In Warren Buffett's office hangs a piece of paper worth less than the nail holding it up: a Western Union stock certificate. Dangling between an award signed by Obama and a diploma from the famed Dale Carnegie public-speaking class. Looking back, Buffett can laugh and say "That's when Charlie and I quit short selling." Charlie would sigh and explain it was his third and final time betting against stocks.

Western Union was a telegraph monopoly running out of reasons to exist. Competitors' long-distance rates were falling while Western Union's kept rising. The uniformed messenger, once an iconic mascot, was being phased out as machines took over. They needed a growth story, and they needed it fast. They settled on Telex, a direct-dial teleprinter service still years from commercial launch and already eating a significant share of the R&D budget. It had to work, or Western Union was in trouble.

Market cap of $118M, earnings of $19M, pension outlays of $10M, reserves of $2.4M. The retirement obligations were growing 15% annually, on track to become a $200-250M liability¹ with no funding plan behind it. Western Union ran it pay-as-you-go, so the earnings looked real and weren't. 17 a share for a stock worth pennies.

Munger saw this as a prime short candidate. The company was overstaffed, surrendering pricing power, and watching its technology become obsolete. He'd learned from a previous investment that setups like this didn't end well. This was his chance to sit on the other side of the table.

We tried outsourcing our support to off-the-shelf software. Their pitch: generate perfect customer support experiences with AI. All their systems did was read our in-house notes and guess.

A user would write in asking if their July payment was recorded. The rented product would send back a generic article about our billing practices. The same one it sent to everyone. It didn't know who the customer was. Nor their history with our company. Nor the specifics of their billing setup. The bots weren't dumb. They just didn't know anything. Context problem.

Enterprise companies insist on shelling out millions of dollars for these AI solutions. Generic products nervously avoiding real problems. Frustrating customers and somehow creating more low-value work for employees. Most companies have yet to see real returns from AI:

We orchestrated a live session this week. The framework was treating AI as a drop-in remote employee.

Making sure you launch a new “agent” effectively will improve it’s performance. From onboarding to consistent performance reviews:

Click here for the whole presentation

A quick ask:

Would you help me compile answers for a new post?

Just reply to this email responding to: What books, PDFs, or articles have had the most impact on your life (and why)?

Special shoutout to Justin, Joe, Grayson, Momin, and Daniel for providing such excellent answers.

In 1953, Charlie Munger's first investment went from boom to broke in one year. It was the leveraged buyout that nearly cost him everything.

He was a successful attorney at Musick, Peeler & Garrett, cashing an hourly rate. Yet he had a burning desire to be rich. And fast. After hundreds of timid commutes, he decided to address himself.

He knew Transformer Engineers and its staff intimately. They were a common job shop. Developing and manufacturing components, magnetic amplifiers, wave filters, and hermetic transformers for missile guidance and telemetry. Sold through the prime contractors in the aerospace missile defense industry. Yet this one was different than the dozens of others that had popped up in the Pasadena highways trolled by our enterprising lawyer.

In 1995, nobody doubted the internet would matter. But holding the exponent in your head was impossible. Nobody doubts AI is a big deal either. But who is accurately imagining how big?

Scale estimation remains elusive in a world where the crowd can decipher direction. The mistake is in the magnitude.

Take Jeremy's viscerally titled "Billion Dollar PDFs". Which he expands on in Patrick's podcast:

every once in a while someone basically crystallizes a notion right at the right time in the right way that becomes the foundational viewpoint or opinion in a certain era...the Billion-dollar PDF thing is this idea that you can form billions of dollars of capital one way or another around simply setting a new idea

Third South Capital

A lot of my time in the past month has went to our two verticals:

• M&A

• Services

May was a great month. For one, Yingyue and I said our vows:

Three weeks in, and I still smile at these photos. Some of the ones that stick:

• too excited to get started, we left the car doors wide open for hours

• the judge went off-script in terms of the vows, confusing everyone

• when we went to take pictures after, my dad took several of a duck

March was a pretty good month.

I got engaged

Thank you to Max and Rachel. Both were massive helps in the process of capturing this (and many other!) moments. They followed us around for over an hour. Hiding in bushes and monuments, to take a picture and video of the Proposal. A moment we will cherish for a lifetime! You guys are the best.

January was a busy month. My parents came to visit Carlisle, which finally gave me a chance to show them the apartment Abi and I have been putting together. Beyond that, I spent a lot of time on calls with customers and deep in the weeds across our various businesses at Third South Capital.

For example, I've been advancing some new features on Git-Zen. It's been a really fun change of pace because it's a completely different product and customer base from EntryThingy, where I’ve historically spent the bulk of my time over the past few months.

As I sit down to write this, two things are heavily on my mind.

1. Doing "Everything" to Win

Hi! It’s been a few months since you last received this newsletter. Here is what has been going on:

EntryThingy Turnaround

When we bought the company, the codebase was a liability. Customers stayed for one reason: we were the cheapest option. We had no roadmap, just a broken asset. A good asset in some regards, but still, a fundamentally broken asset to build upon.

We tore the code down and rebuilt it from scratch. For a specific cohort of art galleries, EntryThingy is no longer just the cheap option, we are the best option. That feels good to say.

You’ll love this: I found out greatmemos.com was available, so I bought it. It’s live right now!

You’ll notice the site is a repository of the best memos out there. You are the first to hear about it. Check it out and reply with any memos that should be added!

On the Third South front, we are doubling down on what works. With EntryThingy, we are simplifying how artists of all sizes find and submit work, which includes promoting our competition. With Third South Services, we are tackling a similar rebuild opportunity, but in the finance world.

The ET V2 launch wasn’t perfect. We found bugs. Some were legacy gremlins…some were our own doing. It’s embarrassing to have to fix things we convinced ourselves were fine. But our goal is to make the most useful platform for artists and galleries. We are almost there.

I hope you had a nice October! Mine constituted trips to Washington D.C. and New York City. There were some wonderful people in both cities, most notably family in D.C. and friends in New York.

I went on a spree and bought about fourteen over the past couple of weeks. The favorite thus far is Elon Musk by Walter Isaacson. Sam Walton’s Made in America is the current read. And I’m most looking forward to Napoleon: How He Did It, which was written by Napoleon’s long-time secretaries.

The finished and wrote-up books include insights gained from the Wright Brothers, Seven Days in the Art World, and The Only Rule Is It Has To Work.

Over the past month, I wrote about my gripes with email hygiene, an unlock in design engineering, and a long-held belief around the world being nothing more than a mirror.

Hi all,

September was mostly spent on rebuilding EntryThingy (more on this soon!) and settling into Carlisle. I didn’t write to you all in August because I traveled to France for a wedding, and to Williamsburg for family.

If you’re in New York, Richmond, DC, or Philly in the coming months, let me know. I'd love to see you!

Here are some pictures from a hike:

So July is wrapped up. I did a lot of writing, a lot of traveling, and had a lot of fun this month. Here are some of the published pieces from the past thirty days:

• The Earned Secrets of Constellation Software

  • for those who read this earlier, thank you for the kind notes
  • please reply with any reactions/questions to this piece, a follow up one is in the works!
  • Assorted Notes on the Union Tank Car Company - practical monopolies
  • Goodbye, Richmond, reflections on the River City
  • How I gave in to the dark side and started breaking habits.
  • Evolving my relationship with writing.

How We Published 3 Interviews in 4 Hours

We’ve unlocked an efficient new way to create content. In about four hours, I recorded, transcribed, and published three hour-long interviews:

• Third South Capital: Learn how our founders are building software companies without knowing any programming.

• Founder Spotlight: An interview with the founder of one of our recently acquired companies.

Our process is simple and AI-powered: Google Gemini handles the transcription, an AI agent refines it with pre-set instructions, and we get a polished Google Doc that requires only a few minutes of final review.

Investing

I joined Colin, Justin, and Harrison at Third South Capital as a full-time partner. We're buying and operating software companies. You can read our first public update here.

Most of my time in public markets has been reduced to special situations and acquisitive conglomerates due to the low research requirements. I’d rather spend the time on building companies with my friends!

Writing