This post is a perfect reminder of why you’re so great, Derek: You find important stuff, and you contextualize it masterfully. Missed reading you! Also, congrats and welcome back!
lol i had to click in to see how you could possibly find the time to READ!!! but then i saw audiobooks. everything is on audiobooks now since my son was born 3 years ago. the eyes can only handle so much!!
Thanks for a wonderful piece; as always, your dissemenation of all the news, writing, and themes really lends a balanced perspective. I feel I learn much, think deeply, introspect, and most of all, feel more grounded and hopeful after reading your work. Thanks again and congrats on the new baby!
Welcome back Derek, you were missed! And congrats on the Dad-dom.
Regarding AI and science. I already see the effect in academic hiring. We have talented young people working in established fields, who will tell you that in practice AI can't be applied to their problem for lack of data. Their work is interesting and important, and they get passed over for someone using AI to innovate in a data heavy field that is increasingly crowded. Of course, there has always been herd following.
In counterpoint, I see the AI revolution as being accompanied by a revolution in Mathematics. A whole slew of new math tools and language were developed (and open source coded) to enable the required AI development. I would say that every time a new branch of mathematics has been discovered, shortly thereafter it has proven useful to model something that was previously not understood... IOW to lead to new scientific insights outside of math.
So I entirely expect a wave of deep, fundamental science insights to be coming shortly, not empowered by AI, but enabled by the mathematical tools used to build AI.
Some surprising and interesting stats in here! Looking forward to seeing more analysis on #10 (paging @noahpinion).
On #11, I think that’s a bit too simple an analogy. There may be some spaces where you want to completely exclude AI—perhaps to try and see what your brain can come up with without being directly assisted or influenced by AI. But to me it’s more about how you use AI rather than if you use AI.
The way I see it, there are two primary ways of using AI—1) to have it complete tasks for you and 2) to have it help you raise the bar, that is to accomplish more than you possibly could on your own.
(1) is extremely useful for those tasks I don’t care to improve independent of AI (much of my housework, personal and professional organization and admin) and frees me up to spend more time doing the things that make me feel most alive!
(2) can help me and my collaborators do the things we love in ways we never thought possible, and to move quickly from just planning to actually executing and building!
Now, just as with any other technology, I think it’s healthy to be able to disconnect in some spaces and times. Just as sometimes I sit around with friends and family and we wonder about certain things that are completely knowable with a quick internet search, but sometimes we just don’t. More fun staying present, paying the price of ignorance to avoid the distractions and just focus on the curiosity and wonder that we share, allowing our minds to run and the jokes to flow.
I’ll try and think of a better analogy for you. Maybe a starter for my own blog post! Thanks for getting the mind running!
Ok now I'm quickly feeling that tension--I start thinking about an alternative analogy... but I feel my mind being pulled towards the ChatGPT and Claude sites. As someone who loves to just spend time thinking about all sorts of things in the world around me (hence why I love your work), LLMs have become incredible, always on, never tiring thinking partners. They're not as good as my real life thinking partners among friends, family, and colleagues, but they're not always around or in the mood. Those sycophantic robots are pretty good at stretching my thinking and filling my knowledge gaps. Ahhhh! Good thing to reflect on for sure...
Welcome back! I was going to joke that you hadn't missed much. Then I realized that if you had managed a news blackout and went to NYT today, you would have no idea that we seemed close to invading Greenland, and that we did (briefly) invade Venezuela to capture their President! Already out of the news cycle. Wild times. Glad you're back to help us make some sense of it all.
Based on title I thought this was going to be an essay on ideas ABOUT paternity leave. One thought on the article, on thought on paternity leave:
1. As people shift toward science that already has large data sets it will become more valuable and unique to create data. That will balance science out.
2. Paternity leave is the answer to ending the male/female income gap. The 'problem' with WOCP (women of childbearing potential) is that you have to accept that there is a high probability that at some point they will suddenly be gone for 3 months, so you take a huge additional risk putting them in positions of extreme importance/responsibility instead of a man. This is where the income gap comes from. The easy fix is to make paternity leave mandatory - not available, but mandatory. That way there is just as much risk of men being gone for 3 months as there is of women. Of course, the decline in birth rates may also change this calculus.
> scientific attention is concentrated on a smaller set of problems and fields that happen to have more available data, while less-charted fields are comparatively ignored
To me this seems like a good thing since research in fields with very little data was probably junk to begin with. I hope it leads to more data collection in the long run though and not just abandoning interesting research areas which are currently data-poor
This post is a perfect reminder of why you’re so great, Derek: You find important stuff, and you contextualize it masterfully. Missed reading you! Also, congrats and welcome back!
Glad you are back and congratulations on the baby. Your insights were missed.
You nailed it with this one. Congratulations, but I’m thankful you’re back.
lol i had to click in to see how you could possibly find the time to READ!!! but then i saw audiobooks. everything is on audiobooks now since my son was born 3 years ago. the eyes can only handle so much!!
Thanks for a wonderful piece; as always, your dissemenation of all the news, writing, and themes really lends a balanced perspective. I feel I learn much, think deeply, introspect, and most of all, feel more grounded and hopeful after reading your work. Thanks again and congrats on the new baby!
Welcome back Derek, you were missed! And congrats on the Dad-dom.
Regarding AI and science. I already see the effect in academic hiring. We have talented young people working in established fields, who will tell you that in practice AI can't be applied to their problem for lack of data. Their work is interesting and important, and they get passed over for someone using AI to innovate in a data heavy field that is increasingly crowded. Of course, there has always been herd following.
In counterpoint, I see the AI revolution as being accompanied by a revolution in Mathematics. A whole slew of new math tools and language were developed (and open source coded) to enable the required AI development. I would say that every time a new branch of mathematics has been discovered, shortly thereafter it has proven useful to model something that was previously not understood... IOW to lead to new scientific insights outside of math.
So I entirely expect a wave of deep, fundamental science insights to be coming shortly, not empowered by AI, but enabled by the mathematical tools used to build AI.
Some surprising and interesting stats in here! Looking forward to seeing more analysis on #10 (paging @noahpinion).
On #11, I think that’s a bit too simple an analogy. There may be some spaces where you want to completely exclude AI—perhaps to try and see what your brain can come up with without being directly assisted or influenced by AI. But to me it’s more about how you use AI rather than if you use AI.
The way I see it, there are two primary ways of using AI—1) to have it complete tasks for you and 2) to have it help you raise the bar, that is to accomplish more than you possibly could on your own.
(1) is extremely useful for those tasks I don’t care to improve independent of AI (much of my housework, personal and professional organization and admin) and frees me up to spend more time doing the things that make me feel most alive!
(2) can help me and my collaborators do the things we love in ways we never thought possible, and to move quickly from just planning to actually executing and building!
Now, just as with any other technology, I think it’s healthy to be able to disconnect in some spaces and times. Just as sometimes I sit around with friends and family and we wonder about certain things that are completely knowable with a quick internet search, but sometimes we just don’t. More fun staying present, paying the price of ignorance to avoid the distractions and just focus on the curiosity and wonder that we share, allowing our minds to run and the jokes to flow.
I’ll try and think of a better analogy for you. Maybe a starter for my own blog post! Thanks for getting the mind running!
Ok now I'm quickly feeling that tension--I start thinking about an alternative analogy... but I feel my mind being pulled towards the ChatGPT and Claude sites. As someone who loves to just spend time thinking about all sorts of things in the world around me (hence why I love your work), LLMs have become incredible, always on, never tiring thinking partners. They're not as good as my real life thinking partners among friends, family, and colleagues, but they're not always around or in the mood. Those sycophantic robots are pretty good at stretching my thinking and filling my knowledge gaps. Ahhhh! Good thing to reflect on for sure...
Welcome back! I was going to joke that you hadn't missed much. Then I realized that if you had managed a news blackout and went to NYT today, you would have no idea that we seemed close to invading Greenland, and that we did (briefly) invade Venezuela to capture their President! Already out of the news cycle. Wild times. Glad you're back to help us make some sense of it all.
On the longevity charts: what I see that all four categories are reverting close
to pre COVID levels, notable but not that encouraging. (and those offset y axis are to be discouraged!)
Based on title I thought this was going to be an essay on ideas ABOUT paternity leave. One thought on the article, on thought on paternity leave:
1. As people shift toward science that already has large data sets it will become more valuable and unique to create data. That will balance science out.
2. Paternity leave is the answer to ending the male/female income gap. The 'problem' with WOCP (women of childbearing potential) is that you have to accept that there is a high probability that at some point they will suddenly be gone for 3 months, so you take a huge additional risk putting them in positions of extreme importance/responsibility instead of a man. This is where the income gap comes from. The easy fix is to make paternity leave mandatory - not available, but mandatory. That way there is just as much risk of men being gone for 3 months as there is of women. Of course, the decline in birth rates may also change this calculus.
> scientific attention is concentrated on a smaller set of problems and fields that happen to have more available data, while less-charted fields are comparatively ignored
To me this seems like a good thing since research in fields with very little data was probably junk to begin with. I hope it leads to more data collection in the long run though and not just abandoning interesting research areas which are currently data-poor
Congratulations and welcome back! And thanks for the recs and links.
Did you happen to pause paid subscriptions during your leave?