DirectedEvolution

Pandemic Prediction Checklist: H5N1

Pandemic Prediction Checklist: Monkeypox

I have lost my trust in this community’s epistemic integrity, no longer see my values as being in accord with it, and don’t see hope for change. I am therefore taking an indefinite long-term hiatus from reading or posting here.
 

Correlation does imply some sort of causal link.

For guessing its direction, simple models help you think.

Controlled experiments, if they are well beyond the brink

Of .05 significance will make your unknowns shrink.

Replications prove there's something new under the sun.

Did one cause the other? Did the other cause the one?

Are they both controlled by something already begun?

Or was it their coincidence that caused it to be done?

Wiki Contributions

Comments

This post and its companion have even more resonance now that I'm deeper into my graduate education and conducting my research more independently.

Here, the key insight is that research is an iterative process of re-scoping the project and execution on the current version of the plan. You are trying to make a product sufficient to move the conversation forward, not (typically) write the final word on the subject.

What you know, what resources you have access to, your awareness of what people care about, and what there's demand for, depend on your output. That's all key for the next project. A rule of thumb is that at the beginning, you can think of your definition of done as delivering a set of valuable conclusions such that it would take about 10 hours for any reasonably smart person to find a substantial flaw.

You should keep on rethinking whether the work you're doing (read: the costs you're paying) are delivering as much value, given your current state of knowledge. As you work on the project, and have conversations with colleagues, advisors and users, your understanding of where the value's at and how large the costs of various directions are, will constantly update. So you will need to update your focus along with it. Accept the interruptions as a natural, if uncomfortable, part of the process.

Remember that one way or another, you're going to get your product to a point where it has real, unique value to other people. You just need to figure out what that is and stay the course.

The advice here also helps me figure out how to interact with my fellow students when they're proposing excessively costly projects with no clear benefit due to their passion for and interest in the work itself and their love of rigor and design. Instead of quashing their passion or staying silent or being encouraging despite my misgivings, I can say something like "I think this could be valuable in the future once it's the main bottleneck to value, but I think [some easier, more immediately beneficial task] is the way to go for now. You can always do the thing you're proposing at a later time." This helps me be more honest while, I believe, helping them steer their efforts in ways that will bring them greater rewards.

The most actionable advice I got from the companion piece was the idea of making an outline of the types of evidence you'll use to argue for your claims, and get a sign-off from a colleague or advisor on the adequacy of that evidence before you go about gathering it. Update that outline as you go along. I've been struggling with this exact issue and it seems like a great solution to the problem. I'm eager to try it with my PhD advisors.

Edit: as a final note, I think we are very fortunate to have Holden, a co-founder of a major philanthropic organization, describing what his process was like during its formation. Exposition on what he's tracking in his head is underprovided generally and Holden really went above and beyond on this one. 

As of October, MIRI has shifted its focus. See their announcement for details.

I looked up MIRI's hiring page and it's still in about the same state. This kind of makes sense given the FTX implosion. But I would ask whether MIRI is unconcerned with the criticism it received here and/or actively likes their approach to hiring? We know Eliezer Yudkowsky, who's on their senior leadership team and board of directors, saw this, because he commented on it.

I found it odd that 3/5 members of the senior leadership team, Malo Bourgon, Alex Vermeer, and Jimmy Rintjema, are from Ontario (Malo and Alex at least are alumns from University of Guelph). I think this is a relevant question given that the concern here is specifically about whether MIRI's hiring practices are appropriate. I am surprised about this both because the University of Guelph, as far as I know, is not particularly renowned as an AI or AI safety research institution, and because Ontario is physically distant from San Francisco, ruling out geographic proximity as an explanation.

A bit of Googling turned up MIRI's own announcement page for Malo Bourgon's hiring as COO (he's now CEO). "Behind the scenes, nearly every system or piece of software MIRI uses has been put together by Malo, or in a joint effort by Malo and Alex Vermeer — a close friend of Malo’s from the University of Guelph who now works as a MIRI program management analyst."

I would like to understand better what professional traits made Malo originally seem like a good hire, both given that his background doesn't sound particularly AI or AI safety-focused. "His professional interests included climate change mitigation, and during his master’s, he worked on a project to reduce waste through online detection of inefficient electric motors. Malo started working for us shortly after completing his master’s in early 2012, which makes him MIRI’s longest-standing team member next to Eliezer Yudkowsky."

I'd also like to know what professional traits led to the hire of Alex Vermeer, given that both Alex and Malo were hired in 2012. Was a pre-existing friendship a factor in the hire, and if so, to what extent?

The three people from Ontario seem particularly involved in the workshop/recruiting/money aspect of the organization:

  • "Malo’s past achievements at MIRI include: coordinating MIRI’s first research workshops and establishing our current recruitment pipeline." (From the hiring announcement page)
  • For another U Guelph alumn listed on their team page, "Alex Vermeer improves the processes and systems within and surrounding MIRI’s research team and research programs. This includes increasing the quality and quantity of workshops and similar programs, implementing best practices within the research team, coordinating the technical publication and researcher recruiting pipelines, and other research support projects."
  • "Jimmy Rintjema stewards finances and regulatory compliance, ensuring that all aspects of MIRI’s business administration remain organized and secure." (This is also from the team page)

In my personal opinion, Eliezer's short response to this post and the lack of other response (as far as I can see here) suggest that MIRI may be either uninterested or incapable of managing its perception and reputation, at least on and around LessWrong. That makes me wonder how well it can realistically fulfill its new mission of public advocacy. I am also curious to know in detail how it came to be that so many people from Ontario occupy positions in senior leadership?

I replicated this review, which you can check out in this colab notebook (I get much higher performance running it locally on my 20-core CPU).

There is only one cluster of discrepancies I found between my analysis and Vaniver's: in my analysis, mating is even more assortative than in the original work:

  • Pearson R of the sum of partner stats is 0.973 instead of the previous 0.857
  • 99.6% of partners have an absolute sum of stats difference < 6, instead of the previous 83.3%.
  • I wasn't completely sure if Vaniver's "net satisfaction" was the difference of self-satisfaction and satisfaction with partner or perhaps the log average ratio. I used the difference (since theoretically self-satisfaction could be zero, which would make the ratio undefined). Average net satisfaction was downshifted from Vaniver's result. The range I found was , while Vaniver's was .

In Vaniver's analysis,  represents an adjustable correlation between a person's preferences and their own traits. Higher values of  result in a higher correspondence between one's own preferences and one's own traits.

One important impact of this discrepancy is that the transition between being on average more self-satisfied than satisfied with one's partner occurs at around  rather than , which intuitively makes sense to me, given the highly assortative result and the fact that the analysis directly mixture an initial set of preferences with some random data to form the final preferences as a function of .

Can we ground these results in empirical data, even though we can't observe preferences and stats with the same clarity and comprehensiveness in real-world data?

One way we can try is to consider the "self-satisfaction" metric we are producing in our simulation to be essentially the same thing as "self-esteem." There is a literature relating self-esteem to partner satisfaction in diverse cultures longitudinally over substantial periods of time. As we might expect, self-esteem, partner satisfaction, and marital satisfaction all seem to be interrelated.

  • Predicting Marital Satisfaction From Self, Partner, and Couple Characteristics: Is It Me, You, or Us?
    • Men and women had similar scores in personality traits of social potency, dependability, accommodation, and interpersonal relatedness.
    • Broadly, self-satisfaction, partner-satisfaction, and having traits in common are all positively associated with marital satisfaction.
  • Partner Appraisal and Marital Satisfaction: The Role of Self-Esteem and Depression
    • "Regardless of self-esteem and depression level, and across trait categories, targets were more maritally satisfied when their partners viewed them positively and less satisfied when their partners viewed them negatively."
  • The Dynamics of Self–Esteem in Partner Relationships
    • "[S]elf–esteem and all three aspects of relationship quality are dynamically intertwined in such a way that both previous levels and changes in one domain predict later changes in the other domain."
  • Relationships between self-esteem and marital satisfaction among women
    • "Marital satisfaction was found to be positively correlated with self-esteem in both cities, so that higher self-esteem was associated with greater satisfaction."
  • Development of self-esteem and relationship satisfaction in couples: Two longitudinal studies.
    • "Second, initial level of self-esteem of each partner predicted the initial level of the partners’ common relationship satisfaction, and change in self-esteem of each partner predicted change in the partners’ common relationship satisfaction. Third, these effects did not differ by gender and held when controlling for participants’ age, length of relationship, health, and employment status. Fourth, self-esteem similarity among partners did not influence the development of their relationship satisfaction. The findings suggest that the development of self-esteem in both partners of a couple contributes in a meaningful way to the development of the partners’ common satisfaction with their relationship."
  • A Mediation Role of Self-Esteem in the Relationship between Marital Satisfaction and Life Satisfaction in Married Individuals
    • "According to the findings of the study, the mediation self-esteem between the marital satisfaction and life satisfaction was statistically significant (p<.001). The whole model was significant (F(5-288)= 36.71, p<.001) and it was observed that it explained 39% of the total variance in the life satisfaction. Self-esteem was positively associated with marital satisfaction and considered one of the most important determinants of life satisfaction."

Finally, I wonder what the value of  is likely to be for participants of rationalist culture? A culture that promotes individual agency and self-improvement, that acknowledges serious challenges in our dating culture, our culture's egalitarian values, the far larger degree of control we have over ourselves than our partners, and the tendency for people to seek a self-justifying, optimistic narrative, all seem to me to point in the direction of  being high. That would suggest a rationalist culture with perhaps higher levels of self-esteem than partner-esteem. Fortunately, that says nothing at all about the absolute level of self- and partner-esteem, which I hope are on average high.

I can't disagree with Vaniver's conclusion that people are "mostly being serious" when they describe their partner as their better half. But I think the results of my reanalysis and my speculation on the value of `corr` (at least in rationalist-type culture) make me think this isn't because people are accurately appraising their partner as satisfying their own preferences better than they do themselves.

I looked around a bit more on Google Scholar (to be honest, just starting with the phrase "my better half"), and found a couple studies.

  • My Better Half: Strengths Endorsement and Deployment in Married Couples
    • "The present study focuses on married partners’ strengths endorsement and on their opportunities to deploy their strengths in the relationship, and explores the associations between these variables and both partners’ relationship satisfaction. The results reveal significant associations of strengths endorsement and deployment with relationship satisfaction, as expected. However, unexpectedly, men’s idealization of their wives’ character strengths was negatively associated with relationship satisfaction."
    • This is on a scale from 1-5 (p < .05).

 Is it me or you? An actor-partner examination of the relationship between partners' character strengths and marital quality

  • "[W]e examined the effects of three strengths factors (caring, self-control, and inquisitiveness) of both the individual and the partner on marital quality, evaluated by indices measuring marital satisfaction, intimacy, and burnout. Our findings revealed that the individual’s three strengths factors were related to all of his or her marital quality indices (actor effects). Moreover, women’s caring, inquisitiveness and self-control factors were associated with men’s marital quality, and men’s inquisitiveness and self-control factors were associated with women’s marital quality (partner effects)."

So idealizing your partner looks like a neutral-to-negative behavior. Inquisitiveness looks like a trait that both genders value. It strikes me that there are many things that you can do for your partner that they can't do for themselves - positive and negative. They can't praise or idealize themselves (or it won't come off the same way, anyway). They can't ask themselves "how was your day?" They can't give themselves a hug in a difficult moment, or if they do, it doesn't feel the same as when their partner does it.

No matter how effective you are at operating in the world, there are certain things that you just cannot do for yourself. In many areas of life, only your partner can. That seems like good reason to call them your better half.

Epistemic status: I read the entire post slowly, taking careful sentence-by-sentence notes. I felt I understood the author's ideas and that something like the general dynamic they describe is real and important. I notice this post is part of a larger conversation, at least on the internet and possibly in person as well, and I'm not reading the linked background posts. I've spent quite a few years reading a substantial portion of LessWrong and LW-adjacent online literature and I used to write regularly for this website.

This post is long and complex. Here are my loose definitions for some of the key concepts:
 

  • Outcome fixation: Striving for a particular outcome, regardless of what your true goals are and no matter the costs.
  • Addiction: Reacting to discomfort with a soothing distraction, typically in ways that cause the problem to reoccur, rather than addressing its root causes.
  • Adaptive entropy: An arms race between two opposing, mutually distrusting forces, potentially arriving at a stable but costly equilibrium.
  • Earning trust: A process that can dissolve the arms race of adaptive entropy with listening, learning how not to apply force and tolerate discomfort, prioritizing understanding the other side, and ending outcome fixation.

I can find these dynamics in my own life in certain ways. Trying to explain my research to polite but disinterested family members. Trying to push ahead with the next stage in the experiment when I'm not sure if the previous data really holds up. Reading linearly through a boring textbook even though I'm not really understanding it anymore, because I just want to be able to honestly say I read chapter 1. Arguing with almost anybody online. Refusing to schedule my holiday visits home with the idea that visits to the people I want to see will "just happen naturally."

And broadly, I agree with Valentine's prescription for how to escape the cycle. Wait for them to ask me about my research, keep my reply short, and focus my scientific energy on the work itself and my relationships with my colleagues. RTFM, plan carefully, review your results carefully, and base your reputation on conscientiousness rather than getting the desired result. Take detailed, handwritten notes, draw pictures, skim the chapter while searching for the key points you really need to know, lurk more and write what you know to a receptive audience. Plan your vacations home after consulting with friends and family on how much time they hope to spend with you, and build in time to rest and recharge.

I think Valentine's post is a bit overstated in its rejection of force as a solution to problems. There are plenty of situations where you're being resisted by an adaptive intelligence that's much weaker and less strategic than you, and you can win the contest by force. In global terms, the Leviathan, or the state and its monopoly on violence, is an example. It's a case where the ultimate victory of a superior force over all weaker powers is the one thing that finally allows everybody to relax, put down the weapons, and gain slack. Maintaining the slack from the monopoly on violence requires continuously paying the cost of maintaining a military and police force, but the theory is that it's a cost that pays for itself. Of course, if the state tries to exert power over a weaker force and fails, you get the drug war. Just because you can plausibly achieve lasting victory and reap huge benefits doesn't mean it will always work out that way.

Signaling is a second counterpoint. You might want to drop the arms race, but you might be faced with a situation where a costly signal that you're willing and able to use force, or even run a real risk a vicious cycle of adaptive entropy, is what's required to elicit cooperation. You need to make a show of strength. You need to show that you're not fixated on the outcome of inner harmony or of maintaining slack. You're showing you can drive a hard bargain, and your potential future employer needs to see that so they'll trust that you'll drive a hard bargain on their behalf if they hire you. The fact that those future negotiations are themselves a form of adaptive entropy is their problem, not yours: you are just a hired gun, a professional.

Or on the other hand, consider How to Win Friends and Influence People. This is a book about striving, about negotiating, about getting what you want out of life. It's about listening, but every story in the book is about how to use listening and personal warmth to achieve a specific outcome. It's not a book about taking stock of your goals. It's about sweetening the deal to make the deal go down.

And sometimes you're just dealing with problems of physics, information management, skill-building, and resource acquisition. Digging a ditch, finding a restaurant, learning to cook, paying the bills. These often have straightforward, "forcing" solutions and can be dealt with one by one as they arise. There is not always a need to figure out all your goals, constraints, and resources, and go through some sort of optimization algorithm in order to make decisions. You're a human, you typically navigate the world with heuristics, and fighting against your nature by avoiding outcome fixation and not forcing things is (sometimes, but not always), itself a recipe for vicious cycles of adaptive entropy.

Sometimes, vicious cycles of competition have side benefits. Sometimes, these side benefits can outweigh the costs of the competition. Workers and companies do all sorts of stupid, zero-to-negative sum behaviors in their efforts to compete in the short run. But the fact that they have to compete, that there is only so much demand to satisfy at any given time, is what motivates them to outperform. We all reap the benefit of that pressure to excel, applied over the long term.

What I find valuable in this post is searching for a more general, less violent and anthropomorphized name for this concept than "arms race." I'm not convinced "adaptive entropy" is the right one either, but that's OK. What concerns me is that it feels like the author is encouraging readers to interpret all their attempts to problem-solve through deliberate, forcing action as futile. Knowing this *may* be the case, being honest about why we might be engaged in futile behavior despite being cognizant of that, and offering alternatives all seem good. I would add that this isn't *always* the case, and it's important to have ways of exploring and testing different ways to conceptualize the problems you face in your life until you come to enough clarity on their root causes to address them productively.

I also think the attitude expressed in this post is probably underrated on LessWrong and the rationalist-adjacent world. I think that my arc as a rationalist was of increasing levels of agency, belief in my ability to bend the world to my will, willingness to define goals as outcomes and pursue them in straightforward ways, create a definition of success and then pursue that definition in order to get 70% of what I really want instead of 10%. That's a part of my nature now. Many of the problems in my daily life - navigating living with my partner, operating in an institutional setting, making smart choices on an analytical approach in collaboration with colleagues, exploring the risks and benefits associated with a potential project - generate conflicts that aren't particularly helped by trying to force things. The conflict itself points out that my true goals aren't the same thing as the outcome I was striving for when I contributed to the conflict, so conflict itself can serve an information-gathering purpose.

I'm doing something dangerous here, which is making objections to seeming implications of this post that the author didn't always directly state. The reason it's dangerous is that it can appear to the author and to others that you're making an implied claim that the author hasn't considered those implications. So I'll just conclude by saying that I don't really have any assumptions about what Valentine thinks about these points I'm making. These are just the thoughts that this post provoked in me.

I think DACs face two challenges.

  1. The cost/benefit ratio for the population of potential projects is bimodal. They're either so attractive that they have no trouble seeking donors and executors via normal Kickstarter, or so unattractive that they'll fail to secure funding with or without a DAC.
  2. Even if DACs were normal, Bob the Builder exposes himself to financial risk in order to launch one. He has to increase his funding goal in order to compensate, making the value proposition worse.

For these reasons, it's hard for me to get excited about DACs.

There is probably a narrow band of projects where DACs are make-or-break, and because you're excited about them, I think it's great if you get the funding you're hoping for and succeed in normalizing them. Prove me wrong, by all means!

Goodness gracious, the reaction to this post has made me realize that I have a fundamental disconnect with the LessWrong community’s way of parsing arguments in a way I had just not realized. I think I’m no longer interested in it or the people who post here in the way I used to be. If epistemic spot checks like this are not valued, that’s a huge problem for me. Really sad.

I’ve taken a break from LessWrong before, but I am going to take a longer one now from both LessWrong and the wider LW-associated online scene. It’s not that the issues aren’t important - it’s that I don’t trust the epistemics of many of the major voices here and I think the patterns of how posts are up and downvotes reflect values that frequently don’t accord with mine. I also don’t see hope for improving the situation.

That said, I’ve learned a lot from specific individuals and ideas on LW over the years. You know who you are. I’ll be glad to take those influences along with me wherever I find myself spending time in the future.

My main aim is just to show that Scott did not represent his quoted sources accurately. I think the Social Model offers some useful terminology that I’m happy to adopt, and I am interested in how it fits into conversations about disability. My main point of frustration is seeing how casually Scott panned it without reading his sources closely, and how seemingly uninterested so many of my readers appear to be in that misrepresentation.

I am really disappointed in the community’s response to my Contra Contra the Social Model of Disability post.

I am only familiar with the interactionist model as articulated by Scott. One difference appears to be that the Social Model carves out the category of “disability” to specifically refer to morally wrong ways that society restricts, discriminated against, or omits to accommodate impaired people. It has a moral stance built in. The Interactionist model uses “disability” as a synonym for impairment and doesn’t seem to have an intrinsic moral stance - it just makes a neutral statement that what people can or can’t do has to do with both environment and physical impairment.

Load More