Nitpicking at the example, worker bees do not have offspring; the best way for them to spread their genes is to protect the queen and thus, the hive.

Birds can have offspring, so self-preservation instead of risky attacks is optimal for individuals of a flock (of genetically unrelated individuals).

It's not that the group is less intelligent, rather that the individuals of the group have different goals (self-preservation vs hive preservation, though the end goal of maximizing fitness is the same).

But genetic fitness breaks down as a metric when you add culture to the system, so application to humans is limited.

It is important to note that people have a wide range of attachment to their gender identity, ranging from willing to undergo extreme body modification in order to match their gender identity, to those who don't care in the slightest.

The issue is that cisgender is the default, and if you don't have a strong attachment to your gender identity, you have no reason to change the label. Hence, cisgendered people have a wide range of attachment to their gender identity, from strongly identifying with it to no attachment at all.

(There is also the group of agender, which includes those who have deeply examined their gender identity and decided that they don't really care (and probably also want to signal their examination and non-caring of gender identity))

Someone who is transgender obviously has an attachment to their gender identity, and this is obviously from which the Pronoun Discourse stems. They have a strong preference for a gender, and a preference to be referred to with the appropriate pronouns, and thus being misgendered is upsetting, as their preferences are violated.

(Of course, most of this rests on the ability to communicate the preference, and accidental violations when the preference was not communicated are less egregious than deliberate violations.)

Otherwise misgendering can be upsetting if it is tied to stereotypes of masculinity and femininity and attempting an insult based off those stereotypes.

Incarceration plays three roles (to varying degrees of success): punishment (and therefore deterrent), rehabilitation and exclusion from society.

One group of people would prefer focusing on rehabilitation rather than punishment, and are likely those who oppose solely serving vegan food to prisoners. Another group of people sees prisoners as no longer human and deserving of moral concern, and think that the cruelty of prison is the point.

The US prison system leans more towards the latter than the former (see: mandatory prison labor), though other places in the world lean differently. How far and in what direction prison systems should lean is a topic for philosophical debate.

Though also note that degree of punishment doesn't really correlate with deterrence from crime (review Apel and Nagin (2011), tl;dr: more about certainty of getting caught, it's complicated)

I think that sentence is required for a complete logical specification of the question.

But by removing that sentence, GPT3.5 still responds popcorn.

Edit: I think the key change is "looks at the bag".

As a human*, I also thought chocolate.

I feel like an issue with the prompt is that it's either under- or overspecified.

Why does it matter if Sam has seen the bag before? Does Sam know the difference between chocolate and popcorn? Does Sam look at the contents of the bag, or only the label?

Revised Prompt:
A sealed bag contains popcorn. There is no chocolate in the bag. The bag is transparent, and its contents can be seen without opening it. There is a label on the bag that reads "chocolate".

Sam can differentiate between chocolate and popcorn. Sam looks at the bag and reads the label. She believes the bag is full of

--

I've tested ChatGPT 3.5 and it works on this revised prompt.

If you check the moderation logs, Roko deleted a recent comment, which probably garnered the downvotes that lead to the rate-limiting.

Good post. This looks possible, if not feasible.

"crazy, unpredictable, and dangerous" are all "potentially surmountable issues". It's just that we need more research into them before they stop being crazy, unpredictable, and dangerous. (except quantum I guess)

I think that most are focusing on single-gene treatments because that's the first step. If you can make a human-safe, demonstrably effective gene-editing vector for the brain, then jumping to multiplex is a much smaller step (effective as in does the edits properly, not necessarily curing a disease). If this were a research project I'd focus on researching multiplex editing and letting the market sort out vector and delivery.

I am more concerned about the off-target effects; neurons still mostly function with a thousand random mutations, but you are planning to specifically target regions that have a supposed effect. I would assume that most effects in noncoding regions are regulator binding sites (alternately: ncRNA?), which are quite sensitive to small sequence changes. My assumption would be a higher likelihood of catastrophic mutations (than you assume).

Promoters have a few of important binding motifs whose spacing is extremely precise, but most of the binding motifs are a lot more flexible in how far away they are from each other.

Also, given that your target is in nonreplicating cells, buildup of unwanted protein might be an issue if you're doing multiple rounds of treatment.

The accuracy of your variant data could/should be improved as well; most GWAS-based heritability data assumes random mating which humans probably don't do. But if you're planning on redoing/rechecking all the variants that'd be more accurate.

Additionally, I'm guessing a number of edits will have no effect as their effect is during development. If only we had some idea how these variants worked so we can screen them out ahead of time. I'm not sure what percent of variants would only have an effect during development, so you'll need to do a lot more edits than strictly necessary and/or a harder time detecting any effects of the edits. Luckily, genes that are always off are more likely to be silenced, so they might be harder to edit.

Though I would avoid editing unsilenced genes anyways, because they're generally off and not being expressed (and therefore less likely to have a current effect) and the act of editing usually unsilences the genes for a bit, which is an additional level of disruption you probably don't want to deal with.

I don't know how the Biobank measures "intelligence" but make sure it corresponds with what you're trying to maximize [insert rehash of IQ test accuracy].

Finally, this all assumes that intelligence is a thing and can be measured. Intelligence is probably one big phase space, and measurements capture a subset of that, confounded by other factors. But that's getting philosophical, and as long as it doesn't end up as eugenics (Gattaca or Hitler) it's probably fine.

Honestly just multiplex editing by itself would be useful and impressive, you don't have to focus on intelligence. Perhaps something like muscle strength or cardiovascular health would be an easier sell.