https://ninapanickssery.com/
Views purely my own unless clearly stated otherwise
Nina Panickssery
Karma: 3,750
I also tried for a few minutes and failed to elicit any goblin talk out of 5.5
Agreed. Also good parents can tell reasonably well when their kids are mature enough to have freedoms, like for some it might be closer to 11, for others closer to 15.
My thought process is “would my kid endorse me having restricted their freedom in this way when they reflect on it as an adult”. I’d make a best-effort prediction of this based on the kid’s personality and qualities.
I don’t buy the “genuinely do not love anyone” assertion
Have you read John Wentworth’s posts…
Cool! I was actually wondering whether outdoor-only preschools were a thing today!
I, using my available tools as an admin
What is the point of this? Say you find this criminal’s LessWrong profile—what is the benefit of exposing it? On the flipside, there are downsides, namely violating norms of privacy.
How AI has changed my approach to writing
I am more likely to post “personal opinion”-style, or even emotional, content that’s not factual / entirely subjective. Since AI has made it much easier to do research and compile factual summaries of any well-understood area of human knowledge, the value of pure-fact posts has decreased (though of course there’s still value in knowing what to ask and curating / clarifying an LLM’s findings). However, AIs cannot know or convey your unique perspective on subjective matters or internal experience.
I am less likely to edit or polish my writing. I used to spend time ensuring my writing was free of grammatical errors. Now I edit much less, only enough to ensure the thoughts flow coherently and it’s easy enough for someone to read. Partly as a signal that the prose was human-written, partly because the effort put into editing is no longer a good signal of caring because editing is effectively free (you can get an LLM to do it).
I am more likely to maintain a distinctive style. Another reason I edit less is that I want to maintain a distinctive, personal style. When the first good LLMs came out I’d often use them to rephrase content, asking them to make my prose flow better or be clearer. But these days I want some clumsy constructions or phrases in my writing that are distinctively mine, I want a personal writing “smell” just like AI writing has a smell. As more people use LLMs to write, their styles become more undifferentiated, and so having your own style lets you stand out. Also when it comes to personal, non-factual content, writing in your own distinctive voice helps convey internal thoughts more precisely; some of the meaning is in the precise choice of phrasing.
The reason democracy is useful isn’t because the majority knows best
Despite the fact that we call countries like the US “democracies”, very little is decided by an actual popular vote of the entire population. Nevertheless, critiques of representative democracies, the sort of which exist in most of America and Europe, often bring up the superior decision-making ability and knowledge of elites compared to the masses. This is irrelevant. The masses decide very little and their intelligence and knowledge of policy is entirely irrelevant to their usefulness as voters. Instead, voters serve a useful purpose as a crude gradient signal—if they perceive the country as having improved during a politician’s term, they’ll vote for them again, otherwise, they’ll vote for someone else. This is the ~only role of the voters and is responsible for the main benefit of democracy—policymakers have at least some incentive to improve the country for the people since that’ll decrease their chances of being voted out.
In summary, the role of modern democracies is to improve the incentives for the actual decision-makers, and not to involve the masses in the decision-making (not only that, but to maintain the incentive, the voters just have to be vaguely aware of whether “stuff has improved” vs. “stuff has worsened”—nothing skilled or detailed).
(Of course this has issues since “making the public think stuff has improved in your term” can be Goodharted in various ways, e.g. by delaying problems; kicking the can down the road, by trying to influence public perception more than real outcomes, etc. Also the public might punish policymakers for things actually out of their control like natural disasters.)
I agree. I think a consequence of this is that making “gippities” very good at teaching people, explaining stuff, and actively involving people in the oversight process is underrated.
Ideally parents can help here by giving best-effort objective assessments.
Yes indeed, I even was thinking to write a jokepost where I reverse every point and make it sound even more convincing. Though as a dogmatic individual I am convinced people should only follow the correct version 🙃
Trigger warning: vague insight-porny content
Often the crux of a disagreement between people boils down to a starting premise, intuition, moral belief, preference, or value that they are unlikely to change (or whose justification or source comes from an entirely different realm, i.e. moral philosophy or personal emotions and psychology). It seems to me that sometimes either one or both people in the disagreement willfully (not always sure whether consciously or subconsciously) ignore this because:
they know that this disagreement is insurmountable, or
they don’t want to explicitly state the premises they’re using, perhaps because they are unpopular, or unlikely to be held by the other person, or because they are not politically correct
So instead, the argument proceeds as if the main question is one of logic or empirical evidence. But this is a waste of time, and risks confusing someone into agreeing if they don’t recognize the reliance on a premise they don’t hold.
Great article, thanks for writing about this
I think there is a tension with saying that society should lower the demands on parents, so that people have more kids
I think the only possible tension here is re. embryo selection. And it’s not a real tension. The claim is something like “if what’s giving you pause is the high demand on parents, just wing it and have kids anyway and anyhow” + “if you already know you want to have a kid and want to optimize their genes/happiness here are some ways to do it”. I think most Rationalists would agree that the life of an additional non-embryo-selected, ordinary-parented child is still worth creating. Or in other words, one set of claims is about the floor of how much effort you can put in per child and it still be a good idea to have the child. The other set of claims is about effective ways to put more effort in if you want to (mainly what’s discussed is embryo selection for health/intelligence).
I assume by “health-optimizing genetic manipulation” you mean embryo selection (seeing as gene editing is not possible yet). Indeed, Rationalists are more likely to be interested in embryo selection. And indeed, it is costly. But I’d say this is different from costly parenting—it’s a one-time upfront cost to improve your child’s genetics.
I ~never hear the 2nd thing among rationalists (“improve your kid’s life outcomes by doing a lot of research and going through complicated procedures!”).
Homeschooling is often preferred not because it substantially improves life outcomes but because it’s nicer for the children and often parents. School involves a lot of wasted time/effort, and is frustrating and boring for many children. And so by homeschooling you can make their childhood nicer irrespective of life outcomes.
I was actually thinking to make a follow-up post like this. I basically agree.
Let’s talk about two kinds of choice:
choice in the moment
choice of what kind of agent to be
I think this is the main insight—depending on what you consider the goal of decision theory, you’re thinking about either (1) or (2) and they lead to conflicting conclusions. My implicit claim in the linked post is that when describing thought experiments like Newcomb’s Problem, or discussing decision theory in general, people appear to be referring to (1), at least in classical decision theory circles. But on LessWrong people often switch to discussing (2) in a confusing way.
the core problem in decision theory is reconciling these various cases and finding a theory which works generally
I don’t think this is a core problem because in this case it doesn’t make sense to look for a single theory that does best at two different goals.
I think those other types of startups also benefit from expertise and deep understanding of the relevant topics (for example, for advocacy, what are you advocating for and why, how well do you understand the surrounding arguments and thinking...). You don’t want someone who doesn’t understand the “field” working on “field-building”.
My bad, I read you as disagreeing with Neel’s point that it’s good to gain experience in the field or otherwise become very competent at the type of thing your org is tackling before founding an AI safety org.
That is, I read “I think that founding, like research, is best learned by doing” as “go straight into founding and learn as you go along”.
I naively expect the process of startup ideation and experimentation, aided by VC money
It’s very difficult to come with AI safety startup ideas that are VC-fundable. This seems like a recipe for coming up with nice-sounding but ultimately useless ideas, or wasting a lot of effort on stuff that looks good to VCs but doesn’t advance AI safety in any way.
Wow our thinking on these topics is very similar