jdilla.xyz

How will agents interact with the world?

2024-07-13

Lattice made a splash this weekwith a pretty crazy announcement about adding AI workers to their platform. It was shambolic and they’ve since walked it back.

While Lattice did this poorly, I think that the question of “how do we integrate Agents into the world?” is an interesting place to dig and experiment right now.

As an example, if you believe in agents, it seems pretty clear that agents are going to need to be able to pay for things subject to certain rules. So... what does it look like to give an agent a credit card?

I could just give it a credit card in my name, but that seems a little risky, and if things go wrong, who’s going to make that right?

But if I hire an agent created by another company to do work for my company, who gives the credit card to them? Is it the creating company? Do they then invoice me after the fact?

It's possible that this looks exactly like how businesses give workers credit cards... but maybe not? It might be better to know that this is the card assigned to system X by entity Y. The entity that is ultimately on the hook for the spending even if things go wrong might want to be able to track that; the credit card issuer might also want to know which of its clients are giving Agents these abilities as the patterns of spending, real and fraudulent, might look different. This transparency probably helps the system overall.

Another example is account creation. There are probably types of services where we want non-human actors to be able to create an account. We could have them pretend to be human, but it might help to let them ask for agent access to a service. This is probably different from API access; in some cases, it probably helps for them to see exactly what I see in the system.

Zooming out a bit, it seems to me that people get really upset when something pretends to be a human but it is actually AI. It also seems likely that we’re going to want to give agents more ability to act in the world and be productive. Yet the systems we have today that are essential for productive work assume human actors or computers acting on behalf of humans (programmatic access), but nothing in between. If we’re going to capture the value from agents, our systems are going to have to adapt.

How children refer to adults

2024-06-25

When I was growing up, I never used first names with adults. The adults in my life were "Mr. Knabe", "Mrs. Stanley", or "Dr. Woods".

Adults reinforced this norm as well. When I met my parents friends, they introduced themselves — in a friendly way — as "Mr. Brinker" rather than Chris. The same with teachers — I had "Mrs. Bryson", not "Deborah".

My parents would’ve corrected me had I tried something else. I’m sure they probably did at some point, but I don’t remember it happening. It wasn’t notable, it’s how the world was. In lots of cases, I'm not even sure I knew the first names of my parents friends until I graduated from college and then someone like Mr. Hehn would say, "please, call me Gunther" in a way that communicated I was now an adult too. This made me feel proud. The only exceptions I can think of here are my Pastors (Jerry) and family (Aunt Julie and Uncle Bert).

As far as I can tell, this has completely gone out of fashion.

With my kids, 2 and 4, no adult uses their last name. My friends introduce themselves as Mr. Jon and Ms. Veronica, not Mr. and Mrs. Flash. I do this too — I introduce my friends to them as Mr. Graham and Mr. Ted not Mr. Rowe and Mr. Strong. Even my daughter’s teacher is Ms. Heather not Ms. Jones. I assume this will change as they enter the formal school system… but who knows!

This new behavior is so consistent that if an adult that I knew well introduced themselves to my child as Mr. Banna instead of Mr. Rami, it would seem overly formal, like wearing a tuxedo to an office.

This doesn’t bother me on a moral level but I am intensely curious about it. When did it change? Why? I assume it’s related to the broader decline of formality in our culture, the way that the hoodie has replaced the sports coat for menswear.

But what is driving this? Is it a desire to be youthful? Relatable? A way of communicating that adults and children are on the same level? As we’ve made this switch, what have we given up? Anything? Nothing? Does this change how children perceive adults? Does it change how children perceive themselves?

I’d love to hear a theory of the case here.

Reactions to Situational Awareness

2024-06-14

My first glance reactions to Situational Awareness: The Decade Ahead by Leopold Aschenbrenner.

I enjoyed reading it a lot.

The most persuasive part of his argument to me is the relationship between compute and intelligence. This is sort of like the New England Patriots to me; I'm going to believe in it until it stops working. I see reasons why it might stop (run out of data, limited by energy / computing power available), but I don't know when or if we'll actually hit those constraints. People are pretty good at avoiding constraints!

I think he underrates the likelihood of a bottleneck somewhere that keeps us from getting to the AGI he imagines. Any individual bottleneck might be unlikely, but as long as one exists, the entire system is constrained.

Something I see Leopold do at points is assume a super AI, in his case, an automated AI researcher that is 100x as competent as today's top AI researcher. With this assumed, any AI research problem is solvable because you can scale up infinite 100x AI researchers to get around the problem. Once any AI research problem is solvable, then any problem is solvable.

What I think will ultimately happen is something like this:

  • An AI will exist that is super human on many dimensions. It will be able to do many things way better than humans and will be inarguably smarter than most humans. [0] Most of todays knowledge work will be offloaded to the AIs. This will be similar to the way that a lot of the production work of 1750 has been moved to machines in factories.
  • That AI will also have limitations. There will be some things that it can't do as well as humans or where humans will have the ability to reliably trip it up, despite it's intelligence. To extend the factory analogy, you'll still have humans pressing buttons for reasons other than just keeping the humans in control.
  • This will be really destabilizing. Society is going to change more between 2020 and 2040 than it did between 1950 and 2020.

Somewhat off topic: earlier this year, I read Meet You in Hell, which is the story of Henry Clay Frick and Andrew Carnegie. The dynamics of that era, with the railroad leading to a spike in demand for steel and steel leading to a spike in demand for coke were very recognizable in today's AI race.

[0]: I think GPT-4 is already this! Do you know a single person who knows as much stuff about as many things as it does? I don't. And yet it still has limitations!

Goodbye, Bill Walton

2024-05-28

240527-bill-walton-al-1323-d43839.png

Bill Walton is my favorite athlete I never saw play. He passed away yesterday.

In the summer of 2007, I read David Halberstam's classic book, Breaks of the Game, which is about the Portland Trailblazers of the Late 1970s who briefly shot to the top of the NBA world before falling back down to mediocrity.

Breaks of the Game is as close to a perfect book as I've ever read. The focus for the story is Bill Walton's foot, which breaks in the 1977-78 playoffs with his Portland Trailblazers set to win a second straight NBA title. This bad break cuts short a championship run, a budding dynasty, and according to Halberstam, the perfect basketball team with Walton the perfect basketball star.

With this as his lens, Halberstam tells the story of how basketball and American culture are changing, becoming more financially driven, and in the process something is being lost. If you have even a passing interest in basketball, you should read it.

I met Bill Walton through this book, but I fell in love with him when he came on to the Bill Simmons podcast. [1]

Simmons wanted to have a fairly conventional conversation about the 2007 NBA season. [0] But Walton wanted to use basketball as a metaphor for life, specifically how to create a meaningful life, especially for young people, through preparation, judgement, and irrepressible joy. He had a sense of the interplay between individual talent and team chemistry that I love most about the basketball and articulated in a way no one else can.

Over the years, I've probably re-listened to that conversation 15 times to try and capture a little bit of his approach to life. [2]

As I've gotten older, the list of athletes I want to meet has dwindled. I still appreciate the skill, but cringe at the awkwardness of approaching another grown man... to talk about what exactly? I'd rather leave them to practice their craft and enjoy their time with their families.

Bill Walton remained the exception. When I moved to San Diego in 2021, I would go on runs past his house in hopes that he might be out checking the mail and I might get a chance to thank him. For what exactly? Having a great basketball career? A wonderful approach to life? I'm not exactly sure, but I know he mattered to me. Of course, I flattered myself that he'd recognize in me a kindred spirit and a friendship would be born. It wasn't to be. With his passing, it never will be.

It's an odd thing to care this much about someone who you never saw play. If anything, it's a reminder that how we live matters and can echo out into the world to the people around us, even people who've never met us and never saw us do the thing that we were best at.

Thank you, Bill Walton, for sharing your gifts with me.

[0]: Such a great time to be an NBA fan.

[1]: At the end of the podcast, Simmons brings up Breaks of the Game and it's so painful for Walton that he has to end the show.

[2]: I also printed out a copy of John Wooden's Pyramid of Success and bought a handful of old Sports Illustrated with Walton on the cover.

ChatGPT Mac App

2024-05-24

I'm one of the early users here, but this app is severely underbaked, even for a beta rollout.

Forget about advanced features — I'm having issues with basic scrolling. It's also slow. The speed of response time that's so exciting on the web isn't there for some reason.

I also miss the ability to cmd + f for pieces of text within a chat. For some reason, on the app this does a search across my chats but not within the chat I'm focused on. Not helpful!

I don't want to draw too many conclusions from a bad week for OpenAI, but it definitely seems like they've lost focus on what matters.