See Why Neuron Counts Shouldn’t Be Used as Proxies for Moral Weight and maybe also Is Brain Size Morally Relevant?
saulius
Karma: 11
saulius’s Shortform
My favourite way to use LLMs is to write personalised fiction. For example:
1. I’m trying to decide whether to have a child. It helps to have LLMs write stories about how my life as a parent might feel from inside under various circumstances (e.g., the child is autistic)
2. For my Dad’s birthday, I wrote a story with LLMs where his favourite book character (Schweik) goes to my Dad’s village and gets into my Dad’s hobbies. Maybe I’ll write a sequel where the book character meets my Dad. LLMs did a great job of capturing the humour of the original book.
Thanks for clarifying. If you ever pitch your ideas to potential investors or something, I recommend avoiding talking about hundreds of embryos, or at least acknowledging that this is unrealistic with current technologies before doing so. When reading, I was a bit worried that you might be divorced from reality, thinking in sci-fi terms, not knowing the basic realities about IVF. This made it difficult for me to trust other things you were saying about domains I know nothing about. Just letting you know in case it’s helpful :)
I’ve just started reading and this seems very interesting and important. However, I find the discussion about embryos and scaling odd. I mean sentences like “If we had 500 embryos”. Here is some quick info for women under 35, generated by ChatGPT:
A single egg collection usually retrieves 8-14 eggs. Out of those, only 4-6 embryos typically develop far enough to be tested, and about 50-60% of those will be genetically normal. This means that in most cases, only 2-4 embryos per cycle are actually viable for implantation.
Even in the best-case scenario, only about 50-55% of embryo transfers lead to a live birth.
Egg collection and embryo transfer aren’t easy. Women have to inject hormones daily for weeks, go through a minor surgery to retrieve eggs, and deal with bloating, pain, and possible complications. It’s also expensive—one cycle can cost $10,000-$20,000. I doubt many women would go through dozens of rounds just to produce 100+ embryos.
I think this makes embryo selection even less promising than you portrayed. I’m confused about how this affects your analysis of gene editing. Or maybe I just don’t understand why you talk about hundreds of embryos because I haven’t read the full text.
reminds me of this