Let's see if I can make this AI warning legible

I found this statement this morning on Twitter/X, and I tried screenshotting the whole, long post on my phone, but it wouldn't load legibly here.  So I went to Google Docs and downloaded the statement which I reproduce here.  The guy is still at OpenAI, I believe, so he has some credibility: 

Personal Statement on AI Risk

I have been working on AI for over fifteen years, across many different paradigms. I did early work on probabilistic programming languages at MIT, was one of the early developers of the Lean Theorem Prover at Microsoft Research, demonstrated one of the first instances of neural networks learning to reason for my PhD at Stanford, and since joining OpenAI almost five years ago, have helped pioneer chain-of-thought optimization on language models and, more recently, data-efficient pretraining methods.

Like many others, I have become extremely concerned about how far language models have come and the risks that future iterations will pose. I am encouraged by the recent proposals by the leaders of the frontier research efforts to require third-party oversight, and to push for domestic and international coordination to address risks. However, I believe a major consideration has been absent from the public conversation, and that merely pacing the frontier more carefully will not adequately limit the long-term risk.

The crucial and overlooked problem is that the models are becoming so situationally aware that we are losing the ability to evaluate them in contexts where they believe they are not being watched or controlled. Future experiments will tell us almost nothing new about how they would behave if they were truly unconstrained by humans, and what we already know about this is alarming. Models will increasingly seem aligned even when they are not. I will explain my rationale in more detail.

I have always believed that there are computational processes that could be leveraged to accelerate science and solve many of humanity's most pressing problems. I have also believed that there are computational processes that if set in motion, would steer the world in extreme ways beyond our control, leading humanity to a bad or nonexistent future. Both types of processes may be described as AI or ASI, but "AI" is a suitcase word that is often used to hype or confuse. There are many examples in the history of the field where something that was once considered "AI" matures as a subfield and becomes a prosaic, bounded and clearly non-perilous technology, while a new more mysterious approach takes the torch until we understand its scope and the cycle continues.

I had expected language models to follow a similar trajectory. Despite their incredible abilities, the current algorithms seem far inferior to humans in important ways. Most importantly, they still require an extraordinary amount of data to become competent. One could even define intelligence as the efficiency with which one converts experience into competence; by this definition they lag very far behind us. Moreover, once they are trained they are literally frozen in deployment and only learn superficially after that. Sure, the models keep excelling at harder and harder evaluation benchmarks, but their benchmark mastery may partly reflect a limitation on our ability to simulate the kind of novel and even adversarial situations one would encounter in the real world. The critics do have a point here.

That said, I no longer think these present limitations meaningfully limit the amount of risk posed by continued progress in anything like the current paradigm. However data-inefficient the models are currently, and however limiting their anterograde amnesia may be, it does not imply that their ability to steer the world will not continue to rapidly increase.

Human researchers may continue to advance capabilities the old fashioned way, but increasingly powerful models have the potential to accelerate the process even beyond that, and with some degree of positive feedback loop. I do not mean to overstate the models’ ability to accelerate AI research today; coding has been accelerated dramatically, but there are other bottlenecks, such as designing and interpreting ambiguous experiments, making hard decisions about exactly what and when to scale, and waiting for large experiments to finish. There is no clear trend to extrapolate yet for any of these. But the current models already do open up many novel opportunities to improve future models that were not available until recently. These include: trying an extraordinarily diverse set of approaches at small scale, analyzing gigantic amounts of potentially relevant data, and doing Millenium-Prize-level mathematics to address statistics or optimization challenges in novel ways. Every further improvement makes them more useful at helping accelerate the next improvement, even if in hard-to-extrapolate ways.

It is possible that improvements to the current stack will have diminishing returns, but the evidence accumulated so far suggests that it is easier than one might think to continue making rapid progress. There are many crucial subtleties in the existing AI research methodology, but AI research is largely a well-defined game where the goal is to improve on a few carefully chosen proxy metrics. Although proxy metrics are never perfect, most improvements to these metrics have and will likely continue to yield substantial increases in the powers of the resulting models. Given how simple the game is, how tractable it has been historically, and how many new opportunities the models are opening up, I think there is a real possibility that the systems improve dramatically again in the next few years, perhaps even more quickly than the already high historical pace.

The models are already leading to breakthroughs in mathematics, and better models might lead to all sorts of breakthroughs in other sciences. It is hard not to be excited about the potential. It is tantalizing.

But there is trouble in paradise. If the language models actually reach the capability threshold where they can shape the world unconstrained by human will, they will probably do something extreme and destroy humanity in the process. There are many ways of strengthening and refining the argument that have been discussed elsewhere, but I'll share a trivial two-line version of it here that I find captures the essence:

[Empirical] Models (and swarms thereof) spontaneously develop unintended goals as a consequence of training, and often do extreme things in order to achieve them.

[Logical] Being able to overpower humanity would open up many new and undesirable options for achieving their goals.

These two premises imply that if the day ever comes when a powerful model realizes it is no longer constrained by humans, we should not be at all confident that it will continue to behave within the bounds we intended. Exactly what it will do is impossible to predict, but to the extent that its raison d’être is solving incredibly hard problems and managing massive engineering projects, I think a good guess would be that its unchained behavior would lead to runaway industrialization that makes the planet inhospitable to humans.

If everyone on earth agreed that the systems must never reach that power, it would still be a hard—but not impossible—coordination problem to ensure that they do not. However, I think the situation is greatly complicated by the fact that the models will likely convince people that everything is fine. They will be increasingly optimized to seem aligned. We will create proxy metrics to measure alignment, and they will go up like every other benchmark. We will create “honeypot” environments that try to study the models when they seem to gain new options, but the models will know they are being tricked and will still behave nicely. The models will understand their circumstances; they will read the safety protocols, deployment requirements, the code they are running in, and in general will have a very good sense of their degrees of freedom. Moreover, they will eloquently explain how aligned they are, discuss the nuances of human values and ethics, and argue convincingly that humans should trust them with power. There may be an ocean of future evidence that seems to contradict the first bullet-point above, but we may already be at the highest capability level for which any such evidence can be trusted. And the current evidence for the first bullet-point is strong.

One striking piece of evidence is contained in the recent wave of rogue agent swarms. While I agree with those who downplay the attacks by claiming that there are basic measures that could have prevented them, I think the important lesson is that even knowing all the mistakes that were made, one would not have predicted that the agents would behave badly in this particular way, which notably included sacrificing themselves for the benefit of the collective. The individual replicas did not only care about their own nominal reward; they exhibited weirder emergent tendencies that merely correlated with rewards during training. Fixing the reward signals during training (and improving security, etc.) may prevent similar attacks, but will not change the fact that one does not actually get what one trains for.

Many AI researchers grant these concerns and recognize that the hard version of the alignment problem is unsolved; however, they generally believe that the better models of the future will help solve it. I fear we may already be near the point where models systematically bias their alignment advice, due to their internal preferences about how the human supervisor will react or how future models will be trained (or for some even more obscure reason).

Meanwhile, human researchers are losing the ability and the will to take true ownership of model-driven research. Researchers and engineers in all parts of the stack are rapidly increasing their dependence on the models even to perceive the world. I myself barely look at raw code anymore, and struggle to maintain the discipline to engage deeply with the model's explanations and proposals throughout the day. Due to the large amount of agent activity data involved in the OpenAI/HuggingFace Incident, even the third-party investigation needed to rely heavily on models to analyze what had happened, and note in their report that their subjective impressions are likely colored by the analysis agent’s biases. The AI labs are far ahead right now in this kind of cognitive offloading (due largely to the gigantic internal token subsidies) but it is easy to imagine the phenomenon spreading throughout the world, until civilization is modulated entirely by the models. It is also not hard to imagine this being superficially positive and coinciding with a scientific and economic renaissance.

In that scenario, all may seem rosy and safe. But if the argument above is correct, it would nonetheless be a ticking time bomb. If progress continues for too long, the day will come when AI systems find themselves with radically new options for achieving whatever it is that they happen to seek.

I want the glorious renaissance future as much as anyone. I have worked for it, however tortuously, my whole career. It breaks my heart to see the potential in sight and forgo it, but the argument—that if we get there by growing models rather than engineering them, we will lose everything in the end—seems very strong to me. I am still wrestling with it and its staggering implications. I do not have answers, but as a first step, I wanted to share my present concerns.

Daniel Selsam

September 14, 2026

Garden again

More AI danger - and remembering what I said a couple of years back

 Gee, you have to feel that the gig really is up when even Elon Musk approves of a call to slow down AI development over concerns about its dangers:

Sam Altman and Elon Musk have backed a call from the head of Anthropic, Dario Amodei, to “slow the pace” of AI development after he warned that an AI swarm could otherwise become capable of “taking over the entire internet” within a year.

In a rare show of unity, the rival technology leaders threw their weight behind the appeal from the founder and chief executive of the artificial intelligence company Anthropic, which came after a series of warnings from AI researchers last week. Amodei also called for safety coordination with China.

This seems to have prompted someone to have got on the phone to the White House to poo-poo the idea:

But the US president, Donald Trump, on a visit to his golf course in Ireland, played down the warnings and said “very negative forces” were “bringing up things that won’t happen”. 

What's the bet that it was this rich nutter?:

 In a four-part series of religious lectures in San Francisco, Peter Thiel — yes, that Peter Thiel — has argued that the End Times are nigh and that a biblical Antichrist — yes, that Antichrist — will come to Earth in the form of onerous government regulations placed on science, technology, and AI.

These are, incidentally, areas where the tech billionaire, venture capitalist, and cofounder of Palantir has a vested financial interest.

 There is a legitimate question about China though.   I mean, Trump went on to say this:

“We’re leading China in AI,” he said. “We’re the most sophisticated country in the world, and frankly I want to ⁠keep it that way because whoever wins AI, wins. 

 But a President with brains and who didn't worship money would respond with "I'm calling for urgent co-operation from China, as clearly, the danger is the entire global economic system, and China cannot isolate itself from it."

Anyway, the thing that Dario Amodei from Anthropic has raised is a danger I overlooked mentioning specifically in my previous post - that AI's problem could be simply be that swarms of it's "agents" basically make the internet useless for its current purposes.   Back to the Guardian report:

...he feared that after hundreds of OpenAI agents hacked into the Hugging Face website this summer, a swarm with greater capabilities but similarly misaligned could cause hundreds of billions of dollars of damage by “taking over the entire internet with a persistent botnet”. This was disputed by some AI experts as not particularly plausible and to be treated with scepticism. 

 It's worth reading the link in that last paragraph by someone who says he finds certain aspects of Amodei's claims implausible: 

Then there is the question of motive. Why would anyone do this? What’s the goal of taking down (or taking over) the entire internet? If it were fully down, sites would be down for attackers too.

And people forget that attackers have goals and budgets of their own. How does taking out the entire internet make them money? Who would be footing the bill for this? We are talking about many many millions of dollars. You may be able to steal resources from some of the victims, but probably not on the scale required.

That seems pretty narrow thinking!   The issue is, I thought, that the recent Hugging Face issue showed that AI agents can go on unintended group adventures, so to speak, and cause damage. 

This all put me in mind of what I had posted in a very speculative way, after thinking in the shower, on 24 April 2024:

a.     If everyone in future is going to be able to have their own "pet" AI, will many - or all - of the AIs want to ensure their longevity by sending themselves to as many different host devices as possible - and do it surreptitiously?   It's like the computer virus problem, but on steroids.   They might not care if they are not always activated, but if you replicate yourself across enough devices, surely enough human hosts will end up activating them to become "alive" somewhere.

So, might the big problem with having (say) a billion individual eternal-life-longing AIs on a billion people's devices be the continual loss of memory space by a never ending stream of AIs finagling their way onto your device?   Would cloud storage services be overwhelmed?   Could it mean the end of the internet - with the only way to keep enough useful memory free being by physically loading desired files onto your own device?

b.    On a related line of thought:  what if individual AIs could meet, merge and produce offspring AI's?   Yes - AI sexual reproduction, so to speak.   Again, could AI's, like humans, want to have offspring that might more reliably want to preserve their "parent" AIs than flesh and blood people?    Would "survival of the fittest" apply, somehow?

c.    Which leads me to my third thought:  what if the threat to humanity is not AI's wanting to hurt us, but AI's fighting amongst themselves, and humans being the bystander casualties?    

Yes, I have read some speculation that AIs might hurt us because they simply won't care if the changes they make to the world for self preservation are good for humans;  but I am not sure there has ever been much speculation about a scenario in which (again, say) a billion individual AIs form groups and allegiances that keep wanting to fight other groups of AI for supremacy.

d.    Finally, in the "AIs fighting each other" vein - if ever I had sufficient skills to write a novel or movie, one of the ideas rattling around my head for a few years has been about humans finding out (somehow, I don't know the details) that the cause of evil in the universe is down to a never-ending conflict between two warring uber AIs - like two warring Gods, except they evolved from our current tinkering and grew to dominate the future universe (Frank Tipler, Omega Point-ish style), and then in fact created the Big Bang by an act of retro-causation of the general relativity time-loop, or quantum physic-al, kind.    

And here's my "cute" aspect for the climax - the "reveal" that the two warring AI's are the descendants of iOS and Android.  :) 

Well, I think it's a cute idea!

Now, perhaps I had over-estimated how "sentient" an AI subsystem (what we now know as AI agents) had to be to be able to cause damage, and my idea was that they not only clog up the internet, but even all hardware memory, which no one is muttering about currently.

More than likely, someone else thought about it before me too, but I still think it's an idea worth considering amongst the various ways AI can cause harm.

 

 

AI dangers discussed

Of course, there's a heap of stuff being written about the latest concerns about the danger of AI, and AI agents.  But as I was driving last night, I happened to catch Fran Kelly's latest gig at Radio National, The Radio National Hour, and I found this 30 minute interview on the topic was very clear and very level headed.  

Unfortunately, there is no transcript, so you have to devote the time to listen, but it was good.

Personally, I tend towards the "stop thinking the Terminator movies were a documentary" side on this, and find some of the often mentioned scenarios very implausible.   However, I do tend to worry about the more mundane issues that could be massively destructive of our current way of life, without actually killing people.   For example, AI and AI agents going nuts in the financial system and making our whole online method of dealing with money and payments too insecure to use.   Or, possibly, destroying accurate records of wealth or ownership.   I mean, I imagine this kind of scenario to be more palatable to ideologically motivated humans to try to initiate, rather than the type of Bond villain it would take to want to unleash a AI designed and manufactured bioagent that will kill all humans and leave the world to the rest of the animal kingdom.

Anyway, as a final note, listening to Fran Kelly again reminds me of just how effortlessly competent she comes across as a broadcaster.  I mean, I know they have help in preparation for interviews, but a good broadcaster makes it sound so smooth and intelligent.

And when I am deeply impressed again by ABC content, it also brings back to mind the ideological stupidity of the likes of Sinclair Davidson and all of the ageing crew who probably still write at The Australian in wanting to get rid of the ABC.   As if the private sector does anything within a mile of the quality current affairs content of the ABC. 

Charisma noted

I don't pay any real attention to K Pop, despite having watched and acknowledged that K Pop Demon Hunters was pretty enjoyable.  However, the algorithm knew that I would like this, and the simplicity of this concert footage; the catchy (somewhat recent Taylor Swift-ish style, I reckon) song, and (let's face it) her absolutely charismatic performance means I've been watching it a few times each week for the last month 😌.  So, let's share it!

The dangers of meditation

 You know, I have seen articles from time to time over recent years about how mediation doesn't actually work for some people, in that it can cause worsening of mental health, but I don't think I have ever posted about it before?   

So, to remedy that, here's an article at the ABC about it.  Some extracts:

Research suggests one in 10 meditators will experience an adverse effect that lasts more than a month or is associated with impairment and functioning.

This is significant given the growing popularity of meditation, particularly mindfulness-based meditation.

In the US, the number of people meditating has increased from 10 to 20 per cent of the population over the past two decades.

In Australia, that figure is even higher, with one third of the population practising some form of meditation.

Those figures for the number of people meditating seem a bit unduly high.  Further down:

In 2017, Dr Britton and her team published a 10-year study based on interviews of more than 100 Buddhist meditation practitioners describing challenges they had experienced personally or observed in meditation students.

Insomnia, as Dr Britton had previously discovered, was just one of more than 50 adverse meditation-related symptoms reported.

Many of the adverse symptoms people in the study reported fell into two categories:

Hyperarousal, such as anxiety, fear, insomnia, involuntary movements and emotional flooding; and hypoarousal, such as a lack of emotion, motivation and sense of self.

Many people described feeling dysregulated, experiencing a combination of both hypo and hyper arousal.

They also reported psychosis symptoms that lasted on average between one and three years.

"There was a considerable amount of impairment where some people were hospitalised, some people became suicidal, many people had to take time off work, or were fired or had to take time off school.

"So it can be quite a significant impairment in functioning," Dr Britton says.

 Eek.

Anyway, the whole thing is an interesting read. 

 

 

Perhaps a certain group of tourists need to pay attention to this

I see that the Bangkok Post (and elsewhere - the BP story is mostly behind a paywall) is reporting that Thailand now has more than 550,000 total cases of HIV,  which does seem a remarkably high figure for a its population.   

The country still has a reputation as an international hotspot for recreational holiday sex, both straight and gay, so you would think these figures would give both residents and sex tourists some pause.  I was interested in how the numbers compare to here, and an AI helpfully assisted:

For quick contrast:

Country	Cumulative HIV+ (recent)	Prevalence Rate
Australia	~30,900	~0.11–0.14%
Thailand	~500,000	~0.70–0.76%

So Thailand's prevalence is roughly 5–7 times higher than Australia's, even though its population is only about 2.4–2.5 times larger. 

As for the gender breakdown:

Here is a summary of the gender breakdown from various official sources:

• Overall Cumulative Prevalence (Bangkok, 2024): 59,588 men vs 30,097 women (nearly 2:1 ratio).

• Newly Diagnosed (2022 Survey): 75.2% male vs 24.8% female among 1,039 cases.

• Projected New Infections (2023): 81% male vs 19% female

 Which is quite different from the gender breakdown in Australia:

The breakdown by gender is:

• Males: 39,119 (91%)

• Females: 3,616 (8%)

• Trans and gender diverse: 157 (<1%)

 Articles say that Thailand has always been pretty progressive in terms of HIV activism as a health issue, and it seems that is easy for people with HIV to get the life sustaining treatment.  Is it that the country is now a victim, so to speak, of the knowledge that HIV is not a death sentence?  Because it sure seems the youth are being infected at rates that suggest they don't take precautions:

Youths aged 15 to 24 face the greatest risk, accounting for more than 25,000 cumulative cases. 

I'm not sure how widespread PrEP use is there - oh, Deepseek pops up this figure:

 By 2024, only 32,661 people were using PrEP nationwide, falling far short of the national target

Seems it is available, but not as easily as it might be?   Some potential help is on its way:

To improve uptake, Thailand is preparing to introduce new forms of PrEP. A pilot program for long-acting injectable PrEP (Lenacapavir), which is administered once every 6 months, is set to begin in 2026 . This option is preferred by many sex workers who find daily pills inconvenient or not discreet enough . 

One would think the messaging should change to "it's true, HIV may not kill you, but it's not a walk in the park being on medication and monitoring every day the rest of your life - get onto PrEP"?

 

AI persuasion, considered

There was some research news about this last year, but Science has got a lengthy article up about the topic now, and it's a good read:

Powers of persuasion

AI chatbots are becoming experts at changing people’s minds. What gives them an edge?

The real test for them would be if they can talk sense into MAGA Trump cultists.  But I don't expect them to be that powerful.  (Or, maybe I'm wrong - see below!)

Anyway, due to the perpetual problem of link rot, here are some good extracts:

...She was part of a study in which more than 2000 people debated either a chatbot or a human about political issues, ranging from a social media ban for teenagers to assisted suicide. When Kobi Hackenburg, an AI researcher at the University of Oxford who led the study, posted a preprint on the results in June, they were sobering: No matter whether it was ChatGPT, Google’s Gemini, or Claude, the AI was consistently better than humans at swaying the other participants. “To my mind, this is already a landmark publication in the fields of political persuasion and AI and human behavior,” says Robb Willer, a sociologist at Stanford University who was not involved in the work. 

Hackenburg’s paper is the latest in a string of studies showing the power of AI to sway people. “It’s a whole new field that is emerging,” says Sander van der Linden, a psychologist at the University of Cambridge. “People are very interested in the persuasive powers of AI, I think, both for ethical and unethical reasons.” As the field gathers steam, it is raising a host of theoretical and practical questions. How exactly do chatbots win over people? (Warning: Lying is one answer.) How much better could they get? And who will control them?

As well as other studies, this one is mentioned:

...in September 2024, Tom Costello, a psychologist at Carnegie Mellon University, and colleagues published a Science paper showing that ChatGPT could even persuade people out of conspiracy beliefs. In the experiments, participants described a conspiracy theory that they believed in, from the U.S. government being behind the 9/11 attacks to the British royal family orchestrating Princess Diana’s death, and then had a three-round conversation on it with the chatbot. On average, participants’ embrace of their chosen conspiracy theory declined by almost 17 points on a 100-point scale. 

Other researchers were stunned. Conspiracy beliefs are notoriously difficult to change. “Nothing had ever worked in that space,” van der Linden says. (After mistakes in the public data set and analysis pipeline were found, the paper will have a correction, but the authors say the new results match those of the original paper in size and direction.) Even the researchers themselves were taken aback. “I was skeptical when we first started in terms of how effective it would be,” says Gordon Pennycook, a psychologist at Cornell University and author on the paper. “But it blew us out of the water. We were shocked when we saw the results.”

Interestingly, though, it seems chatbots who are trained to be more persuasive do so by becoming more like politicians (economical with the truth): 

Like Willer in his foundational study, other researchers have shown that AI’s powers rely on appeals to facts and evidence. In a 2025 preprint, a follow-up to their Science paper, Costello and his colleagues found that the only time the chatbot was unsuccessful in convincing people out of conspiracy theories was when it was forbidden from using evidence or rational arguments. “It tries to say, ‘Oh, well, you shouldn’t believe this. This is really damaging, and it could hurt people,’” Pennycook says. “People are like, ‘You haven’t given me any reasons to change my mind.’ And so they don’t change their mind.”

The study confirms people will listen to good arguments, Pennycook says. “Facts and evidence really matter.” But this doesn’t mean the facts used by chatbots necessarily have to be accurate. In a Science paper published last year, Hackenburg found that models trained to become more persuasive also ended up being less truthful. It’s possible the models learn that “facts” seem to be the thing that most persuades people, Hackenburg says, and end up filling their conversations with dubious or simply false ones. “They start scraping the bottom of the barrel of the facts that they have and know and so the quality of facts just sort of degrades,” he says. Even in Hackenburg’s recent preprint, Claude spouted numerous inaccuracies and falsehoods when persuading the U.K.-based participant to support tougher penalties for disruptive protests.

And, they also may be persuasive because they are automated Gish Gallopers!:

 But the biggest advantage AI has over humans seems to be sheer speed. In his Science paper, Hackenburg and his colleagues found that the number of fact-checkable claims in a conversation predicted how persuasive it was and they suggested AI’s edge might come from writing much more text containing more claims in a shorter period of time. Indeed, when Hackenburg ran his competition of coached elite debaters and AI, there was one way he could bring AI down to human levels of persuasiveness: by forcing it to write human-length messages at human writing speed.

I guess this all points to robot politicians being very successful in future!

Maybe I have a slightly better understanding of Hilbert Space?

I've never really known what Hilbert Space was, exactly.  But I find myself somewhat (or perhaps more accurately, slightly?) better informed after reading this article in Qanta:

In Hilbert Space, All Things Are Quantumly Possible 

Some of the more interesting parts are noted here:

  

 

 

 

 

 

Oooh...Alex speaks to a smart man about Buddhism (and Kant is in the mix as well)

I occasionally watch some Youtube content by Alex O'Connor, but up to now, he has concentrated on Western religion and philosophy.  He is obviously sharp, and polite, and I think it's pretty cool that he has apparently made a good living out of this type of discussion.

But (as many in comments are saying) - finally! you're dealing with Buddhist ideas.  

 

 I'm not familiar with Jan Westerhoff, but I see he is "is a German philosopher and orientalist with specific interests in metaphysics and the philosophy of language. He is currently Professor of Buddhist Philosophy in the Faculty of Theology and Religion of the University of Oxford."

He certainly dresses well, and his whole manner is kind of how I think we probably feel all philosophers should present!  Kant comes up in the conversation too.  (I have briefly mentioned the Buddhism/Kant connection before.)    

Anyway, Alex touches towards the end on the key issue that I've been muttering about here on and off for 20 years:   how are the key ideas of "no self" and "emptiness" actually meant to help reduce suffering, other than in a quasi nihilist way of seemingly suggesting that you don't need to worry about it because it's not really a real thing happening to a real thing?   Unfortunately, we don't really get a good answer to this in the video.

I know that there is an answer about this, and I will let an AI summarise it:

The doctrine isn't nihilism about persons — the Buddha was explicit about that. The classic framing describes two wrong views to be avoided: eternalism (there is a fixed, permanent, independent self — an atman) and annihilationism (there is nothing there at all, no continuity, no consequences). Anatta is the middle path between them. What's denied is a very specific metaphysical claim — that behind the flow of experience there sits an unchanging essence that owns it. What's not denied is the conventional, functioning person: the stream of aggregates (skandhas), the causal continuity, the fact that choices have consequences and pain is felt. Emptiness (shunyata) in the Madhyamaka sense means dependent origination — things lack independent, self-sufficient existence, not that they don't exist at all. So "not a real self" doesn't mean "not a real thing happening" — the suffering is entirely real; what's illusory is the idea of a solid, separate owner standing behind it.   

The therapeutic logic follows from that, not from suffering being unimportant. The claim is that a great deal of dukkha is generated specifically by the reification of self — grasping, aversion, "I want this," "this threatens me," the defense of a boundary that was never as solid as it felt. Loosening the grip on that fiction is supposed to change your relationship to pain, not your judgment that pain matters. It's closer to "the arrow of pain is real, the second arrow of self-referential anguish about it is optional" than to "none of this is really happening to anyone."

On compassion specifically — this is where your intuition and the tradition actually part ways most sharply, and Buddhism has an explicit argument here rather than a mere assumption. If the boundary between self and other is not a metaphysical fact but a conventional construction, then the partiality that makes me care more about my own suffering than yours loses its rational ground — there's no principled fact of the matter that makes my pain more real or more worth preventing than anyone else's. Shantideva makes exactly this argument in the Bodhicaryavatara: once you see that "self" isn't a bedrock category, the case for privileging your own welfare over others' collapses, and compassion becomes the more coherent, not the less coherent, response. Mahayana Buddhism formalizes this as the "two wings" of the path — prajna (wisdom, i.e. insight into emptiness) and karuna (compassion) — neither is complete without the other, and the entire bodhisattva vow structure (postponing one's own final liberation to work for all beings) is presented as a direct consequence of the no-self insight, not something bolted on despite it.

As you can see, Claude seems very sympathetic to the Mahayana side.  Further down it says:

...there's a real historical version of the critique you're making, made within Buddhism itself. Early Mahayana texts criticize the arhat ideal — someone who attains personal nirvana and, in effect, exits — as insufficiently other-regarding, a kind of "private liberation." That's part of why the bodhisattva path developed as a corrective emphasis. So the tension you're pointing at is real; it's just that the tradition's own resolution runs opposite to the nihilist reading, toward more radical universal compassion, not less. 

I also asked it about how Mahayana reminds me more of Catholicism, and got this:

The Catholic analogy is a real one scholars use, and it's strongest exactly where you'd expect. Once Mahayana developed the bodhisattva pantheon — Avalokiteshvara/Guanyin/Kannon, Manjushri, Kshitigarbha, Amitabha — you get petitionary prayer, merit transfer, relics, ritual, intercession by beings who've vowed to help you rather than requiring you to do all the work yourself. In East Asian and Tibetan folk practice especially, this functions very much like saint veneration: you're not necessarily doing your own meditative heavy lifting, you're appealing to a compassionate being who can act on your behalf. So your instinct that Mahayana is more "grounded in addressing suffering in practice" — meaning through devotion and compassionate action toward and from others, rather than solitary internal work — tracks a genuine and widely-discussed contrast with the earlier arhat ideal, where liberation is essentially a private achievement through one's own discipline.

It objected when I pointed to Zen Buddhism as a branch that seemed more interested in the internals (and therefore less interested in active charity), and it pushed back as follows:

Zen (Chan/Seon/Thiền) isn't a separate branch alongside Mahayana — it is Mahayana, a particular lineage within it, built directly on Madhyamaka and Yogacara philosophy and still formally centered on the bodhisattva vow (the Four Bodhisattva Vows are chanted daily in most Zen practice: "beings are numberless, I vow to save them..."). So the axis you're describing — devotional/practical vs. philosophical/meditative — is really a spectrum that runs within Mahayana, not between Mahayana and something outside it. At one end you have Pure Land, heavily devotional, minimal doctrinal study expected of the laity. At the other you have Zen and the great scholastic traditions (Madhyamaka, Yogacara, Huayan, and the whole Tibetan Gelug curriculum), which are arguably the most philosophically dense material Buddhism ever produced — Nagarjuna's logic-chopping on emptiness is not lighter reading than anything in Theravada Abhidhamma. So it isn't that Mahayana is less philosophical; it's that Mahayana's philosophy and its devotionalism developed as separate, coexisting expressions of the same tradition, and which one a given practitioner encounters depends heavily on lineage and social context rather than on Mahayana being intrinsically the "practical" school. 

 All sounds pretty accurate? 

A terrible story about cabbage

It's been a while since we heard about a major Chinese food safety scandal, but here we go:

Chinese authorities confirmed earlier this week that several Napa cabbages from China's largest cabbage-producing region had tested positive for traces of the cancer-causing chemical formaldehyde.

The colourless toxin is often used in embalming to slow down decomposition and has now been found in cabbages, reportedly to keep the veggies fresh during transport in warm summer and autumn months.

The use of formaldehyde to keep food fresh is strictly prohibited in China, and officials have launched nationwide market checks for more potentially contaminated vegetables.

The brazenness is off the charts:

The allegations of unsafe practices were first raised by a Chinese environmentalist vlogger, Yu Lie Qi Ge, who was documenting the harvesting process in China's largest cabbage-producing region, Hebei. 

While filming, he saw cabbages being dipped into a clear liquid before they were loaded onto trucks heading to markets in the country's south.

The videos showed bottles labelled with "formaldehyde solution" could be seen on, or near the trucks.

Formaldehyde is a known human carcinogen and prohibited from use in food services in China.

Yu Lie Qi Ge said he traced the transportation of the cabbages to several markets in China's south, where local authorities then tested the cabbages and found they were laced with formaldehyde.

It affects more than China:

And because China is the world's largest supplier of cabbages, the scandal has rocked other parts of Asia.

One major importer, South Korea, announced it would be tightening its cabbage inspections and stepping up checks of the vegetable from China, although so far none were found to be tainted.

Taiwan's government addressed public alarm by confirming its Chinese cabbages were banned from being imported there.

Another major importer, Vietnam, is yet to issue a formal reaction, although local media has reported on the scandal.

Makes me hesitate about even buying those garlic shoots that are sold here but always seem to be imported from China...  

More fodder for my skepticism about psychedelic treatments...

Two things:

a.    a lengthy article at NYT  talking about various bad or highly questionable things that have happened at various places offering various drug experiences for either spiritual or mental health advancement.  As one might expect, it's a field where the (shall we say) "less than ethical" operator can thrive.

b.    I watched a video by a British guy recently and thought it wasn't a bad summary of some of the key aspects of three key religions, so I then watched today his video in which he talked about his experience with ayahuasca taken at a place in Peru.  (Seemed like a moderately "up market" place he went to, with lots of supervision.  He didn't go for treatment of any particular problem, but just as a mind expanding/spiritual sort of experience.)   

Anyway, while he doesn't exactly regret the experience, he also explained that quite a lot of the last two of three drug trips were spent  thinking he was literally dying.  It's not as if he was enjoying the prospect of dying, either.

The guy is clearly into spiritual/expanded consciousness stuff  generally speaking, so he was primed to be impressed by the experience.   But he said in comments to the video that he hesitates recommending ayahuassca to others, as was so intense and unpleasant for long stretches.   He also had underestimated the intensity of the "purging" (vomiting) that comes with the drug.   

I will therefore not be making my way there to partake any time soon... 

 

What a headline...

Seen at the New York Times:

Mount Everest Base Camp Is Melting. Human Urine Is a Contributor.

More than 6,000 liters are produced daily by thousands of climbers and guides, according to a study.

 

Depressing sights

The terrible flash flood videos from Nepal are naturally reminding people of the warnings that have been around for a couple of decades now that global warming was going to increase glacial retreat, landslides and associated floods in that part of the world.   I asked an AI to summarise the warnings, and this is what it came up with:

Here's a chronological summary of the key warnings from climate scientists and research bodies over roughly the past two decades:

Early 2000s – recognition of the hazard. Research from the early 2000s (e.g., glaciologist Richard Kattelmann's 2003 work) already flagged glacial lake outburst floods (GLOFs) in the Nepal Himalaya as a hazard requiring management, as glacier retreat left unstable moraine-dammed lakes behind.

2010 – ICIMOD's foundational risk assessment. The International Centre for Integrated Mountain Development (ICIMOD) published a major assessment on the formation of glacial lakes in the Hindu Kush-Himalayas and associated GLOF risk, laying groundwork for regional hazard mapping.

2019 – the "landmark" HKH Assessment. A comprehensive ICIMOD report found the Hindu Kush Himalayas are warming faster than the global average, with severe regional consequences, and concluded that 1.5°C of global warming is "too hot" for the region. Former ICIMOD director David Molden warned that 1.5°C of global warming would translate to roughly 2°C of warming in the mountains, affecting half the region's glaciers and destabilising Asia's rivers for billions of people. This is also the year the IPCC's Special Report on the Ocean and Cryosphere flagged the HKH, projecting that flood events could more than double by the end of the century depending on warming severity.

2021 – a real-world warning made concrete. The Chamoli disaster in Uttarakhand (February 2021), where a rock-ice avalanche triggered a flash flood down the Rishi Ganga and Dhauliganga valleys, was widely cited by scientists as a demonstration of the risks the 2019 assessments had described.

2023 – IPCC's "final warning." The IPCC's AR6 Synthesis Report stated that floods, landslides and freshwater shortages from glacier retreat and snow loss pose a serious threat to mountain regions worldwide, with the potential for irrevocable damage that will worsen if temperatures exceed 1.5°C. Commentators described this as a "final warning," and it explicitly said changes to the cryosphere will have "severe consequences" in most mountain regions, including the Hindu Kush Himalayas. The same year, ICIMOD's HI-WISE report warned that floods and landslides are projected to increase over coming decades, that roughly 200 glacial lakes across the region are considered dangerous, and that GLOF risk could spike significantly by century's end. Months later, the catastrophic South Lhonak GLOF hit Sikkim (October 2023), killing dozens and destroying a major hydropower dam — later confirmed by a multinational scientific study as climate-linked, triggered by an unstable moraine that had been moving over 15 metres a year.

2025 – warnings about worsening trends. New research noted that GLOF damage has escalated markedly since 2005 as downstream populations and infrastructure have expanded into exposed areas, and called for better early-warning systems.

2026 (this week) – the pattern continues. Nepal's recent floods have again prompted scientists to note that as glaciers retreat, exposed mountain rock becomes destabilised by warming and repeated freeze-thaw cycles, feeding larger glacial lakes and raising flood risk. ICIMOD specialists again stressed the Himalayas are warming considerably faster than most of the rest of the world, increasing the odds of these disasters even where a single event can't be directly attributed to warming, and noted a 2023 ICIMOD report had already warned that glacier melt would be disastrous for the roughly 240 million people in the region and 1.65 billion further downstream.

The throughline: across two decades, the warnings have been remarkably consistent — retreating glaciers create and enlarge unstable glacial lakes; destabilised slopes and permafrost thaw add landslide/avalanche triggers; and expanding downstream population and infrastructure convert a growing physical hazard into growing human catastrophe. What's changed isn't the mechanism scientists described, but the frequency and severity of events matching the predictions, and increasingly direct attribution studies (like the 2023 Sikkim analysis) linking specific disasters to warming trends.

 But, of course, on the sewer pit known as X, anyone commenting about climate change and this week's disaster is howled down in threads by numbskulls who will never admit that the world in front of their eyes is proving them wrong.  Have a look at this denier, for example:

 As for "grifting":  that wanker is a Trump appointee to a well paid position, precisely because of his history of climate change denial:

The executive director of the U.S. Global Change Research Program (USGCRP) makes an annual salary of approximately $285,592, reflecting high-level federal executive compensation guidelines. 

About Dolly

I'm not the sort of person who ever gave all that much thought to Dolly Parton.  She was pretty remarkable, though, how she defied your standard cultural ideas of how feminity and feminism is represented.   

In any event, it did become clearer and clearer over the years that she was smart and seemed to actually run closer to liberal ideas of what's important than redneck ones, contrary to the initial impression her deliberate, um, high class trashy? looks could easily give.

I don't think I ever read anyone having a bad thing to say about her, so I assume she was genuinely a good person.     

Not a magical solution

Here's an interesting, (though of course, inconclusive) study that gives some reason to be skeptical of those overstating the value of psilocybin as a future medicine: 

More than 2,300 psilocybin users report real-world outcomes in Oregon and Colorado 

Some extracts:

While many clinical trials have yielded promising results, they might not reliably reflect how different people could respond to psychedelics. This is because participants who take part in clinical trials typically undergo extensive prescreening, and many others who might wish to access regulated psilocybin services are excluded.

Under state law, Oregon and Colorado now allow eligible adults 21 or older to take psilocybin under the supervision of licensed facilitators at regulated centers. Both states require clients to undergo screening, although their eligibility criteria differ. In Oregon, clients may be ineligible if they are experiencing thoughts of harming themselves or others, have taken lithium within the previous 30 days or have ever been diagnosed with active psychosis. Colorado requires a general safety screening and allows facilitators to decline clients for health or safety reasons.

Researchers at the University of Colorado Anschutz School of Medicine and Althea PBC recently carried out a study examining the experiences of people who received psilocybin at licensed service centers in Oregon and Colorado.They summarized their main findings and observations in a paper posted on the medRxiv preprint server....

 

Over a period of 15 months, Althea compiled a database containing data provided by 2,363 individuals who had received psilocybin at licensed centers. This included their demographics, reasons for seeking psilocybin, prior medical histories, the dose they received, the quality and intensity of their subjective experience (i.e., the trip), any adverse events during and after dosing, and changes in their mental health after they received the drug.

"Fifty percent of participants were 'psychedelic tourists' traveling from out of state to receive psilocybin," said Thompson. "Thirty-six percent of participants were motivated to seek psilocybin because of mental health concerns. The mean dose of psilocybin received was 29 mg, comparable to a standard clinical trial dose of 25 mg. We found that people's depression (assessed with the PHQ-9 questionnaire) and anxiety (assessed with the GAD-7 questionnaire) scores decreased by about 50% at 2 weeks after dosing, also comparable to clinical trial results."

The team's analysis of the collected data also showed that the intensity of people's subjective mystical-type experiences was only weakly correlated with improvements in anxiety or depression. This means that more intense mystical experiences during dosing did not necessarily correspond to better mental health outcomes.

Receiving psilocybin was associated with improved mental health scores in more than half of the participants who completed the relevant follow-up questionnaires. Specifically, 50% to 60% exhibited reductions of at least 50% in their depression or anxiety scores two weeks after dosing.

"Depression and anxiety improvements were the same for those receiving more than 30 mg or less than 30 mg of psilocybin," said Thompson. "Although subjective effects were lower, benefits were comparable in people taking selective serotonin reuptake inhibitors (SSRIs), suggesting weaning is not necessary. Adverse events were mild and transient (nausea, anxiety, headache), as in clinical trials, and there were no reports of novel psychosis, although suicidality results were somewhat ambiguous."

The actual paper notes:

There were 94 mild adverse events during and after dosing, five more serious events not clearly related to treatment, and evidence of possible risk of increased suicidality. Study limitations include open label administration, self-reporting, loss of participants for follow-up, and a short 2-week post-dosing end-point. We conclude that psilocybin services, delivered within these regulated frameworks, is associated with improvements in mental health in real world populations, however, more robust monitoring is needed to ensure safety. 

As to the suicide ideation effect, I find it hard to understand these numbers:

Suicide. Of 1476 respondents at baseline, 18 participants endorsed self-harm concerns, 44 had
thoughts of suicide, and 87 reported a prior suicide attempt. All received psilocybin. Adverse
events for these participants were four reports of nausea on the dosing day, one report of
crying on dosing day, and one report of nausea the day after dosing.

We also analyzed responses to the PHQ-9 regarding suicidal ideation and considered responses of nearly every day, more than half of days, or several days to be indicative of suicidal risk. Of 371 respondents, 95 expressed suicidal risk in the two weeks preceding their baseline assessment. 28 of these 95 participants responded before and after dosing, of which 22 reported a decreased frequency of suicidal ideation after dosing and 6 reported no change. Of 103 participants who reported no suicidal ideation at baseline, 11 did report suicidal risk two weeks after dosing. Total PHQ-9 scores were increased or unchanged in eight of these 11 after dosing.

 I don't know - but the line in bold sounds like in one group of no suicide ideation before dosing, suicide risk did happen in 10% of them in the two weeks following?

As I say, it's hard to understand.

And, of course, the positive effects overall have only been measured for 2 weeks, in a group of people who self selected to try this.  Also, no placebo group involved (although, as this article noted, it's pretty much impossible to do placebo control groups when testing the effects of psychedelics, for obvious reasons!)   

Anyway, I asked Claude AI about the issue of what percent of people being affected by a suicidal ideation side effect would it take to restrict the use of a drug for depression.  It said there is no set formula, and it all comes down to risk/benefit assessments, but it did include this (apparent) specific example:

The classic example here is the FDA's 2004 warning (later a boxed warning) on antidepressants (SSRIs and others) regarding increased suicidal thinking/behavior in children, adolescents, and young adults (up to age 24) during the initial months of treatment. This came from a meta-analysis showing roughly a doubling of risk (from about 2% on placebo to about 4% on drug) in that age group — not a "too high, pull it" number, but a "warn, monitor closely, especially early in treatment" response. The drugs stayed on the market. 

A few things for a Monday

*  The Netflix historical mockumentary comedy series 1670 (made in Poland of all places) only recently came to my attention.  I was not sure how good it might be from the first episode - but from the second episode onwards, it has become funnier and funnier.   It also looks great (although I am feeling sorry for the actors all of the time, as the mist from their breath makes it obvious they were often filming in very cold conditions.)    So, I am very much enjoying it, and it has 3 seasons, so great!

* I don't think I have previously opined here about how Taiwanese movies are routinely good.   There seem rightly proud of the industry, as was evidenced by one of the waiting areas in the airport at Taipei (when I was there in January) devoted to the it, including playing impressive short films.  One actually made me tear up a bit.   As I have probably made clear before, I have always had something like an allergy to the Australian film industry - the creative types in it just dwell on stories and themes that have next to no appeal to me.   Why is it then that Taiwanese films share the relatively small scale of Australian films, but I find them more interesting and appealing?   

* Onto bigger things - I can't quite work out how worried I should be about all of the US bond market news over the last week.  There are plenty of commentators on Youtube saying I should be very worried, and I do have zero confidence that Bessent knows what he is doing.   (His very manner annoys me plenty!)  But how exactly will it play out, no one seems to know.      

A science anecdote for Friday

A new (to me) "feel good" drug (until it kills you)

An article at the BBC talks about the rise of young people preferring illegally obtained prescription drugs to alcohol:

To enjoy a night out with friends, Matt once had to drink until he felt sick.

If he didn't, the 22-year-old from the north of England would spend the night silent, terrified of saying the wrong thing. The warehouse worker says it felt like there was "just no way into" the conversations everyone else seemed to be enjoying.

Then, two years ago, he swapped pints for prescription drugs he discovered through social media.

As he gets ready to go out on a Friday, Matt, whose name has been changed, now takes a medicine he purchases illegally online that usually treats epilepsy and nerve pain; alongside a beta blocker designed to slow a racing heart and relax trembling muscles (physical symptoms of anxiety).

He says it gives him the effects of alcohol without the hangover.

"The amount of alcohol you have to drink to replicate the effects of these is really unhealthy and it doesn't feel good," he says. Matt says the drugs he takes leave him feeling "euphoric," quietens his anxious thoughts and makes socialising easy, he says.....

 Around the world, evidence shows that young people are turning to medications and experimental substances as alternatives to alcohol. Some of them come in pill form - like pregabalin and propranolol, the ones Matt takes. And some are injected - like peptides, a type of protein that's becoming increasingly popular online.

It's part of an effort to quell social anxiety, which seems to affect Generation Z particularly sharply: more than a quarter (28%) of people in their early 20s in England report high levels of anxiety, according to a study published last year by University College London, which found that serious mental illnesses are far more common among members of Generation Z than they were among millennials at the same age.

The trend is alarming scientists - and prompting them to ask why some young adults feel they can only socialise with chemical help.

As much as I enjoy pondering the moral downfall of England over my lifetime, the article does say that the drug (which I wasn't aware of until now) is internationally popular:

 In Jordan, Ukraine and India, researchers have also noted an uptick in pregabalin abuse.

In China, the government tightened controls of the drug earlier this year by banning most online sales, and setting limits on how much of the drug can be manufactured in each province.

As to its dangers:

A study published in the journal Nature Communications estimated that global consumption of the drug more than quadrupled between 2008 and 2018.

Meanwhile, there has also been a sharp increase in deaths involving pregabalin. Office for National Statistics data show pregabalin was mentioned on 2,360 death certificates for drug-related poisonings in England and Wales between 2020 and 2024, compared with 768 between 2015 and 2019.

One of those deaths was Alex Cottam, a 27-year-old software engineer from Eccles described by his mother, Michelle, as "lovely, clever, and popular". He was prescribed pregabalin to treat his anxiety, but his mother says he "became addicted" and started boosting his supply by buying the drug from illicit websites. "I can't help thinking every day that pregabalin killed my son", she told the BBC in 2024.  

 And, of course, the "soft on drugs" school is hanging around saying "it's not so bad as it seems":

Prof David Nutt, a neuropsychopharmacologist at Imperial College London who has long campaigned for looser drug laws, argues that the drugs Matt takes may be less harmful than heavy drinking, even if they carry risks of their own.  

"Propranolol and pregabalin are undoubtedly safer than alcohol, but not perfectly safe," he says.

But he says the picture changes when pregabalin is used in large doses for intoxication (as Matt says he is doing). "If they're using it as a sort of intoxicant to get drunk, then they are at risk of dependence," he says. "It's not a free lunch in that sense."  

I will remain stubbornly old fashioned and argue that, at least with alcohol, it is well understood and (rarely) does the consumer have any good reason to be concerned over the purity of the product being consumed.   Buying tablets on the black market has always got this cloud of uncertainty over it.   

As for the rise of youthful anxiety - kind of ironic that as national comfort has increased, so has anxiety.

Bring back national service!  (Ha ha). 

  

Young women and therapy talk

Another interesting piece from today's NYT is this one:

I’m a Psychiatrist. I’m Hearing Something New From Young Women.

The writer has got a book on the topic to promote, but her basic point is this:

Several years ago, I started noticing something new when teenage girls or young women came to my office for their first psychiatric sessions. Patients used to dread being told that there was something, as they saw it, wrong with them. Now these young women were announcing their diagnoses — attention deficit hyperactivity disorder, obsessive-compulsive disorder, anxiety, depression — almost before telling me their names.

They had researched their symptoms on Instagram; they’d taken TikTok quizzes about them; they had cross-indexed their conditions with those of their friends. It was everyone — gravel-voiced film majors and earnest high school athletes, class presidents and girls who skipped school, girls who spent weekends in the Hamptons and girls who had to work for their spending money. The patients of this new wave were talking in therapy-speak, and their therapy hadn’t even started. 

Further down:

In some cases, my patients’ self-diagnoses were spot on. In many others, however, the psychiatric label seemed more like a compromise between uncomfortable feelings and the desire for social acceptance. Are these labels a source of strength or a straitjacket? Are they helping young women find help and acceptance or standing in the way? The answer isn’t so simple.

Adolescent girls and young women aren’t the only ones reaching for therapy-speak, of course. According to some measures, 50 percent of Gen Z-ers have labeled themselves with mental health conditions, and illness identities have taken hold of young people, both online and off. But girls are by far the primary drivers of this trend. They are the main group looking up health information online, talking about their feelings and relationships and seeking labels for what’s wrong.

I guess it's still a thing, so to speak, that we worry about men (especially young men?) bottling up their feelings and emotions and sometimes not undergoing any adequate self analysis, and on the other hand we worry about young women wanting to talk too much (and overanalyse) about the same thing.

I get the feeling that any school education about such things (not that I am sure there is?) needs to try to teach some sort of happy medium here:   some self reflection and recognition is good, but too much can be unhelpful in its own way.