât break down inside the brain, and we define the constraints that a satisfactory theory of consciousness should fulfill, then I believe we can make very good progress. For instance, one such constraint is that the theory should solve the binding problem (see also here). So if we cannot observe a binding mechanism in digital computers, or explain how digital computers could in principle give rise to binding (as I and other at QRI argue), then that should be very strong evidence against digital computers being conscious. More in: Digital Sentience: Can Digital Computers Ever âWake Upâ?
(Note that this doesnât rule out other forms of artificial, non-digital consciousness!)
Thanks for sharing this! You might be interested in this: Consciousness Isnât Substrate-Neutral: From Dancing Qualia & Epiphenomena to Topology & Accelerators
Leaving this here: âTestimonial of a chronic cluster headache patient after using DMT to abort attacksâ. @jonleighton of the Organisation for the Prevention of Intense Suffering interviews John Fletcher, chronic cluster headache patient.
JF: Iâm 65 years old now and they [cluster headaches] started for me in late September 1973, so about 51 years Iâve had cluster headaches. [âŚ] I came out of a seven-year remission and as soon as I did, I havenât had a break since, and that was 6 months ago. Iâve been getting at least 10 attacks a day, every day.
JL: Can you tell us a little bit more about the other medical conditions that youâre suffering from at the moment, just to give a bit of context?
JF: Besides stage five COPD [chronic obstructive pulmonary disease], I have adenocarcinoma lung cancer. I have achalasia disease, which my esophagus swells up and ultimately causes me to aspirate food and causes bacteria pneumonia. And Iâve got multiple abdominal hernias from surgery I had 20 major surgeries. I broke my back at the beginning of the year that was caused by severe osteoporosis from steroids, from my COPD. And Iâve broken my ribs 30 or 40 times in the last couple of years from severe osteoporosis.
JL: How would you compare the suffering due to cluster headaches compared to everything else that youâre experiencing?
JF: I wouldnât trade anything for cluster headaches. None of it. Cluster headaches is the worst pain Iâve ever had in my life. Iâve never felt anything worse. And, I mean, including being terminally ill and breaking and just so many other really severe painful conditions, but cluster headaches is in a in a category by itself. Iâve never felt anything like it. Itâs such severe pain, itâs literally violent screaming pain and, to put it bluntly, it feels like youâre being murdered. Itâs unbelievable pain. And you expect to see blood. Telling people about my first attack, I thought I was shot in the head. I thought somebody shot me and I was dying. Every attack pretty much feels like youâre dying. Itâs just horrible, horrible thing. So no, I wouldnât trade any of it for cluster headaches.
JL: And what medications have you tried? And have any of them worked for you over the years?
âŚ
Thanks for sharing, Deborah! Iâll add these resources to my list of interventions. :)
Thanks so much for your comment!
Actually, someone else brought up this point separately, so I agree thereâs more to say here. Iâd love to dig deeper into this question and possibly write a paper on the topic (e.g. for this collection). If you have literature to recommend (either by you or otherwise), please send it my way. And also let me know if youâd like to get involved in such a project. :)
Gee, not sure what happened thereâthanks for pointing that out! Iâve edited the footnote.
Thank you for your comment, Tim!
Indeed, the choice of e is arbitrary and used for illustration purposes. And the base 6 is simply the choice for which the total burden of CH is larger than that of migraines, so itâs also not derived from first principles. This footnote is relevant:
The resulting scaling as would mean that the 0â10 scale would have to span 4 orders of magnitude. While GĂłmez-Emilsson & Percy (2023) suggest the scale spans âat least two orders of magnitudeâ, private communication with the authors indicates their central estimates might be closer to 4 orders of magnitude, with uncertainty ranging from 2 to 8 OOMs.
The paper cited also mentions the possibility of a linear relationship for lower pain intensities and an exponential relationship at higher intensities (a âkinkedâ distribution), highlighting the fact that there are more possibilities beyond a uniform exponential increase.
I personally donât have a good intuition for what the base should be but might do more work on this specific question.
Iâm also not sure what the optimal mapping of intensities for the Russell vs Torelli & Manzoni scales is, also considering the fact that the two studies had different methodologies. I think thereâs no correct answer, so that was my best guess (though I could also imagine âVery slightâ being more intense than a 1.5). Do let me know if you have other suggestions! (Or feel free to fork the code and play around with the parameters. :) )
Thanks! Itâd be great if someone (maybe myself, but ideally someone with more experience in the field) published a summary of the existing literature (more research here). Having spent so many hours reading up on the topic these past few months, Iâm optimistic about the efficacy. I think funding and/âor running a large scale RCT in particular for N,N-DMT (in a country where it is legal) would be a great use of EA money/âtime.
I think the EA community has shown incredible initiative in tackling major global health issues, making a lot of progress on problems such as malaria (which causes 600,000 deaths/âyear) and lead poisoning (which causes 1.5M deaths/âyear), among so many others. These efforts really show our ability to mobilize resources and drive change when we identify pressing problems.
My hope is that we can direct a similar amount of attention to helping the ~3 million people worldwide who have this terrible condition. Even if my quantitative estimates of the burden of pain were off by an order of magnitude, the situation would still be tragic (and, as @algekalipso has pointed out, somewhat analogous to times when anesthesia had already been invented but not adopted, given the promise of low dose psychedelics1). I think it would be an incredible success story for our community if we managed to eliminate (or at least significantly reduce) this source of enormous suffering. If youâd like to contribute in any wayâeither with time or fundingâplease get in touch!
1 Coincidentally, when I asked Claude to estimate the lifetime prevalence of undergoing major surgery without general anesthesia before it was invented, its initial guess was surprisingly similar to the lifetime prevalence of cluster headachesâ0.2%.
Done! Thanks for sharing this!
I think that a corollary of the first point is that we can learn a lot about alignment by looking at humans who seem unusually aligned to human values (although I think more generally to the interests of all conscious beings), e.g. highly attained meditators with high integrity, altruistic motivations, rationality skills, and a healthy balance of sytematizer and empathizer mindsets. From phenomenological reports, their subagentic structures seem quite unlike anything most of us experience day to day. That, plus a few core philosophical assumptions, can get you a really long way in deducing e.g. Anthropicâs constitutional AI principles from first principles.
(The link isnât working for me.)
Nearly five years since this post was written, and still so little attention and virtually no funding in EA is dedicated to this problem. Ignoring the logarithmic distribution of suffering continues to be my top contender for biggest blindspot in the EA community. My heart goes out to the likely millions of people worldwide who suffer hell-level pain on a regular basis. Iâll continue to think of ways to help.
I really liked your talk! Congratulations, and I hope the idea continues to catch on!
Qualia Research Institute, plus some pro bono time for a project at the intersection of law and AI.