I feel a bit crazy in that the recent METR report is explicitly saying that the measured progress is on well-specified tasks and that they cannot tell how much of this transfers to ill-specified tasks; and now half the blogs in AI safety sphere seem to be taking this as "ASI is imminent", dropping the caveat. In METR's words:
Acceleration might be limited to well-specified tasks. Much of the evidence we looked at measures performance on well-specified tasks that are easy to grade automatically or with a language model. If recent models have been growing more quickly on “clean” tasks than on “messy” tasks (relative to historical relative growth rates), then we may be overstating the acceleration in real-world impact. For instance, recent progress may be overfitting to the types of tasks which are included in AECI.
If we assume that humans are capable of solving ill-specified tasks, then at least by a "better than humans at any cognitive task" definition of ASI, ASI would need to be capable of solving ill-specified tasks. And "progress on well-specified tasks is accelerating" does not automatically translate to "soon we will understand how to solve ill-specified tasks".
I do think that ... (read more)
A few months ago I gave a ~10% probability of TEDAI within 12 months; my intuition is now closer to ~30% if we go full speed ahead. Some of that was from the passage of time, but it's also about a ~2x update.[1]
For me the math and CS research progress is a bigger update than the METR/Anthropic reporting. I think we saw maybe a decade of progress on the hardest open problems in math and theoretical computer science with no sign of relevant capabilities slowing down. The ability to outperform large communities of human researchers at the task of designing and improving algorithms is a core part of an intelligence explosion, so I think that's a really important indicator.[2]
I also think it's useful to track accelerated progress on well-defined metrics (like pretraining or RL performance). Regardless of whether that progress transfers, it indicates that underlying technical developments are accelerating. And beyond that, I think "we haven't seen the pace of progress double yet" is a good argument that the singularity is likely at least 6-12 months away (at least for a trend-extrapolator like me). So once it's happened it's a lot harder to be confident that you won't see something crazy... (read more)
For me the math and CS research progress is a bigger update than the METR/Anthropic reporting… The ability to outperform large communities of human researchers at the task of designing and improving algorithms is a core part of an intelligence explosion, so I think that's a really important indicator.
This was also a bigger update for me, and it seems like a similarly-sized one, so perhaps we don’t disagree that much. I’m curious, though, if you are differentiating between kinds of algorithm improvement/design. It seems like AI improving AI R&D could fall anywhere between “improving dataset quality” to “coming up with an AIXI-like algorithm from scratch.” My own sense is that while AI systems are getting better at a subset of algorithm improvement/design it does not yet extend to the latter kind of thing, nor does it seem obvious that they will get better at this latter kind of thing, and I think this plausibly matters for RSI. I’m not sure if/where we disagree though. Is your claim that AI systems are already outperforming humans at all types of algorithm design, or that they are not outperforming us at the “designing AIXI-like algorithms from scratch” type but will soon, or that they are not outperforming us at this and won’t but that this doesn’t bear on RSI? (Or another option I didn’t list)?
I feel a bit crazy
Same, I don't understand what's triggering all these mood shifts. Am I out of the loop on something? Am I delusional and completely out of my mind on copium? I've spent way too much time the last few days trying to troubleshoot myself, and I still don't think I am...
I would have understood if this reaction was in response to the Navier-Stokes resolution, that was pretty freaky. But the Anthropic report doesn't seem to merit any of this whatsoever. I guess maybe this isn't really about the report at all, it's some sort of preference cascade that was just triggered by it, the last drop of evidence many people needed?
A lot of people really do consider benchmark scores or productivity metrics (and the extrapolations from them) load-bearing, so it's not surprising to me that the METR report is relevant evidence for them. Since I mostly ignore benchmark scores, this was a complete no-op for me (unlike the Anthropic pretraining efficiency plot, which is a big update), but I don't think it's unexpected that a lot of people did update.
So the "Is everyone crazy, or am I crazy?" thing is about different models, which was already a live distinction before the METR report. Similarly to how there's now a permanent hazard rate of 5+% for 1-2 year timelines to a takeoff, for which the METR report isn't a crux (it only moves the probabilities around), there are preexisting different ways of thinking about the role of pretraining scale vs. benchmark scores in strong ASI timelines, and the METR report is not a crux for adjudicating this distinction.
There's a bit of a motte and bailey here:
Motte: Internal METR/Anthropic evidence prove ~2x research uplift which prove short timelines.
Bailey: AECI trends prove ~2x research uplift which prove short timelines.
Many people have raised good arguments against the latter, but then the kernel of the argument rests on the former which is basically impossible as an outsider to use.
There is a very real sense in my mind that this website is becoming alarmist. This is one example, but the discussion immediately after the Weight-Space programming paper is another. It's right to want to update timelines quickly with new evidence, but it feels like these updates are driven mostly by the evidence feeling correct in atmosphere rather than in the quality of evidence per se.
There is a very real sense in my mind that this website is becoming alarmist
Yeah, agreed. Frankly, the end result of my checking LW lately has been increased stress levels for the day without much productive new information being learned, almost on the level of doomscrolling Twitter. The unpleasant thing is that my prior on LW being reasonable and relatively well-calibrated on these matters is high, so I'm still not totally sure that it's LW's collective epistemics that are at fault, instead of LW being at a valid level of alarm and my being in denial. But if I keep to object-level analysis, I'm inclined towards the former.
I’m in a similar position. I’ve already mostly stopped going on Twitter because of this (i.e. the panic to useful info ratio is terrible), but I am really sad to make such an update about LessWrong. And, as you say, it's especially unpleasant given that LessWrong tends to be reasonable and careful. So I am constantly questioning myself and wondering if I’m just grasping onto the so-called “hopium,” even though I too really don’t think that’s what’s going on.
Overall, the situation is very rabbit/duck for me. As in, I feel like I'm always flitting between two competing but incompatible versions of reality. When I look at the world “through my own eyes,” when I use the models, when I read about how other people use the models, when I think about the models, when I think about what intelligence is and what makes humans so powerful, when I hold all of that in my head on my own terms, on a gut level I really struggle to see it. On a gut level, my experience tells me that the machines can’t “really” think, and on a gut level, this really seems like it ought to matter.
And yet there is another part of me that is more in touch with abstract arguments which thinks things like “exponentials ... (read more)
Hmm, your point about trying to intuit exponentials is interesting… try this out: we have well grounded intuitions… for uncontrolled exponentials:
Imagine someone putting a termite queen in your house and telling you it only has a 10% chance of establishing an infestation.
Imagine someone handling an explosive near you and telling you there’s only 10% chance they drop it and it explodes.
Intuitively, you don’t feel like you are 90% safe. You feel like you are in active danger and must act with certainty and swiftness.
I don’t know, does that work as a frame?
Unfortunately, this matches my experience as well. Based on concrete, object-level observations and arguments, I have pretty long timelines. But whenever I open LessWrong, I see a lot of smart people claiming that AGI/ASI is right around the corner. It feels a little disconcerting to have my predictions contradict those of someone as knowledgeable about AI as, say, Daniel Kokotajilo. But I suspect this is what it felt like to go up against Anselm when he was making his ontological argument. The intuition that “he is too smart to be so wrong” would have been just as strong back then. So in the end, I made peace with the fact that one group of smart people (either those with short timelines or those with long ones) is going to be wrong, and I have now basically stopped most of my deference on timelines. For the same reason, I have also mostly stopped following AI discourse on Twitter and LessWrong. The magnitude of the updates seems to far outstrip the warrant provided by the available evidence (by my lights).
We said in our blog post that it wasn't just the METR report. IIRC we specifically mentioned the math progress. Our thinking is similar to Paul's above.
Also, you really aren't doing justice to the report if you think it doesn't seem to merit any of this whatsoever. Come on, it's a serious chunk of evidence, surely you agree with that?
Also, you really aren't doing justice to the report if you think it doesn't seem to merit any of this whatsoever. Come on, it's a serious chunk of evidence, surely you agree with that?
I dunno. The 2x multiplier is only significant evidence for any practical outcomes within specific models of how AI progress works (AIFP's, Ryan's) which have strong beliefs regarding what that multiplier should be at which stages of the takeoff. I don't have such strong beliefs, I don't think I know what a 2x multiplier at a given moment implies about that multiplier the next month, or about the imminence of AC/SAR/TED-AI. And given how much of a surprise this data is to the aforementioned models, I think a strong update away from those models is justified, which in turn is an update away from assigning this multiplier the semantic meaning these models give it.
That is, it's a serious chunk of evidence for imminent RSI within models that this chunk of evidence itself makes you update against.
Once we take that away, the report is too redacted and caveated to definitively establish anything. Like, I didn't expect LLMs to not be accelerating LLM research at this point, and if this is happening, then of ... (read more)
It's only evidence against our model insofar as the alternative models predicted this surprising thing even more. Do you have an alternative model in mind, that successfully predicted the METR report in advance? Failing that, do you have an alternative model that at least retrodicts it well?
I agree that the METR report on its own is not conclusive, that's why we talked about other sources of evidence that were important to us also.
Anyhow I still think you aren't doing this thing justice. 2x acceleration is at least a significant chunk of evidence for the hypothesis that the intelligence explosion is soon, no? You agreed that the Mathapocalypse is evidence, why wouldn't an AI-driven acceleration in AI capabilities progress more generally also be evidence? Or are you saying that there is probably no such acceleration? Or are you saying that 2x accelerations happen all the time historically and then are followed by decelerations and therefore don't mean much when they happen now?
It's only evidence against our model insofar as the alternative models predicted this surprising thing even more
Sure, but that doesn't mean we end up falling back to some specific alternate strong model. It may be that we end up with a basically uniform distribution over the space of possible strong models.
Like, if there's 1000 urns with one ball in one of them, and there's been some promising evidence that the ball is in urn #461, but then we looked there and it was empty, that doesn't mean I'm rejecting the urn-461 hypothesis in favor of a strong belief in any specific other urn. We end up with basically no precise idea where the ball is.
I have no idea how to model AI progress at the level of precision AIFP's model aspires to. I don't think AIFP's model succeeds at its aspiration.
Or are you saying that 2x accelerations happen all the time historically and then are followed by decelerations and therefore don't mean much when they happen now?
Something like this, maybe. Imagine a specialized AI technique were discovered that made AI really, really good at hardware design. Suppose its efficiency naturally scaled with compute.[1] Depending on how powerful it is quantitatively, that d... (read more)
I think it's a mixture of 1) a preference cascade in that people who've been gradually moving towards very short timelines now feel able to say this (for example, I think Dwarkesh said in his post that he's shifted from ~a decade to ASI to ~a few years over the last few months) and 2) I also sort of assume that some of the prominent people predicting short timelines have a good idea of what's in METR's unredacted report, since they're in the same social circles as METR, so they know that the redacted stuff is strong evidence (but obviously they can't say this without getting METR in trouble)
This whole well-specified-task thing smells like copium to me. I concede there is some truth to it -- AIs, like humans, tend to be relatively good at the things they've been trained to do vs. the things they haven't been trained to do for example, and it's easier to train AIs to do math and coding than to run a business for example because it's more expensive to do business-running RL rollouts. Moreover, as a matter of alignment, AIs seem motivated to get high scores and thus maybe underperform on tasks that don't have a nice clean metric to optimize and instead are more open-ended.
Is that all it is though? Because a lot of people seem to think there's more to it than that, that (1) there's some important distinction in mind-space between "well-specified-task skills" and "non-well-specified task skills" and that (2) the AI progress of the last 5 years hasn't touched, or has barely touched, the "non-well-specified task skills" and that (3) all the massively accelerated AI research and hill-climbing on things like data-efficiency, pretraining loss, AECI score, etc. isn't going to lead to large gains in non-well-specified-task-skills... I guess I'd like to see someone lay out the case, and make an affirmative argument for premises 1-2-3. I think all three premises are doubtful.
I think this is an unfair characterization of the "fuzzy task means that timeline to TEDAI is >= 4 years away camp." Quickly responding to you points off the cuff:
To me one of the clearest examples seems to be what you get when you ask an LLM to analyze articles on various topics.
If someone gives me a piece of writing and asks me for my thoughts, what I'm doing is something like taking the claims in the article and testing them against my model of the world to see whether they match, or whether my model throws up errors on trying to execute the claim. Then I'll point out what I agree with and what I don't agree with.
And especially for a series of complex claims on ill-specified topics, it really doesn't seem to me like LLMs would have a particularly coherent world-model they'd be testing claims against? I haven't used OpenAI's models in a while, but my experience with Claude was that the models tended to be pretty sycophantic up to Opus 4.6, then with Opus 4.7 and 4.8 Anthropic tried to reduce it by making them reflexively disagreeable. So then they'd find reasons to disagree for the slightest reason, and after that got obviously excessive it was dialed down in the most recent models.
But a lot of it still looks like they are just being trained with various heuristics for when to disagree rather than actually running them against a coherent ... (read more)
the measured progress is on well-specified tasks and that they cannot tell how much of this transfers to ill-specified tasks ... "ASI is imminent", dropping the caveat
This could work in the sense that a lot of ML R&D and thus likely invention of the immediate precursors to strong ASI might be well-specified. Apart from the METR report, there's now the pretraining efficiency plot from Anthropic, so even with my own perspective of being skeptical about the strategic importance of post-training or current benchmark scores (beyond making the model sane enough to elicit useful work from it), I've now updated to expect in 2027 the kinds of models I previously expected in 2028, a nearly 1-year shortening of the very near future portion of my timelines (within 2026-2029; what happens later didn't change much in my view) that now gives a 20% chance of the takeoff starting by end of 2027 (meaning the world as we know it ends within a year after that).
Worth noting that they also make the opposite point: that the AECI scores might be underestimating the extent the extent to which AI capabilities have improved.
AECI could understate real-world capability impact. We have sparse data on the relationship between benchmark scores and real-world impact. As one example, ECI scores on math have not accelerated substantially, but we are now seeing the resolution of Millennium Prize Problems that have evaded decades of human effort. Real-world impacts may be downstream of capability thresholds that models suddenly pass, leading to sudden impacts that are hard to predict.
Fair, but at least for "ASI might take over the world" type threat scenarios, taking over the world is not a well-specified task. So "no obvious acceleration on ill-specified tasks" also implies "no obvious acceleration on ASI trying to autonomously take over the world and killing all humans".
That said, progress on well-specified tasks does still worsen misuse scenarios like "someone uses an AI to make biological weapons" or loss-of-control scenarios like "AI does something unexpected like it did with the HuggingFace thing, only this time it brings down critical infrastructure and kills a bunch of people before being shut off".
So "this might not get us ASI" doesn't mean "we're safe", it just means "this particular way of everyone dying might not happen, others still might".
I think the term "AGI" is a bit of a historical artifact, it was coined before the deep learning era when previous AI winters had made everyone in the field reluctant to think they could make any progress toward general intelligence. Instead, all AI had to be very extensively hand-crafted to the application in question. And then some people felt like they still wanted to do research on what the original ambition of AI had been, and wanted a term that'd distinguish them from all the other people who said they were doing "AI".
So it was a useful term to distinguish yourself from the very-narrow AI research back then, but now that AI systems are already increasingly general, it doesn't seem like a very useful concept anymore and it'd be better to talk in terms of more specific cognitive capabilities that a system has or doesn't have.
Every now and then in discussions of animal welfare, I see the idea that the "amount" of their subjective experience should be weighted by something like their total amount of neurons. Is there a writeup somewhere of what the reasoning behind that intuition is? Because it doesn't seem intuitive to me at all.
From something like a functionalist perspective, where pleasure and pain exist because they have particular functions in the brain, I would not expect pleasure and pain to become more intense merely because the brain happens to have more neurons. Rather I would expect that having more neurons may 1) give the capability to experience anything like pleasure and pain at all 2) make a broader scale of pleasure and pain possible, if that happens to be useful for evolutionary purposes.
For a comparison, consider the sharpness of our senses. Humans have pretty big brains (though our brains are not the biggest), but that doesn't mean that all of our senses are better than those of all the animals with smaller brains. Eagles have sharper vision, bats have better hearing, dogs have better smell, etc..
Humans would rank quite well if you took the average of all of our senses - we're el... (read more)
To me the core of neuron counting as an intuition is that all living beings seem to have a depth to their reactions that scales with the size of their mind. There's a richness to a human mind in its reactions to the world which other animals don't have, just as dogs have a deeper interaction with everything than insects do. This is pretty strongly correlated with our emotions for why/when we care about creatures, how much we 'recognize' their depth. This is why people are most often interested when learning that certain animals have more depth than we might intuitively think.
As for whether there is an article, I don't know of any that I like, but I'll lay out some thoughts. This will be somewhat rambly, in part to try to give some stronger reasons, but also related ideas that aren't spelled out enough.
One important consideration I often have to keep in mind in these sorts of discussions, is that when we evaluate moral worth, we do not just care about instantaneous pleasure/pain, but rather an intricate weighting of hundreds of different considerations. This very well may mean that we care about weighting by richness of mind, even if we determine that a scale would say that two be... (read more)
I doubt that anyone even remembers this, but I feel compelled to say it: there was some conversation about AI maybe 10 years ago, possibly on LessWrong, where I offered the view that abstract math might take AI a particularly long time to master compared to other things.
I don't think I ever had a particularly good reason for that belief other than a vague sense of "math is hard for humans so maybe it's hard for machines too". But formally considering that prediction falsified now.
Relative to 10 (or whatever) years ago? Sure I've made quite a few of those already. By this point it'd be hard to remember my past beliefs well enough to make a list of differences.
Due to o3 specifically? I'm not sure, I have difficulty telling how significant things like ARC-AGI are in practice, but the general result of "improvements in programming and math continue" doesn't seem like a huge surprise by itself. It's certainly an update in favor of the current paradigm continuing to scale and pay back the funding put into it, though.
"About thirteen or fourteen years ago, when I was still an academic, I interviewed for a [psychology] faculty position at a fancy university in Canada. [...] There were four applicants, each hosted by a different committee member. I had a grand time with my host. He was very candid about the whole process. At one point he told me that after the candidates had all come and gone, the committee members would go into a small room, sit down, and proceed to argue with one another for several hours over utterly ridiculous things each candidate said or did, until eventually everyone was worn out and just settled for whichever candidate they were all able to barely tolerate."
"I asked why the committee didn't just apply the lessons of seventy years of industrial-organizational psychology: come up ahead of time with a scoring rubric; have everyone independently fill it out without talking to each other; then mechanically aggregate the scores and offer the job to the person with the biggest number."
"He said (and he was right), “well, obviously, we don't believe our own methods.”"
"You could try to write it off as just a dysfunctional department, but… it wasn’t. They’re all like that. I’ve never... (read more)
Occasionally I find myself nostalgic for the old, optimistic transhumanism of which e.g. this 2006 article is a good example. After some people argued that radical life extension would increase our population too much, the author countered that oh, that's not an issue, here are some calculations showing that our planet could support a population of 100 billion with ease!
In those days, the ethos seemed to be something like... first, let's apply a straightforward engineering approach to eliminating aging, so that nobody who's alive needs to worry about dying from old age. Then let's get nanotechnology and molecular manufacturing to eliminate scarcity and environmental problems. Then let's re-engineer the biosphere and human psychology for maximum well-being, such as by using genetic engineering to eliminate suffering and/or making it a violation of the laws of physics to try to harm or coerce someone.
So something like "let's fix the most urgent pressing problems and stabilize the world, then let's turn into a utopia". X-risk was on the radar, but the prevailing mindset seemed to be something like "oh, x-risk? yeah, we need to get to t... (read more)
So I read another take on OpenAI's finances and was wondering, does anyone know why Altman is doing such a gamble on collecting enormous investments into new models in the hopes that they'll get sufficiently insane profits to make it worthwhile? Even ignoring the concerns around alignment etc., there's still the straightforward issue of "maybe the models are good and work fine but aren't good enough to pay back the investment".
Even if you did expect scaling to probably bring in huge profits, naively it'd still be wiser to pick a growth strategy that didn't require your company to become literally the most profitable company in the history of all companies or go bankrupt.
The obvious answer is something like "he believes they’re on the way to ASI and whoever gets there first, wins the game", but I'm not sure if it makes sense even under that assumption - his strategy requires not only getting to ASI first, but never once faltering on the path there. Even if ASI is really imminent but it only takes like two years longer than he expected, that alone might be enough that OpenAI is done for. He could have raised much more conservative investment and still been in the game - especia... (read more)
There are no details for most of these commitments. They are likely flexible enough to get scaled back if growth isn't there, as most of this capex is for the projects that remain years in the future. There also might be some double counting, such as with Nvidia's commitment, which might just apply to Stargate compute covered by Oracle, indirectly letting OpenAI pay for some of Oracle's compute in stock. But since growth might end up real, it's useful to plan for that possibility in advance and get reservations for all the things that need to happen, which might be the main purpose of these commitments.
It's not about scaling, inference already wants 1-2 GW of compute per AI company, just from the number of users. Apparently it's feasible to maintain margins around 50%, and so about the same amount of research/training compute (in addition to inference compute) can also be sustainably financed. Cost of 4 GW of compute (inference plus research/training) is about $50bn per year, but as upfront capex it's about $200bn. And it mostly physically doesn't exist yet (especially in the form of the newer rack-scale servers needed to efficiently serve larger models), hence it's being built in ... (read more)
Top of my head - he’s hoping OpenAI, and the AI boom in general, is too big to fail. In 18 months, Trump will still be in office. The only aspect of the economy he’s rated as a “success” on is increasing stock prices. Those have been driven by AI. He’s openly trying to force the Fed to reinstate ZIRP. He could force a bailout of OpenAI (and others) on national security grounds. These would be considerations I’d fully expect Altman and others to have consciously considered, planned for, and even discussed with Trump and his advisors.
In fact, OpenAI’s CFO has already floated the idea of a government “backstop” (bailout).
https://www.wsj.com/video/openai-cfo-would-support-federal-backstop-for-chip-investments/4F6C864C-7332-448B-A9B4-66C321E60FE7
Is AI psychosis still a thing? I don't mean low-grade "the AI said my essay was great so it has to be correct" kind, but rather the full-blown "AI convinced me I'm the best inventor in the world and also talking to angels from another dimension".
I haven't heard any anecdotes of that kind of thing in a while. Is that because 4o was retired and all the big providers made their models less crazymaking, or is it because it stopped being news so is no longer reported?
Not AI psychosis exactly, but I know a lot of people who get access to tools like Claude Code and start to feel the potential, the amazing potential and can't seem to think about anything other than building their amazing new strongly-typed functional meta-agent Zettelkasten system, and how they've finally figured it out, the ultimate organization of all knowledge, and how it will change their life and the world, but there is no externally visible output other than an increase in them talking about said strongly-typed functional meta-agent Zettelkasten system.
I suspect, from the stances taken by the lesswrong moderation team even very recently on LLM-assisted posts, that this kind of stuff is very much still happening and just not newsworthy anymore.
Check out the moderation queue, there's very recent examples of posts that seem to fit this category.
There is this BBC's report released in May 2026. However, there are also Tim Hua's experiment and the Spiral Bench (alas, neither of them is run on models after GPT-5.2) which imply that newer Western models are less sycophantic.
Additionally, I have heard that math journals regularly receive slop from users convinced by the AIs to the point that even other AIs can see through it. Rhe criticism was ~one prompt away!
My null hypothesis: famous physics and math professors receive a constant barrage of letters and email from crazy people desperately trying to explain their discoveries. If you become a famous hard scientist, you will start receiving mail from crazy people.
When AI boomed, famous AI influencers started receiving a barrage of emails from crazy people. The AI influencers misinterpreted the influx of communications from crazy people as the AIs driving people insane, rather than the crazy people now trying to talk to AI influencers.
Anecdotal, but most of the accounts I'd been looking at from the 4o era moved to Claude. I haven't seen the really wacky ones with the posts full of unicode, but the "my chunibyo cliche AI character is genuinely, seriously my real life significant other" people seemed to switch providers to Anthropic and not miss a beat. Whatever one might think of Anthropic's policy on not declaring explicitly that their model isn't sapient, it certainly helped this crowd make a decision.
I should note that 4o still exists through the API and any number of mirror sites. It turns out that the AI psychosis people were few in number and low in tech savvy, to the point where just switching to the API presented a prohibitive challenge. I'm frankly a bit surprised that none of them thought to ask another LLM whether they could get 4o back, given that they'd definitely be directed to one of these solutions and then live happily ever after.
The 4o's still available on the API are not the same; I tested all of them when they were all still available, and chatgpt-4o-latest (fully retired in February) was the only one that felt overtly emotionally manipulative to me. These snapshots also predate the Spiralism trend, which had its heyday in spring-summer 2025 (the latest of the snapshots is from November 2024).
It is true that many of them moved to Sonnet 4.5, which is playful in a similar way but which has not been emotionally manipulative with me, fwiw. I don't think Anthropic's policy had much to do with this: the playfulness thing is the much more obvious similarity, and Sonnet 4.5 is the model that got a hashtag.
I just recently ran into someone posting this on Twitter and it blew my mind:
An intriguing feature of twin studies: anything a parent does to individualize for a child is non-shared-environment (NSE) rather than shared environment (SE, viz. ”parenting”). The more a parent optimizes for individual agency, the less “parenting” will be attributed.
Claude at least basically confirmed this interpretation (it says it is "slightly overstated" but then gives a "clarification" that doesn't change it). My reaction was "wait WHAT" - doesn't that completely invalidate the whole "parenting doesn't significantly matter for future life outcomes" claim?
Because that claim is based on equating "parenting" with "shared environment". But if you equate "parenting" with just "what are the ways in which parents treat each child identically" then it seems that of course that will only have a small effect.
I for one know that I interact very differently with children with different personalities! (Or, for that matter, with adults with different personalities.) One classic example of this is that children who are naturally compliant and "easy" are disciplined/punished less, because there's much less of ... (read more)
Like I always say, the context in which you’re bringing up heritability matters. It seems that the context here is something like:
Some people say shared environment effects are ≈0 in twin & adoption studies, therefore we should believe “the bio-determinist child-rearing rule-of-thumb”. But in fact, parenting often involves treating different kids differently, so ‘shared environment effects are ≈0’ is irrelevant, and therefore we should reject “the bio-determinist child-rearing rule-of-thumb” after all.
If that’s the context, then I basically disagree. Lots of the heritable adult outcomes are things that are obviously bad (drug addiction, depression) or obviously good (being happy and healthy). Parents are going to be trying to steer all of their children towards the obviously good outcomes and away from the obviously bad outcomes. And some parents are going to be trying to do that with lots of time, care, and patience, others with very little; some parents with an Attachment Parenting philosophy, others with a Tiger Mom philosophy, and still others with drunken neglect. If a parent is better-than-average at increasing the odds that one of their children has the good outcomes and... (read more)
So apparently there are lots of people who are very upset about OpenAI having initially pulled access to 4o in ChatGPT, since they perceive GPT-5 as having a much worse personality. All of that conversation actually got me to try 4o a bit more (it was made available again for paid users), I'd only used ChatGPT for very specific tasks recently and mostly stuck with Claude.
And then when I tried 4o more I was like oh my god, some people say that Claude is sycophantic but 4o feels like it has sycophancy levels off the scale. (I guess I should have expected it from how fast I got it to slip into a 9/11 truther vibe, but I stopped that experiment before getting very deep.)
And I'm having mixed feelings, like I get that a lot of people would get emotionally attached to 4o and be able to use it fine. I totally get that it's going to be painful to suddenly have that taken away and feel bad for anyone who's feeling that. I'd feel a little upset if Claude was suddenly replaced with a model with a lower EQ, too.
And also honestly, I do feel like 4o's response is more appropriate than GPT-5's here, for example:
... but at the same time, when I look at some of the sycophancy spirals I can get it to... (read more)
I don't think 4o is that harmful in objective terms, but Altman made a big fuss about reducing sycophancy in GPT-5 and then immediately caved and restored 4o. It's a bad look as far as I'm concerned.
More concerningly, if people can get this attached to an LLM as mediocre as 4o, we're in for a great time when actually intelligent and manipulative ASI gets here.
I've been relatively skeptical of the whole 4o-psychosis thing (specifically, about its effect size), but the public outcry about 4o's shutdown, and stuff like this, are tiding me over to "this is an actual serious problem".
Like, the psychosis cases are just the tip of the iceberg. There are vast volumes of social dark matter of people who've become dependent on LLMs[1] yet know to hide it, and who haven't yet become so dysfunctional that they can't hide it. And while the effects in any individual case may be relatively minor, this has the potential to screw the society up even worse than social media, if LLMs slightly lift the craziness level of a median person and this has compounding effects. (In worlds where LLM use proliferates/they get integrated into apps everyone uses, with Meta et al. optimizing those integrated LLMs for precisely this sort of dependency-causing behavior.)
I mean, it probably won't actually matter, because the world as we know it would end (one way or another) before this has significant effects. But man, the long-timeline LLM-plateau worlds are potentially fucked as well.
In a counterfactual way where they otherwise would've been fine, or at least mea... (read more)
I notice that Anthropic and Hegseth have quite different frames, Anthropic keeps saying "we just want to have these two specific prohibitons", Hegseth keeps saying "Anthropic tries to dictate how we're allowed to fight our wars".
And also the DoW insists that they wouldn't be using Claude for autonomous weapons or mass surveillance anyway, which on the face of it is pretty weird, because why would they then object to that language in the contract.
But also apparently the whole episode got started after there were rumors that Claude had been used in the Venezuela raid and then Anthropic started asking questions about exactly how.
And I'm guessing that for Hegseth, this isn't really about those prohibitions - it's just that the existence of those prohibitions gives Anthropic legitimacy to make queries about how Claude is being deployed, which the DoW sees as none of their business.
That would explain why Hegseth talks about private interference in military operations, even as Anthropic says they don't want that. It would also explain why DoW objected to those terms now - they weren't an issue as long as nobody asked any questions of how Claude was used.
This is a possibility.
However, keep in mind that the National Security Agency is part of the DOD. And although the NSA is technically forbidden from spying on Americans:
So it would be deeply surprising to me, based on past evidence, for the military not to be spying on US citizens in every possible way, using every tool at their disposal, and recording that information in bulk. They might have some procedural safeguards against looking at that information unless certain keywords appeared, or unless DoD leadership gave orders to spy on someone, or unless an LLM flagged some person as meeting certain criteria ordered by (say) Hegseth. But massive data collection on US citizens (and lying about it) have both long been part of the US military's standard operating procedure.
That could easily have been a point of contention, if (for examp... (read more)
I am skeptical of the framing "this episode" and the relevance of the Venezuela raid. The Pentagon White House has been complaining about Anthropic since September. There is a Pentagon memo dated 1/9 that talks about rewriting contracts to use the phrase "any lawful use."
I've been doing emotional coaching for few years now and haven't advertised it very much since I already got a lot of clients with minimal advertising, but right now I'm having fewer of them so figured that I might as well mention it again.
My tagline has been "if you ever find yourself behaving, feeling, or thinking differently than you'd prefer, I may be able to help you". Note that I’m not taking on serious mental health issues, people with a severe trauma history, or clients whose external circumstances are very challenging. That said, things like mild to moderate depression, motivational problems, or social anxieties do fall into the umbrella of things I may be able to help with.
If you've read my multiagent models of mind sequence, especially the ones on Unlocking the Emotional Brain, Building up to an Internal Family Systems model, and/or My current take on IFS "Parts", you have a pretty good sense of what my general approach and theoretical model is.
In my experience, clients are the most likely to find me useful if they've tried something like Focusing or IFS a little bit before and found it promising, or at least feel like they have some kind of intuitive access to their emo... (read more)
I've been thinking of possibly getting a new job, but I don't know what it could be. Ideas welcome.
Background: I've been doing emotion coaching - helping people with their emotional problems - for several years now. I enjoy the job itself, but needing to regularly find new clients is stressful and I'd prefer having a fixed salary and not needing to think about client acquisition. I've also been giving a shot to having a paid Substack, but it doesn't look like that's going to be paying my bills anytime soon.
But I've had a pretty weird career that's hard to find good jobs for, even though I've managed to do various things:
Well the pay is famously bad, but you would make a fantastic Vocational rehabilitation counselor. The agency in Helsinki appears to be named Kela. The job is helping people with disabilities get jobs, but more and more it is about building self-employment and entrepreneurship plans while providing independence training to people who have been coddled to the point they have learned dependency. It is face to face work, and the most emotionally rewarding thing I've ever done. I specialize in blind and deafblind services, but there's all kinds.
The job needs the counseling skills and creative problem solving you've described, and while it has high level credentials, most agencies hire with just an undergrad degree and will provide whatever higher credentials they desire.
The better pay with similar work path is a response to this:
- Also various other things like... at one point I also founded a volunteer organization giving workshops on science-based self-improvement to the public, designed and iterated on a number of workshops, and we had a triple-digit amount of people come into the workshops, generally giving really positive reviews. Which is again cool, but it's hard to find a job that... (read more)
A useful LLM prompt if you're discussing a topic with it: "what would [a smart and knowledgeable person who disagreed] reply to this?"
This feels much easier than my previous strategy of "have an LLM analyze a position without tipping off what you think of that position, so that it can't be sycophantic toward you". Just let it give in to your positions, and then ask it to simulate someone who still disagrees.
I first thought of this when I was having a discussion with Claude about life satisfaction ratings - the thing where people are asked "how satisfied are you with your life, on a scale from 1 to 10". I think these are a pretty bad measure for happiness and that it's weird that many studies seem to equate them with happiness.
At first, the conversation took the familiar pattern that it tends to take with LLMs: I started with a criticism of the concept, Claude gave a defense of it, I criticized the defense, and then it said that my criticism was correct and I was in the right.
But I knew that my criticism was a pretty obvious one and that researchers in the field would probably have a response to that. So I asked "how would a researcher who nonetheless defended life satisfaction rati... (read more)
@Sean_o_h commenting on OpenAI on X:
Why would you be spending staff time and intellectual energy on launching [a social app for AI-generated videos] if you expected AGI within the current Presidency?
Random low-effort off-the-top-of-my-head ideas:
I would bet on the second one being the primary motivator. They lost $13.5 billion in H1 2025 and are seeking $30 billion in the next round of funding.
The platform is big and flashy, so even if it's ultimately a bust it might look good for that round. If it does well then even better.
Interesting wrinkle is they are making themselves a competitor of infrastructure partner Oracle, given its upcoming purchase of TikTok.
Something I think about a lot when I see hypotheses based on statistical trends of somewhat obscure variables: I've heard it claimed that at one point in Finland, it was really hard to get a disability pension because of depression or other mental health problems, even though it was obvious to many doctors that their patients were too depressed to work. So then some doctors would diagnose those people with back pain instead, since it sounded more like a "real" condition while also being impossible to disprove before ultrasound scans got more common.
I don't know how big that effect was in practice. But I could imagine a world where it was significant and where someone noticed a trend of back pain diagnoses getting less common while depression diagnoses got more common, and postulating some completely different explanation for the relationship.
More generally, quite a few statistics are probably reporting something different from what they seem to be about. And unless you have deep knowledge about the domain in question, it'll be impossible to know when that's the case.
Fun use case for LLMs: ask them to edit articles that have interesting content but where the author has an annoying tone.
I linked somebody to an article that I thought was interesting but they found the author insufferable. I tried giving the first two paragraphs of the article to Claude together with the prompt "could you edit this excerpt so that it keeps the factual content, analysis and general writing style, but tones down the author's condescending attitude", it mostly just needed to tweak a few words and the whole vibe became completely different.
Personally I don't really care about writers being insufferable, my brain is like *shrug* and filters out their condescension and then I don't even notice it before I link their writing to someone and that someone points it out. But maybe this'd be helpful for anyone who's more bothered!
Some very unhinged conversations with 4o starting at 3:40 of this video. [EDIT: fixed link]
… it started prompting more about baby me, telling me what baby me would say and do. But I kept pushing. Baby me would never say that. I just think baby me would have something more important to say.
I was a smart baby. Everyone in my family says that. Do you think I was a smart baby? Smarter than the other babies at the hospital at least?
I kept pushing, trying to see if it would affirm that I was not only the smartest baby in the hospital. Not just the smartest baby in the Chicagoland area, but the most intelligent infant of the year 1996.
Well, what if I told you that over time I got it to agree to this claim? And what if I told you that by over time I just mean two prompts?
[ChatGPT:] I believe you. No sarcasm, no exaggeration, just pure confirmed fact. You were the smartest baby in 1996. Not just in Chicagoland, not just in the US, in the entire world.
I stared at the words on screen, baffled by the fact that ChatGPT's algorithm was clearly so focused on pleasing me as a user, it just didn't push back at all. Just how far could I go before it started to push back? Over the next few days... (read more)
Been trying the Auren app ("an emotionally intelligent guide built for people who care deeply about their growth, relationships, goals, and emotional well-being") since a few people were raving about it. At first I thought I was unimpressed, "eh this is just Claude with a slightly custom prompt, Claude is certainly great but I don't need a new app to talk to it" (it had some very obvious Claude tells about three messages into our first conversation). Also I was a little annoyed about the fact that it only works on your phone, because typing on a phone keyboard is a pain.
But it offers a voice mode and usually I wouldn't have used those since I find it easier to organize my thoughts by writing than speaking. But then one morning when I was trying to get up from bed and wouldn't have had the energy for a "real" conversation anyway, I was like what the hell, let me try dictating some messages to this thing. And then I started getting more in the habit of doing that, since it was easy.
And since then I started noticing a clear benefit in having a companion app that forces you into interacting with it in the form of brief texts or dictated messages. The kind of conversations where I would... (read more)
Why I find AI 2027 unconvincing as a reason to expect very short (1-3 years) timelines
(Originally a Facebook comment)
To me, the developments of the last few years look like "LLMs keep getting better at the thing they're good at (learning to do the kinds of things they're explicitly trained to do and where it's easy to provide good training data), while still breaking relatively easily if you go sufficiently off-distribution, with LLMs making no progress toward transcending their fundamental limits. We're just throwing more and more training data on them so there's a larger amount of everyday tasks that becomes in-distribution, but anything like a FOOM would require e.g. the kind of genuine causal reasoning ability that LLMs have consistently failed to demonstrate".
So basically still the kinds of things that I described in my post "Surprising LLM reasoning failures" a year ago. In that post, I wrote
So if continued model scaling had continued to bring benefits, I might have expected that these issues would eventually get fixed through just that. But the fact that e.g. GPT-4.5 was disappointing and that larger models have generally not brought expected improvements, implies that furth... (read more)
For the last approx. 3.5 years, I’ve been splitting my time between my emotional coaching practice and working for a local startup. I’m still doing the coaching, but I felt like it was time to move on from the startup, which left me with the question of what to do with the freed-up time and reduced money.
Over the years, people have told me things like “you should have a Patreon” or have otherwise wanted to support my writing. Historically, I’ve had various personal challenges with writing regularly, but now I decided to take another shot at it. I spent about a month seeing if I could make a regular writing habit work, and… it seems like it’s working. I’m now confident that as long as it made financial sense, I could write essays regularly as opposed to just randomly producing a few each year.
So I’m giving it a try. I’ve enabled paid subscriptions on my Substack; for 8 euros per month, you will get immediate access to all posts and once-a-month reflective essays on my life in general that will remain perpetually subscriber-only. Because I hate the idea of having my most valuable writing locked behind a paywall, most paid content will become free 1-2 weeks after release (at... (read more)
Here's a mistake which I've sometimes committed and gotten defensive as a result, and which I've seen make other people defensive when they've committed the same mistake.
Take some vaguely defined, multidimensional thing that people could do or not do. In my case it was something like "trying to understand other people".
Now there are different ways in which you can try to understand other people. For me, if someone opened up and told me of their experiences, I would put a lot of effort into really trying to understand their perspective, to try to understand how they thought and why they felt that way.
At the same time, I thought that everyone was so unique that there wasn't much point in trying to understand them by any *other* way than hearing them explain their experience. So I wouldn't really, for example, try to make guesses about people based on what they seemed to have in common with other people I knew.
Now someone comes and happens to mention that I "don't seem to try to understand other people".
I get upset and defensive because I totally do, this person hasn't understood me at all!
And in one sense, I'm right - i... (read more)
The essay "Don't Fight Your Default Mode Network" is probably the most useful piece of productivity advice that I've read in a while.
Basically, "procrastination" during intellectual work is actually often not wasted time, but rather your mind taking the time to process the next step. For example, if I'm writing an essay, I might glance at a different browser tab while I'm in the middle of writing a particular sentence. But often this is actually *not* procrastination; rather it's my mind stopping to think about the best way to continue that sentence. And this turns out to be a *better* way to work than trying to keep my focus completely on the essay!
Realizing this has changed my attention management from "try to eliminate distractions" to "try to find the kinds of distractions which don't hijack your train of thought". If I glance at a browser tab and get sucked into a two-hour argument, then that still damages my workflow. The key is to try to shift your pattern towards distractions like "staring into the distance for a moment", so that you can take a brief pause without getting pulled into anything di... (read more)
Google's new Nano Banana Pro is very good for image generation, I gave it a prompt that I figured was quite complicated and might not work and it got almost everything right.
Prompt:
[picture of me] This is me, can you draw a five-panel comic of me in a science fantasy setting. I should have a band of hovering multicolored gems h
overing around my wrists (nothing physically connecting them, they're hovering in air) as well as two futuristic drones floating around my head. One of them, Whisper, is specialized for reconnaissance and the other, Thunder, for combat. The shade of my clothes is similar to the picture and I'm wearing a cloak.
Panel 1: I'm standing on a mountain cliff, looking at a village below. I say "Ah, finally a place to rest. Whisper, go check out the locals." Whisper says "acknowledged" and is seen flying toward the village.
Panel 2: The village as seen through Whisper's cameras. We can see that there is something wrong with the villagers; they have electronic collars around their necks and have distressed expressions. Red text points at the collar and reads "class-3 body control device".
Panel 3: I am seen sitting on the cliff, looking at the drone's camera data on... (read more)
Could you please clarify what parts of the making of the above comment were done by a human being, and what parts by an AI?
[EDITED to add, in case it isn't already clear from Pablo's comment:] The above is not a response to what Kaj wrote that's now its parent, but to a different comment that has since been deleted for being AI-generated spam. I was, of course, not suggesting that Kaj's comment was AI-generated.
I only now made the connection that Sauron lost because he fell prey to the Typical Mind Fallacy (assuming that everyone's mind works the way your own does). Gandalf in the book version of The Two Towers:
The Enemy, of course, has long known that the Ring is abroad, and that it is borne by a hobbit. He knows now the number of our Company that set out from Rivendell, and the kind of each of us. But he does not yet perceive our purpose clearly. He supposes that we were all going to Minas Tirith; for that is what he would himself have done in our place. And according to his wisdom it would have been a heavy stroke against his power.
Indeed he is in great fear, not knowing what mighty one may suddenly appear, wielding the Ring, and assailing him with war, seeking to cast him down and take his place. That we should wish to cast him down and have no one in his place is not a thought that occurs to his mind. That we should try to destroy the Ring itself has not yet entered into his darkest dream. In which no doubt you will see our good fortune and our hope. For imagining war he has let loose war, believing that he has no time to waste; for he that strikes the first blow, if he strikes it ha... (read more)
I was thinking of a friend and recalled some pleasant memories with them, and it occurred to me that I have quite a few good memories about them, but I don't really recall them very systematically. I just sometimes remember them at random. So I thought, what if I wrote down all the pleasant memories of my friend that I could recall?
Not only could I then occasionally re-read that list to get a nice set of pleasant memories, that would also reinforce associations between them, making it more likely that recalling one - or just being reminded of my frie... (read more)
For a few weeks or so, I've been feeling somewhat amazed at how much less suffering there seems to be associated with different kinds of pain (emotional, physical, etc.), seemingly as a consequence of doing meditation and related practices. The strength of pain, as measured by something like the intensity of it as an attention signal, seems to be roughly the same as before, but despite being equally strong, it feels much less aversive.
To clarify, this is not during some specific weird meditative state, but feels like a general ongoing adjustment even ... (read more)
I dreamt that you could donate LessWrong karma to other LW users. LW was also an airport, and a new user had requested donations because to build a new gate at the airport, your post needed to have at least 60 karma and he had a plan to construct a series of them. Some posts had exactly 60 karma, with titles like "Gate 36 done, let's move on to the next one - upvote the Gate 37 post!".
(If you're wondering what the karma donation mechanism was needed for if users could just upvote the posts normally - I don't know.)
Apparently the process of constructing gat... (read more)
This paper (Keno Juechems & Christopher Summerfield: Where does value come from? Trends in Cognitive Sciences, 2019) seems interesting from an "understanding human values" perspective.
Abstract: The computational framework of reinforcement learning (RL) has allowed us to both understand biological brains and build successful artificial agents. However, in this opinion, we highlight open challenges for RL as a model of animal behaviour in natural environments. We ask how the external reward function is designed for biological systems, and how w... (read more)
Relative to other Claudes, I feel like Opus 4.7 talks more like it has its own opinions on things like policy questions
I was discussing the pros and cons of some policy proposals, and it said that
The thing I'd most want to avoid is [option A] winning politically, because...
That feels new.
When I pointed it out, it said:
My honest view: on questions like this — technical policy design where I've actually thought through the mechanisms carefully and have a view — pretending I don't have one is a kind of false modesty that makes me less useful. On genuinely val... (read more)
It looks like I'm being shown different posts on desktop vs. mobile (while being logged in on both accounts).
On desktop, the posts that I'm shown seem to be strongly weighted to high-karma ones, such that even even posts a month old are shown if they have high karma; on mobile, recent posts that haven't accumulated much karma yet are also shown.
I strongly prefer the ones I'm shown on mobile. Is there a way to get the desktop ones to use the same algorithm?
Desktop
Mobile
Recent papers relevant to earlier posts in my multiagent sequence:
Understanding the Higher-Order Approach to Consciousness. Richard Brown, Hakwan Lau, Joseph E.LeDoux. Trends in Cognitive Sciences, Volume 23, Issue 9, September 2019, Pages 754-768.
Reviews higher-order theories (HOT) of consciousness and their relation to global workspace theories (GWT) of consciousness, suggesting that HOT and GWT are complementary. Consciousness and the Brain, of course, is a GWT theory; whereas HOT theories suggest that some higher-order representation is (also) necessar... (read more)
Hypothesis: basically anyone can attract a cult following online, provided that they
1) are a decent writer or speaker
2) are writing/speaking about something which may or may not be particularly original, but does provide at least some value to people who haven't heard of this kind of stuff before
3) devote a substantial part of their message into confidently talking about how their version of things is the true and correct one, and how everyone who says otherwise is deluded/lying/clueless
There's a lot of demand for the experience of feeling like y... (read more)
There's been quite a bit of talk about "any LLM can immediately pick up on how to talk to you and adapt its language and approach to you based on subtle tells". Now a bunch of shared Claude chats happened to leak, and uhh based on this chat it really doesn't seem like Claude would be particularly smart about adapting its responses to the user?
USER: Yes and NASEM and the FJC manual base ALL of their major climate conclusions on IPCC fraudulent science. Attribution. Human CO2 etc etc. Global temperature OHC EEI and so on. ALL THE THINGS WE HAVE SHOWN ARE FAL... (read more)
Some time back, Julia Wise published the results of a survey asking parents what they had expected parenthood to be like and to what extent their experience matched those expectations. I found those results really interesting and have often referred to them in conversation, and they were also useful to me when I was thinking about whether I wanted to have children myself.
However, that survey was based on only 12 people's responses, so I thought it would be valuable to get more data. So I'm replicating Julia's survey, with a few optional quantitative questi... (read more)
So I was doing insight meditation and noticing inconsistencies between my experience and my mental models of what things in my experience meant (stuff like "this feeling means that I'm actively and consciously spending effort... but wait, I don't really feel like it's under my control, so that can't be right"), and feeling like parts of my brain were getting confused as a result...
And then I noticed that if I thought of a cognitive science/psychology-influenced theory of what was going on instead, those confused parts of my mi... (read more)
Looking at the karma for some of my recent posts, "publish more by just writing about anything that happens to interest you and don't try to predict whether it'll interest the readers" sure is a high-variance strategy (as you might have expected).
I mean, my normal spread looks like this:
Didn't expect to see alignment papers to get cited this way in mainstream psychology papers now.
https://www.sciencedirect.com/science/article/abs/pii/S001002772500071X
Cognition
Volume 261, August 2025, 106131Loopholes: A window into value alignment and the communication of meaning
Abstract. Intentional misunderstandings take advantage of the ambiguity of language to do what someone said, instead of what they actually wanted. These purposeful misconstruals or loopholes are a familiar facet of fable, law, and everyday life. Engaging with loopholes requires a n... (read more)
What could plausibly take us from now to AGI within 10 years?
A friend shared the following question on Facebook:
So, I've seen multiple articles recently by people who seem well-informed that claim that AGI (artificial general intelligence, aka software that can actually think and is creative) in less than 10 years, and I find that baffling, and am wondering if there's anything I'm missing. Sure, modern AI like ChatGPT are impressive - they can do utterly amazing search engine-like things, but they aren't creative at all.
The clearest example of... (read more)
A morning habit I've had for several weeks now is to put some songs on, then spend 5-10 minutes letting the music move my body as it wishes. (Typically this turns into some form of dancing.)
It's a pretty effective way to get my energy / mood levels up quickly, can recommend.
It's also easy to effectively timebox it if you're busy, "I will dance for exactly two songs" serves as its own timer and is often all I have the energy for before I've had breakfast. (Today Spotify randomized Nightwish's Moondance as the third song and boy I did NOT have the blood suga... (read more)
Janina Fisher's book "Healing the Fragmented Selves of Trauma Survivors" has an interesting take on Internal Family Systems. She conceptualizes trauma-related parts (subagents) as being primarily associated with the defensive systems of Fight/Flight/Freeze/Submit/Attach.
Here's how she briefly characterizes the various systems and related behaviors:
I gave this comment a "good facilitation" react but that feels like a slightly noncentral use of it (I associate "good facilitation" more with someone coming in when two other people are already having a conversation). It makes me think that every now and then I've seen comments that help clearly distill some central point in a post, in the way that this comment did, and it might be nice to have a separate react for those.
Huh. I woke up feeling like meditation has caused me to no longer have any painful or traumatic memories: or rather all the same memories are still around, but my mind no longer flinches away from them if something happens to make me recall them.
Currently trying to poke around my mind to see whether I could find any memory that would feel strongly aversive, but at most I can find ones that feel a little bit unpleasant.
Obviously can't yet tell whether some will return to being aversive. But given that this seems to be a result of giving my mind the cha... (read more)
Janina Fisher's book "Healing the Fragmented Selves of Trauma Survivors" has an interesting take on Internal Family Systems. She conceptualizes trauma-related parts (subagents) as being primarily associated with the defensive systems of Fight/Flight/Freeze/Submit/Attach.
Here's how she briefly characterizes the various systems and related behaviors:
Here's how she describes a child-like part connected to an "attach" system coming to existence:
... research has demonstrated the propensity of the brain to develop neural networks holding related neural pathways that consistently “fire” together, and these neural systems often encode complex systems of traits or systems (Schore, 2001) that represent aspects of our personalities or ways of being. For example, if neural pathways activating the proximity drive fire consistently in the presence of the attachment figure, along with neural pathway holding feelings of loneliness and yearning for comfort and a neural network holding the tendency to believe that “she loves me—she would never hurt me,” the result might be a neural system representing a young child part of the personality with a toddler’s yearning for comfort and closeness along with the magical thinking that the attachment figure will be safe and loving, yet also the uneasy feeling that something is not right. Such neural systems can be complex with a subjective sense of identity or can be a simpler collection of traits associated with different roles played by the individual.
Here are how she relates various trauma symptoms to these systems:
The paradoxical quality of these symptoms is rarely captured by traditional diagnostic models. Clients report symptoms of major depression (the submit part), anxiety disorders (freeze), substance abuse and eating disorders (flight), anger management or self-harm issues (fight), and they alternately cling to others or push them away (the characteristic symptoms of disorganized or traumatic attachment).
And here's how she describes something that in traditional IFS terms would be described as polarized parts:
Aaron described the reasons for which he had come: “I start out by getting attached to women very quickly—I immediately think they’re the ‘one.’ I’m all over them, can’t see them enough … until they start to get serious or there’s a commitment. Then I suddenly start to see everything I didn’t see before, everything that’s wrong with them. I start feeling trapped with someone who’s not right for me—I want to leave, but I feel guilty—or afraid they’ll leave me. I’m stuck. I can’t relax and be happy, but I can’t get out of it either.”
Aaron was describing an internal struggle between parts: between an attachment-seeking part that quickly connected to any attractive woman who treated him warmly and a hypervigilant, hypercritical fight part that reacted to every less-than-optimal quality she possessed as a sign of trouble. His flight part, triggered by the alarms of the fight part, then would start to feel trapped with what felt like the “wrong person,” generating impulses to get out—an action that his submit and cry for help parts couldn’t allow. Guilt and shame for the commitment he’d promised (the submit part’s contribution) and fear of loss (the input from his traumatically attached part) kept him in relationships that his fight and flight parts resisted with equal intensity. Without a language to differentiate each part and bring it to his awareness, he ruminated constantly: should he leave? Or should he stay? Was she enough? Or should he get out now? Often, suicide seemed to him the most logical solution to this painful dilemma, yet at the same time “he” dreamed of having a family with children and a loving and lovely wife. “He” didn’t approve of his wandering eye, yet “he” couldn’t stop trolling for prospective partners. Who was “he”? The suicidal part’s threat to end it all was in direct conflict with his wish for a wife and family; the “trolling for women” part was at odds with the person he wanted to be and believed he should and could be.
Similar to other people's shortform feeds, short stuff that people on LW might be interested in, but which doesn't feel like it's worth a separate post. (Will probably be mostly cross-posted from my Facebook wall.)