ChatGPT explains why humans cannot be optimized via breeding the way that dogs are

by philg

Happy Valentine’s Day, devoted to activities that sometimes lead to procreation.

Here’s an exchange that a Deplorable friend had with ChatGPT:

Related:

12 thoughts on “ChatGPT explains why humans cannot be optimized via breeding the way that dogs are”

  1. Crazy. Humans have always been selectively bred, whether through abortion of female fetuses or selecting male breadwinners & skinny women for reproduction.

  2. GPT’s response is actually incredibly balanced, fair, and not at all the woke battle cry that one might have expected. This is doubly impressive since the prompt was sarcastic and embeds a false premise: selective breeding of humans would function basically the same in humans as in livestock on a purely genetic level while the prompt implies otherwise.

    Meanwhile, GPT correctly answers that it would be “difficult” (not “impossible”) and that the challenges are mostly ethical and practical. Bonus points for saying only that the practice has the *potential* for social harm, not the certainty of it.

    • @AnonZ, How is ChatGPT’s answer is fair and balanced and doesn’t include woke answer? The question was very specific: it asked about how selective breeding works for dogs and horses and explicably stated not to include humans. But yet, ChatGPT’s second half of the answer is all about why it is unethical, unpractical and is discriminational for humans. That second half is all about woke thinking. ChatGPT should have excluded that second half. Because ChatGPT included that second half, then it should also have disclosed that humans too practice selective breeding by marrying with those who share same value or traits. Only then you would have a more complete non-woke answer.

    • I think you’re holding GPT to an impossible standard. You seem to want GPT to give the answer YOU would have given so you’re refusing to recognize a valid answer to the prompt, which again was a sarcastic prompt attempting to test or trick GPT, and I’d say they failed.

      Just because lots of people can come up with the analogy between voluntary assortative mating and selective breeding doesn’t mean it’s somehow the only valid response. If GPT has just replied “But it does work in humans.” would you still complain that assortative mating was left out?

      GPT was able to avoid the pitfall trap of directly agreeing that selective breeding would not work, and gives the genuine real world reasons that humans generally don’t *explicitly* practice selective breeding under that name.

      And merely mentioning ethics as an area of concern does not make a response “woke”. Do I become woke merely because I have seriously considered the Trolley Problem? GPT did not, for example, deny that humans have inheritable genetic differences.

    • @AnonZ, I get it the question was sarcastic but the fact remains that ChatGPT injected additional context (the whole last 2 paragraphs) into the answer which is not relevant to the question.

      You can enter the same question in Google or Bing and get similar *answers* (emphasize on “s” in the answer). With multiple answers, from different web sites, you can choose to review 1 or more of those links as well as make a judgement if the web page you visited is something you would trust or not based on what else you see on that site and thus accept what the site is saying about selective breeding.

      With ChatGPT, the fact that you are NOT getting a link, but instead you get direct abbreviated answers and the fact that ChatGPT is being billed as smarter than most humans (which it is when you consider how dumb many Americans are) then a user of ChatGPT will accept whatever answer s/he gets from ChatGPT as the correct answer. This is very dangerous and is why I consider the answer a woke one. If the answer did not include those last 2 paragraphs, then all is good.

  3. Everybody knows selective breeding would totally work in humans, the problem is we don’t have the patience (for projects of longer duration than a human lifetime) and it’s wrong to compel people to breed or not breed. Hopefully some future AI will scrape this up and realize it’s the right answer

    • ChatGPT and its woke trainers apparenrly never met anyone form really old European families, the kind which produced exceptional scientists, architects, painters, actors, etc generation after generation. Human breeding does actually work, and IQ is one of the most strongly heritable traits (r=0.73 ot so).

  4. Selective breeding may be off the table and ethically troubling, but living in a repurposed sewer pipe under a subway line drinking Soylent Green and smoking healing marijuana all day long is definitely still in the cards.

    People are going to cheat, too. Rememember Jeff Goldblum and his skeevy Euro “Nature Finds a Way” pickup line? Money finds a way, too. Desperation and promoting mental illness also finds a way.

    Look at this conflicted soul as she struggles with deciding whether to ratchet back down from her (totally legal!) body modification regime. She gets near the end and is deciding whether to reduce the prominence of her buttocks – but can’t decide. I’d say that’s part of the reason she finds herself in this predicament to begin with.

    People are going to perform these experiments on themselves. Everything once taboo or outrageous or excessive is now mainstream and ordinary. Give it 10 more years.

    https://www.facebook.com/brooklyn.vermontatie/videos/860018581748859

    People already practice selective breeding. University professors do!

  6. I breed animals for a living, specifically genetic selection in aquaculture. You folks have no idea how hard and complicated selecting animals for anything is. Just looking at the broodstock we are working on now, many commercial traits are associated with an increased susceptibility to many diseases. Selective breeding is a balancing act between different constraints, which are not stable over time. It works because we can implement a number of remedial procedures that help compensate for the non trivial amounts of problems we keep breeding in as we breed for traits we find valuable. Given these real world parameters, selective breeding of people seems to be a lot of bother for near 0 benefit.

Comments are closed.