7 ms·

Programming My Child

83 points by exolymph 8y ago

AstralStorm 8y ago
The property you're looking for is called robustness. An algorithm that works with predictable results from many initial starting conditions without failure modes is called robust. Biological systems, unlike computers, are usually extremely robust. It is not just one feedback, it is a highly complex set of multidimensional interacting feedback loops all around. Randomness and chaos are also crucial in some of these systems. And most of them actually are coded and not trained - but coded to be robust and adaptive. The whole staging of development is likely either a side effect of this property or an additional cause of it. You cannot truly reset any nontrivial (Esp. Nonlinear) unknown system with feedback where you don't have the direct access to states. This has nothing to do with training and everything to do with control theoretic property of observability.
- j7ake 8y ago
  The property you want is beyond robust. You actually want the algorithm to constantly go beyond its comfortable prediction range and "fail" repeatedly. That's how you learn.
  - ericdykstra 8y ago
    Indeed, not just not fragile, not just robust, but you might even say antifragile: gaining from disorder.
    - AstralStorm 8y ago
      Sometimes, but that property is hard to actually ensure. (and partly why so many people break down too - their cognition is lacking in it) Most of the bodily biological systems are not antifragile or are to a very small degree.
xyzzy123 8y ago
I feel like this is hugely glamourising the rote business of extracting value from data. I liked the story and did feel that it covered well the human aspect, but it was vastly more sympathetic towards our, uh, “digital children” than i think is really warranted. The algorithms will do surprising things, partly because the world is surprising, but mostly as long as that is sufficiently valuable. Otherwise they’ll get replaced by the next person with a jupityr notebook who can optimise better for $objective_function. Or maybe they have enough political control to redefine the objective. The outputs of these things are superficially mysterious but for all practical purposes are dictated by the values of the people who create them. It’s not legal to lobotomise your child but that’s exactly what will happen to any “algorithm” which hurts the bottom line.
tw1010 8y ago
Are there any good books about ideas or a history regarding how societal values and technological tools of the parent generation has influenced their children? How did the philosophical insights of e.g. Kant or Humboldt influence western consiousness, and how did that influence the growing up period of the generation that followed, and what were some reasonable consequential attitude updates when those children grew up into adults? It'd be interesting to see if there are models or any reasonable patterns that can be seen in the adult world view -> child upbringing and subsequent next-iteration adult world view, and what that might say could happen (or might not happen, contrary to what we suspect will happen) about the current generation of gamified children growing up with trump in the background.
mxfh 8y ago
What a weird way to run all the way with an ill-fitting analogy just because; Google. If anything parenting is running DevOps with a stubborn dev-team.
brianzelip 8y ago
> In the first months of her life, I kept a spreadsheet of my daughter’s milestones. I’d love to see some spreadsheets to read other parents’ column headers beyond weight, height, head size.
- allcentury 8y ago
  This is a slippery slope, most parents already do heavy comparisons with their child to others. When there are differences and the parent isn't educated on norms you introduce a lot of self doubt and anxiety.
- docker_up 8y ago
  It would be largely useless. Most children's development is extremely varied. Books like to use average values but they never talk about how wide the standard deviation. It's so wide that an average is actually a useless value. Each child is completely different, so knowing the milestones is basically useless. My son was slow to smile, and very slow to walk. But he was evaluated with a 160 IQ at age 4. I have a friend who didn't talk until he was 5 years old. His parents were afraid he was developmentally delayed. In fact, he went to Berkeley for undergrad and Stanford for his PhD and you never would have guessed he didn't talk until he was 5. So development speed is extremely wide standard deviation to the point where data from the first several years is fun if you want to brag but useless for the purpose of generating any useful data on whole, except for the severely delayed.
le-mark 8y ago
Something about this makes my skin crawl; conflating blossoming sentience with soulless machines. Children are banal and miraculous at the same time. The default human state was to create tiny humans without end until the advent of birth control 100(?) years ago. One thing that's remarkable to me was how divorced I came from children as an adult. I was 40 when we had our first and although I'd been around neices and nephews years earlier, I'd forgotten so much. After having a second, another remarkable thing is how different they are. The younger one does and says things the older one never would have. Children aren't for everyone, and not everyone wants or can have them. But there is something distinctive they add to life.
- SketchySeaBeast 8y ago
  > One thing that's remarkable to me was how divorced I came from children as an adult. My wife and I are in our early 30's and we're struggling with the thought of children because of this - they're utterly alien to us. I feel on an intellectual level that I think I'd like to have children and like to be a father, but with the exception of one friend (who because they have kids we've largely lost contact with), we don't really have any peers that have children, so we don't ever interact with them, and it's all terribly abstract to us.
  - emodendroket 8y ago
    This seems like one of those things where no matter how prepared you think you are it'll be nothing like what you expect.
    - castlecrasher2 8y ago
      My experience has been 1) don't think about it too much beforehand and 2) if you have empathy to see situations from their perspective you'll be able to look up and learn what you need.
  - alexpotato 8y ago
    I had my first (of two) kid when I was 35 and a couple things helped a lot: - I was one of the older cousins on both sides of big families so I got to see a lot of little kids running around. - I read this amazing post from Jeff Atwood: https://blog.codinghorror.com/on-parenthood/ https://blog.codinghorror.com/on-parenthood/ - I watched this talk: https://www.ted.com/talks/rufus_griscom_alisa_volkman_let_s_talk_parenting_taboos?language=en https://www.ted.com/talks/rufus_griscom_alisa_volkman_let_s_... - I read a good quote that went something like this: "I asked my dad if I was ready for kids and he said 'If you are worried having kids will mess up your life, then you're not ready because you are still thinking about yourself. If you are worried that you might mess up THEIR life, then you are probably ready because it means you are already thinking about them.' " Children definitely change your life and are a lot of work. That being said, the first time they smile at you or call you "Dada" or "Mama" it feels like you won the emotional lottery. Then each milestone big or small just keeps adding to that feeling.
trukterious 8y ago
>As with children, we don’t debug these networks; we educate them Mostly we aren't educating young children. Rather, they are the recipients of memes (e.g. words, waving). These are ways of behaving in context. The transmission is mostly unintentional and unavoidable. It has been speculated that if we tried to educate children how to walk and how to talk then they wouldn't succeed at either! Yet they succeed and improve these skills indefinitely. By contrast, arithmetic is intentionally and exhaustively taught, yet most adults cannot perform it. >But you cannot reset a human being. https://www.youtube.com/watch?v=DMKcO-T5Y4o https://www.youtube.com/watch?v=DMKcO-T5Y4o Sure you can. See Tom Scott's face around 3:30. It's a slightly unpleasant facial sequence that I'd only ever witnessed in babies previously; a sort of soft reset. Another form of reset is evident when addicts are administered with ibogaine.
- craftyguy 8y ago
  > By contrast, arithmetic is intentionally and exhaustively taught, yet most adults cannot perform it. The adults that can perform it practice it regularly. Would those same children that can talk continue to know how to do it if they were exposed to it briefly and then never given the opportunity to do it again? The social aspect of talking (understanding context, etc) requires a lot of practice, and regular exercising of it (just like arithmetic..)
- sjg007 8y ago
  We do teach children how to walk and talk. These disciplines are known as PT, OT and speech therapy.
  - monadgonad 8y ago
    Oh come on, a speech therapist isn't needed to get the vast majority of children talking. You know what they meant.
    - AstralStorm 8y ago
      We still teach kids to talk... By talking at them all the time. This includes mute and deaf people who will talk at kids using signs. These kids which are not mute or deaf will learn to speak much later since most others will still talk at them. The hardwired part is the sign, grammar, logic and meaning. These essentially happen over development with no intervention. We do not teach kids logic - they already have it and we teach them formal reasoning and that was not invented but discovered.
shalmanese 8y ago
Keep in mind this is the same David Auerbach who passed along a bunch of intimate details of prominent feminists to Milo Yiannapoulous so he could smear them: https://www.pastemagazine.com/articles/2017/10/slates-david-auerbach-alleged-breitbart-ratfink-te.html https://www.pastemagazine.com/articles/2017/10/slates-david-...

83 points by exolymph 8y ago

The property you're looking for is called robustness. An algorithm that works with predictable results from many initial starting conditions without failure modes is called robust. Biological systems, unlike computers, are usually extremely robust. It is not just one feedback, it is a highly complex set of multidimensional interacting feedback loops all around. Randomness and chaos are also crucial in some of these systems. And most of them actually are coded and not trained - but coded to be robust and adaptive. The whole staging of development is likely either a side effect of this property or an additional cause of it. You cannot truly reset any nontrivial (Esp. Nonlinear) unknown system with feedback where you don't have the direct access to states. This has nothing to do with training and everything to do with control theoretic property of observability.

j7ake 8y ago
The property you want is beyond robust. You actually want the algorithm to constantly go beyond its comfortable prediction range and "fail" repeatedly. That's how you learn.
- ericdykstra 8y ago
  Indeed, not just not fragile, not just robust, but you might even say antifragile: gaining from disorder.
  - AstralStorm 8y ago
    Sometimes, but that property is hard to actually ensure. (and partly why so many people break down too - their cognition is lacking in it) Most of the bodily biological systems are not antifragile or are to a very small degree.

xyzzy123 8y ago

I feel like this is hugely glamourising the rote business of extracting value from data. I liked the story and did feel that it covered well the human aspect, but it was vastly more sympathetic towards our, uh, “digital children” than i think is really warranted. The algorithms will do surprising things, partly because the world is surprising, but mostly as long as that is sufficiently valuable. Otherwise they’ll get replaced by the next person with a jupityr notebook who can optimise better for $objective_function. Or maybe they have enough political control to redefine the objective. The outputs of these things are superficially mysterious but for all practical purposes are dictated by the values of the people who create them. It’s not legal to lobotomise your child but that’s exactly what will happen to any “algorithm” which hurts the bottom line.

tw1010 8y ago

Are there any good books about ideas or a history regarding how societal values and technological tools of the parent generation has influenced their children? How did the philosophical insights of e.g. Kant or Humboldt influence western consiousness, and how did that influence the growing up period of the generation that followed, and what were some reasonable consequential attitude updates when those children grew up into adults? It'd be interesting to see if there are models or any reasonable patterns that can be seen in the adult world view -> child upbringing and subsequent next-iteration adult world view, and what that might say could happen (or might not happen, contrary to what we suspect will happen) about the current generation of gamified children growing up with trump in the background.

mxfh 8y ago

What a weird way to run all the way with an ill-fitting analogy just because; Google. If anything parenting is running DevOps with a stubborn dev-team.

brianzelip 8y ago

> In the first months of her life, I kept a spreadsheet of my daughter’s milestones. I’d love to see some spreadsheets to read other parents’ column headers beyond weight, height, head size.

allcentury 8y ago
This is a slippery slope, most parents already do heavy comparisons with their child to others. When there are differences and the parent isn't educated on norms you introduce a lot of self doubt and anxiety.
docker_up 8y ago
It would be largely useless. Most children's development is extremely varied. Books like to use average values but they never talk about how wide the standard deviation. It's so wide that an average is actually a useless value. Each child is completely different, so knowing the milestones is basically useless. My son was slow to smile, and very slow to walk. But he was evaluated with a 160 IQ at age 4. I have a friend who didn't talk until he was 5 years old. His parents were afraid he was developmentally delayed. In fact, he went to Berkeley for undergrad and Stanford for his PhD and you never would have guessed he didn't talk until he was 5. So development speed is extremely wide standard deviation to the point where data from the first several years is fun if you want to brag but useless for the purpose of generating any useful data on whole, except for the severely delayed.

le-mark 8y ago

Something about this makes my skin crawl; conflating blossoming sentience with soulless machines. Children are banal and miraculous at the same time. The default human state was to create tiny humans without end until the advent of birth control 100(?) years ago. One thing that's remarkable to me was how divorced I came from children as an adult. I was 40 when we had our first and although I'd been around neices and nephews years earlier, I'd forgotten so much. After having a second, another remarkable thing is how different they are. The younger one does and says things the older one never would have. Children aren't for everyone, and not everyone wants or can have them. But there is something distinctive they add to life.

SketchySeaBeast 8y ago
> One thing that's remarkable to me was how divorced I came from children as an adult. My wife and I are in our early 30's and we're struggling with the thought of children because of this - they're utterly alien to us. I feel on an intellectual level that I think I'd like to have children and like to be a father, but with the exception of one friend (who because they have kids we've largely lost contact with), we don't really have any peers that have children, so we don't ever interact with them, and it's all terribly abstract to us.
- emodendroket 8y ago
  This seems like one of those things where no matter how prepared you think you are it'll be nothing like what you expect.
  - castlecrasher2 8y ago
    My experience has been 1) don't think about it too much beforehand and 2) if you have empathy to see situations from their perspective you'll be able to look up and learn what you need.
- alexpotato 8y ago
  I had my first (of two) kid when I was 35 and a couple things helped a lot: - I was one of the older cousins on both sides of big families so I got to see a lot of little kids running around. - I read this amazing post from Jeff Atwood: https://blog.codinghorror.com/on-parenthood/ https://blog.codinghorror.com/on-parenthood/ - I watched this talk: https://www.ted.com/talks/rufus_griscom_alisa_volkman_let_s_talk_parenting_taboos?language=en https://www.ted.com/talks/rufus_griscom_alisa_volkman_let_s_... - I read a good quote that went something like this: "I asked my dad if I was ready for kids and he said 'If you are worried having kids will mess up your life, then you're not ready because you are still thinking about yourself. If you are worried that you might mess up THEIR life, then you are probably ready because it means you are already thinking about them.' " Children definitely change your life and are a lot of work. That being said, the first time they smile at you or call you "Dada" or "Mama" it feels like you won the emotional lottery. Then each milestone big or small just keeps adding to that feeling.

trukterious 8y ago

>As with children, we don’t debug these networks; we educate them Mostly we aren't educating young children. Rather, they are the recipients of memes (e.g. words, waving). These are ways of behaving in context. The transmission is mostly unintentional and unavoidable. It has been speculated that if we tried to educate children how to walk and how to talk then they wouldn't succeed at either! Yet they succeed and improve these skills indefinitely. By contrast, arithmetic is intentionally and exhaustively taught, yet most adults cannot perform it. >But you cannot reset a human being. https://www.youtube.com/watch?v=DMKcO-T5Y4o https://www.youtube.com/watch?v=DMKcO-T5Y4o Sure you can. See Tom Scott's face around 3:30. It's a slightly unpleasant facial sequence that I'd only ever witnessed in babies previously; a sort of soft reset. Another form of reset is evident when addicts are administered with ibogaine.

craftyguy 8y ago
> By contrast, arithmetic is intentionally and exhaustively taught, yet most adults cannot perform it. The adults that can perform it practice it regularly. Would those same children that can talk continue to know how to do it if they were exposed to it briefly and then never given the opportunity to do it again? The social aspect of talking (understanding context, etc) requires a lot of practice, and regular exercising of it (just like arithmetic..)
sjg007 8y ago
We do teach children how to walk and talk. These disciplines are known as PT, OT and speech therapy.
- monadgonad 8y ago
  Oh come on, a speech therapist isn't needed to get the vast majority of children talking. You know what they meant.
  - AstralStorm 8y ago
    We still teach kids to talk... By talking at them all the time. This includes mute and deaf people who will talk at kids using signs. These kids which are not mute or deaf will learn to speak much later since most others will still talk at them. The hardwired part is the sign, grammar, logic and meaning. These essentially happen over development with no intervention. We do not teach kids logic - they already have it and we teach them formal reasoning and that was not invented but discovered.

shalmanese 8y ago

Keep in mind this is the same David Auerbach who passed along a bunch of intimate details of prominent feminists to Milo Yiannapoulous so he could smear them: https://www.pastemagazine.com/articles/2017/10/slates-david-auerbach-alleged-breitbart-ratfink-te.html https://www.pastemagazine.com/articles/2017/10/slates-david-...