Whenever I talk about a knowledge win via robots on the socials or with humans, someone snarks, “Well, how do you know it’s true? How do you know the robot isn’t hallucinating?” Before I explain my process, I want to point out that I don’t believe humans are snarking because they want to know the actual answer; I think they are scared. They are worried about AI taking over the world or folks losing their job, and while these are valid worries, it’s not the robot’s responsibility to tell the truth; it’s your job to understand what is and isn’t true.

You’re being changed by the things you see and read for your entire life, and hopefully, you’ve developed a filter through which this information passes. Sometimes, it passes through without incident, but other times, it’s stopped, and you wonder, “Is this true?”

Knowing when to question truth is fundamental to being a human. Unfortunately, we’ve spent the last forty years building networks of information that have made it pretty easy to generate and broadcast lies at scale. When you combine the internet with the fact that many humans just want their hopes and fears amplified, you can understand why the real problem isn’t robots doing it better; it’s the humans getting worse.

I’m working on an extended side quest and in the past few hours of pairing with ChatGPT, I’ve found myself constantly second guessing a large portion of the decisions and code that the AI produced.

This article pairs well with this one I read today about a possible social exploit that relies on frequently hallucinated package names.

Simon Willison writes:

Bar Lanyado noticed that LLMs frequently hallucinate the names of packages that don’t exist in their answers to coding questions, which can be exploited as a supply chain attack.

He gathered 2,500 questions across Python, Node.js, Go, .NET and Ruby and ran them through a number of different LLMs, taking notes of any hallucinated packages and if any of those hallucinations were repeated.

One repeat example was “pip install huggingface-cli” (the correct package is “huggingface[cli]”). Bar then published a harmless package under that name in January, and observebd 30,000 downloads of that package in the three months that followed.

I’ll be honest: during my side quest here, I’ve 100% blindly run npm install on packages without double checking official documentation.

These large language models truly are mirrors to our minds, showing all sides of our personalities from our most fit to our most lazy.

I’m a sucker for this style of post. This one in particular is jam packed with so many great pieces of advice that I had to read it three times before sharing it.

Here’s the very first item on her list. If it speaks to you, take ten minutes and thoughtfully consider the other 100 items.

1. You are overly obedient. You not only do what people tell you to do, but find it hard to imagine any world other than the one they present to you. Spend more time thinking about what you want, in isolation from the pressures of the world. (Keep this in mind while you read the rest of this very prescriptive document.)

At some point you have to accept that other people’s perceptions of you are as valid as (and probably a lot more objective than) your own.

This may mean letting go of a false or outdated self-image, including some cherished illusions of unique unlovability.

I recently had a talk with Shannon that was eerily similar to the central conceit of this article.

We don’t get to pick how we show up in other people’s interpretation of ourselves. The author’s story about his dad sleeping at the movie theater next to him is a great example.

Whenever I mentioned to people that I was working on a story about friendship in midlife, questions about envy invariably followed. It’s an irresistible subject, this thing that Socrates called “the ulcer of the soul.” Paul Bloom, a psychology professor at the University of Toronto, told me that many years ago, he taught a seminar at Yale about the seven deadly sins. “Envy,” he said dryly, “was the one sin students never boasted about.”

He’s right. With the exception of envy, all of the deadly sins can be pleasurable in some way. Rage can be righteous; lust can be thrilling; greed gets you all the good toys. But nothing feels good about envy, nor is there any clear way to slake it. You can work out anger with boxing gloves, sate your gluttony by feasting on a cake, boast your way through cocktail hour, or sleep your way through lunch. But envy—what are you to do with that?

Die of it, as the expression goes. No one ever says they’re dying of pride or sloth.

This is one of those articles that is hard to pull one single quote from, because it’s just so damn good.

The whole piece hits me right in the chest, and I’m sure you, dear reader, have someone you should be reaching out to after reading this too.

In the best case scenario, we create routines to hypnotize ourselves into repetition. We have loved ones and mentors who tell us to keep going, and help us figure out when we’re on the wrong track. We look for signs that we’re getting better, but we also understand that the process of getting really, really good at something sometimes just feels like a incoherent slog. If we’re lucky and resourceful and creative, we’ll eventually break through the membrane and find ourselves on the other side we’ve been clawing towards for so long.

I read Good to Great a few years ago, but I admittedly never finished it. After hearing this interview though, you'd better believe I'm gonna go back and pour over it.

This interview with Jim Collins was absolutely awe-inspiring. Among the nuggets I took away from this episode:

• You should strive to be a "Level 5 Leader", which means you are simultaneously headstrong and humble. You have to put your organization before any personal gain.

• Jim organizes his time according to the 50/30/20 rule, which means he spends 50% of his time in a given 365 day period on creative activities, 30% of his time teaching, and 20% of his time on everything else.

• On that same vein, Jim has a spreadsheet where he tracks how many hours a day he gets creative pursuits, and in any given 365 day period, he has to have over 1000 total hours. He also tracks what he did on a given day, as well as a rating from +2 to -2 for how he felt on that day. I've been trying to do something similar with tracking the big three things I need to get done each day, and I think I should expand that out a little bit to include these variables.

• You should not do what you’re good at, but do what you’re coded for. This really struck a chord with me, because I think I'm pretty good at developing, but I'm pretty sure I'm coded to be a leader.

• There was a lot mentioned around the flywheel principle, and I think this is something we're just starting to see happen with our own business pursuits.

There's a ton in this episode, so I'm going to stop writing in order to let you start listening.