jdilla.xyz

Are word processors creating all our problems?

2023-09-22

Speculative, but interesting. A form of the medium is the message.

Book thoughts: A Slave's Cause

2023-09-20

I did my best with this book, but I couldn't make it through.

A Slave's Cause by Manisha Sinha is a history of the abolition movement.

I picked it because I was interested in understanding how slavery came to be abolished.. I am still interested in this topic, but one of my goals for the year is to be more willing to put away books that aren't holding my interest, so that's what I'm doing for now. This says more about me as a reader than it does about the author and the book. I'm sure I was not her target reader!

The book is fantastically researched. It seems like the author found every single person in the historical record that opposed slavery and told their story. There are so many people who gave so much to the cause.

My biggest takeaways from the portion of the book I read are:

  • The sheer number of people across races that saw slavery as evil basically from the beginning. While viewing slavery as immoral was a minority position, it wasn't entirely uncommon.. People knew it was a bad idea.
  • The tremendous dilemma slavery posed to people who did believe it was wrong and wanted to do something about it. A modern day challenge with some similarity might be trying to avoid anything with greenhouse gas emissions (obviously not a parallel on a moral level). I'm grateful to be born into an era where I don't have to confront this.
  • The years of work, arguments, and missteps that went into abolition coming to be.

Perhaps I'll revisit this one in the future and if there is book written for more of a general audience that you know of, let me know as I'd happily start there.

The implications of AI as a skill leveler

2023-09-18

Ethan Mollick's fantastic One Useful Thing newsletter has an overview of a recent paper he did studying the impact of AI tools on BCG consultants.

One observation in particular stood out to me:

970b3354-1bac-4146-a92f-57f65440f872_1228x610.jpg

We also found something else interesting, an effect that is increasingly apparent in other studies of AI: it works as a skill leveler. The consultants who scored the worst when we assessed them at the start of the experiment had the biggest jump in their performance, 43%, when they got to use AI. The top consultants still got a boost, but less of one. Looking at these results, I do not think enough people are considering what it means when a technology raises all workers to the top tiers of performance. It may be like how it used to matter whether miners were good or bad at digging through rock… until the steam shovel was invented and now differences in digging ability do not matter anymore. AI is not quite at that level of change, but skill levelling is going to have a big impact.

This data is invaluable, but I think the framing of it (through no fault of the author's) obscures how individuals should be using LLMs. From the perspective of studying the impact of LLMs on a population of BCG consultants, there are low performing consultants and high performing consultants. But as individuals, we are a mix of low performers and high performers depending on the task.

Therefore the implication is that we should be much less afraid of our weaknesses, especially in areas that are complementary to our strengths. The quality (skill?) of being willing to learn by doing is going to be increasingly important, since the LLM will help cover the flaws. Then the way to maximize one's impact is to pick projects where you have a relative strength (beyond the jagged frontier of AI, in Ethan's framing) and pair it with tasks where the LLM can provide complementary, replacement level support.

I can't claim to have mastered this, but over the past 6 months, I've experienced this first hand across a number of domains: programming and in my day job, about material science, chemistry, and technical writing about cosmetics.

My friend Kenneth quipped recently: "What will you do with infinite junior software engineers?", but it's even broader than that. You have infinite access to baseline expertise in basically anything. What will you do with it?

ChinaTalk on Congress

2023-09-14

Fantastic ChinaTalk episode on the role of Congress in American civic live with Philip Wallach:

The specific things that stood out to me:

  • The role of Congress in creating legitimate legislation: The Civil Rights act of 1968 was given legitimacy by the process it went through. The same Southern Senators who filibustered it and fought it all along the way were then able to go back to their constituents honestly and say "we lost, it's time to move on," which gave the legislation staying power. Similarly, debates over WWII policies gave those policies legitimacy with a public who felt burned by WWI.
  • The facade of representative government: Wallach makes the point that virtually every country in the world has a body that is nominally representative, but in authoritarian countries, they have little to no real power. This is one of the things I find most interesting about the Roman Empire. The Republic falls, but until the end there continue to be Senators running around.

ChinaTalk continues to be the podcast most likely to introduce me to something I didn't know I was interested in

Alaska kelp harvest falls by 30%

2023-09-14

From Macro Oceans:

We spoke to half a dozen farmers across the state (if we missed you, please reach out!). All the data was self-reported and although we received updates from the same number of farmers as last year, we know of a few farmers who did not respond, so this total is probably a bit lower than reality. The results: total production in 2023 was 389,900 wet lbs, down 30% on last year.

It's clear to me that the bottleneck for the seaweed industry is customer demand.