Curran

@quinnmbarry @AaronBergman18 Might be the biggest problem in medicine. Many of the worst diseases have incredibly effective treatments limited only by chronic use. Yet highly effective means of treating tolerance exist, several with phase 3 evidence. We’re developing drugs based on some of these mechanisms.

@AaronBergman18 cc @Curranjans

If I was an equitied Ant(hropic employee) I’d at least seriously look into somehow funding anti-tolerance (to drugs) interventions

@QualiaNerd Likely the greatest source of humanities pain and it’s cure is limited only by human coordination. Don’t let EA charities or the medical establishment think they know better! The evidence is radically clear that cluster headaches are the #1 most cost effective health intervention

Seconded. I must stress: If you are suffering from cluster headaches, a safe and effective cure already exists (DMT). It won’t magically change by being rebranded by some pharma company. You don’t have to wait until Science™ catches up with science.

@favoredprisoner @algekalipso Yep definitely agree. Haven’t asked explicit permission of them but happy to talk about it 1-1 (I’ll DM!). An organization with that specific goal would be super valuable and something I’m looking to support

@Curranjans @algekalipso Any reason not to mention any names? (maybe with their permission if they might care). I think I and many other rats/EAs would take it a lot more seriously if we saw others doing so.

If I ever feel that you are grasping the fundamentals of the phenomenal binding problem, I'll let you know. It's incredibly rewarding for me when I see someone start to get it. It usually takes hours of thinking and conversation. But ohhhh my god, it's such a satisfying joint attention event, when someone can see what I see! 😀

@favoredprisoner @algekalipso Feels a lot like periods between discovery and public recognition of evolution or relativity. Not everyone will understand but a competent and motivated enough individual can become almost certain in it

@favoredprisoner @algekalipso I’ve convinced about half the high profile EA’s/rats I’ve tried talking to in the last couple months. Multiple CEO’s and leaders of organizations you would know. When I learned about it, it raised a lot of red flags for the same reason that have since been personally answered

@algekalipso 🥳🥳

I’m currently staying at the very core of Södermalm, Stockholm’s “Old Town”, getting my jetlag-ruffled bearings straight before I proceed onto Borderland to deliver a couple workshops on QRI topics (esp. visualizing phenomenology, mapping qualia, and philosophy of mind). I will then be participating in a low-key research retreat with influential figures in the field of consciousness for a few days, and finally on the 3rd of August, we shall host the very 1st Swedish QRI Meetup! Date: August 3rd 2024, 2PM onwards. Location: Riddargatan 18, Östermalm (please greet us outside, the place is on the 3rd floor*) Schedule: 2-3PM Casual Hangout 3PM Welcome Speech and Introductions 3:30-4:30PM Andrés will deliver a presentation on a surprise QRI topic 4:30 Participants will have an opportunity to share with the group who they are, what they are interested in, and what kinds of collaborations (if any) they would like to pursue 6PM Latest QRI Technology Demos 7PM onwards: returning to casual hangout until the end QRI Meetups are excellent places to connect with other people interested in consciousness, meditation, psychedelics, AI, math/physics, and reducing suffering at scale. We’ve hosted meetups in many cities and countries already**, and we consistently get the feedback that they play the role of a Schelling Point for “qualia people” to meet one another. You can think of it as a way to “activate latent connections” in a city and kickstart a community of like-minded individuals. Qualia of the Day: Please feel encouraged to bring an interesting experience to share with others. This could be a perfume, a candy, a toy, a gadget, a poem, or a brief (couple minutes) activity. One of the long-term goals of the Qualia Research Institute is to map out the state-space of consciousness. Qualia of the Day activities are a great way to enrich our evidential base in the pursuit of this quest.

@richardludlow @scottdomes @jonnym1ller @chrislakin @algekalipso @Plus3Happiness @m_ashcroft @visakanv @RogerThisdell For basic emotions like love or okayness I assume a simple low level physical implementation, likely the release of certain muscle tension in the brain. This tension may have been evolutionarily beneficial by making these emotions feel conditional, spurring action

Hey @scottdomes, @jonnym1ller, @chrislakin, @algekalipso, @Plus3Happiness, @m_ashcroft, @visakanv, @RogerThisdell do you have a definition for “emotion” that you like?

@teortaxesTex @algekalipso Smarter than me too. So was Einstein. But learning about relativity did not leave any room for “luck” for me crediting Einstein with a world changing discovery. The same for the Slicing problem. Skill issue for those relying on luck and wisdom of the crowd

@Curranjans Fair enough, I'm happy to admit @algekalipso is smarter than me. But so is Roger Penrose and yet I can smell bs in his thesis. Do as smart guy does is not enough of a heuristic. Ultimately, we are cursed to rely on luck for bets we can't truly evaluate.

Chances are, if you visit Argentina around 2025-2026 (because universal basic AGI will make travel cheap), you might run into a homeless but strangely smug Russian dude wearing this. Treat him to some beer and empanadas please. Buckwheat will be appreciated too

@teortaxesTex Sounds like another missed opportunity for teortaxes. I’m happy being selectively contrarian and finding confidence where others can’t though. You’re definitely smart enough to make a big impact here if only you could understand!

@teortaxesTex Haha real easy to say! A lot of people were bearish on Einstein before evidence proved him right. A lot of people bearish on Nvidia before earnings proved it right. But a smart enough actor can have a really high probability in all three of these events and be rewarded massively

@teortaxesTex You’re better than most experts and the leading AI labs. This is wrong though. There’s very few useful conversations about consciousness and they pretty much only come from those familiar with QRI ;)

@Curranjans Out of curiosity, what is your probl with my takes on consciousness

@teortaxesTex Sounds like a skill issue. Way too many AI grifters are getting way too much success for you to squander your AI intuitions. Will you suddenly stop being an expert on global AI dynamics or will it not be valuable? I can think of a lot of profitable benevolent pre AGI jobs

@Curranjans I'm currently unable to buy stocks of American tech, and thus do not expect the transition* to go well for me or Argentina *to AGI era

@eshear @nowtheo It’s like critiquing weightlifting because you tried it once and it didn’t work and then saying maybe it just doesn’t work for healthy people. True belief or not it’s not useful and I’m disappointed in Scott for being so dismissive of the best mental health intervention I’ve seen

@nowtheo It feels totally sincere to me, that’s 100% Scott’s true beliefs. Ignorant, maybe! I think Scott would be the first to admit he is capable of error and would welcome criticism. I’m no IFS expert so I can’t judge. But jumping to personal attacks like this is uncharitable.

Yikes, Scott Alexander's intro section in his latest post (Book Review: The Others Within Us) where he introduces IFS feels quite insincere and ignorant. The whole bit about “trance” or “bargaining” or ironically saying “discover not invent” etc makes me feel like he doesn’t get it. I don't see IFS as being about bargaining or negotiating with parts. In fact that kind of frame can be counter-productive to the process. The intention is to connect with yourself, trying to "win" against your parts or impose agendas would lead to disconnection. You don't need to go into any "trance" to connect with your parts. And you don't even need to take the "parts" frame that seriously. You "discover" parts by paying attention to the nature of suffering in your bodymind. You feel into the suffering that arises for you in a particular context and then notice what kind of tensions or constrictions or unease show up in your raw sensory experience. The tension is usually in place to avoid some (perceived) unpleasant emotional experience. To get a better sense of it, you can play with active imagination and narrativization to get a handle on the causality of the suffering you are investigating. This is where the idea of parts can be quite useful. You imagine the tension as a part and then relate to it from a place of openness/curiosity/compassion etc so that you don't inadvertently end up tensing against the tension (which would just create more tension and suffering in the system). Then the various steps of IFS (or whatever flavour of parts work resonates for you) are ways of playing with your attention/awareness and connecting with the part in a welcoming way so that you can create emotional safety for the tension to relax. When that relaxation happens, the emotional experience that was being suppressed/avoided comes to the surface and gives you the opportunity to again similarly connect with that part. You don't need "tricks" or "plans" to do this exploration, you need attunement and openness. Scott's description feels like a poor caricature of an interesting process of exploration, which has worked beautifully for many many people (not just "severely traumatized borderlines"). Given how he approached it, it's not surprising that it didn't "work" for him.

@teortaxesTex @xlr8harder Now we’re lost

@xlr8harder yes any object can be construed as having a perspective it's not "consciousness" in the sense that it has complex contexts like self-attention

Nora often bullies me about rigorous definition of consciousness but this feels 100% correct Moreover, I think ≈every thing can be meaningfully said to have such a "perspective". The question of moral salience is: what is a person? (Intuitively, not the persona Opus emulates)

@teortaxesTex You can induce experience to be in the third person with meditation. The first person part is hallucinated and just a part of modeling the world. The experiences feel just as valuable

@teortaxesTex @JacquesThibs Haha maybe, It’s always a fine line. If they don’t scrap it there’s ways to choose projects with different goals here and the new leads may have different motivations

@Curranjans @JacquesThibs Cartoonish

I wonder what’s going to happen to that 20% of “current compute power” that was promised for Superalignment

@teortaxesTex @JacquesThibs That’s like a billion dollars of GPU hours. Seems like it’s be pretty easy to turn it into synthetic data for capabilities with the label of safety research

@JacquesThibs Do you assume they'll scrap it?

@teortaxesTex Yea like all safety research is bananas. I think things like fear are much more explanatory here than uncuriosity or a pivot. It’s easy to say you’re not driven by emotion if you haven’t worked on it very much. If you’re not genuinely annoyed by people your takes become better

@Curranjans I don't think this is emotion-driven at all, anyway this guy's safety "research" is bananas too x.com/teortaxesTex/s…

> We get to decide what is/isn't a warning shot > We missed the opportunity… > We should heavily mediatize what is happening What is? Such is the new generation. EU AI decels: uncurious people who pivoted from dead-end ML careers into the burgeoning field of safety activism.