The Case for the Old Guard
I have been best man at two weddings. First for my brother and again for my best friend.
My brother’s wedding was late 2020, so things were still in flux from COVID. He and his now-wife lived here in New Jersey, but she grew up in North Carolina, so the wedding would be in Wilmington. Due to the pandemic, by that time gatherings were allowed, but they were capped around 50 people.
Point is, things were hectic. It was the first time most of us were going to be traveling since the start of the pandemic, it was being planned from several states away, and it meant there was no single home base for anyone.
For the groomsmen’s attire, the default choice would have been to go to a local tux shop that’s been around for decades. But given all the complexity, it was decided that we’d try an online service instead. It seemed straightforward; provide sizing, select a few options, and your tux arrives ahead of schedule to try on and make potential alterations ahead of the event. In the end it was not. Two received the wrong shoes, one was missing a belt, and the sizes were acceptable at best. Shipping wouldn’t make it back in time and we had to scramble to find close comps locally, all while not letting this stress leak back to the bride-to-be. Figures.
Fast forward to a few years later, for my best friend’s wedding we all went to the local tux shop. We were measured and fitted and asked our questions to a man who’d been answering the same ones for 50 years. That tux fit me better than any piece of clothing ever had. The materials felt so good I wanted to take the shirt home with me and wear it to bed. One groomsman lost his socks, so he swung by and picked up a new pair the morning of the ceremony.
Some new online tux service could read this story, find all the issues we encountered, and start a competing one focused on fixing all of those things. And maybe it would be more expensive or slower or maybe it would be cheaper and faster.
Experimentation is essential, but trust and reliability claims are earned by doing something many many times in many many hectic situations.
In the blockchain space and especially regarding DeFi protocols, there’s common reference to the Lindy Effect, or my alliterative preference, Lindy’s Law. You might even hear people and things described as Lindy on Gen Z streams.
I learned today that the term got its origin from Lindy’s Delicatessen, a famous restaurant in the NYC Broadway theater district. It was coined by a reporter who identified an interesting phenomenon while observing comedians and performers at the deli: the longer a Broadway show had already been running, the longer it could be expected to continue running.
That story was later referenced by mathematician Benoit Mandelbrot, who extended this concept to power-law distributions, and later again by Nassim Nicholas Taleb into its modern interpretation.
In DeFi, it’s used to explain how a protocol that has existed for many years and has had significant value moved through it has stronger empirical evidence of security than a newer, even heavily audited protocol. It has presented an economic incentive for attackers for years, yet it remains standing today.
In short: survival itself provides evidence of durability.
Over the past few months we’ve seen dozens of critical vulnerabilities across technologies old and new. Given the pace of new AI model development, they will not slow down. Things may get a lot worse before they get better.
But eventually, the continued survival of what remains running provides evidence of its durability. What has survived repeated attempts to break it has demonstrated something that no amount of theoretical assurance or “Claude, find all bugs. Make no mistakes.” could ever replace.
That’s Lindy’s Law. And that’s a big case for The Old Guard.
But there’s another case for The Old Guard too.
In 2023, Silicon Valley Bank collapsed. The reasons why are a story for another day. But as the founder of a company in the tech space, this meant many of the companies we relied on for critical business infrastructure were out of their primary source of operating funds, at least until a backstop was found.
We were contacted by our PEO, a kind of payroll / hiring service which was responsible for paying our team, and reassured that if we were temporarily out of operating funds while the SVB situation was being figured out, that they had a distributed set of operating funds and were in a position to cover all payrolls for as long as necessary so our team would still get paid.
Their competitor was sure to jump on X and explain how this was their legal obligation as our PEO anyway. I think that misses the point.
It’s not a given that the company offering these services has sufficiently distributed and extensive funds to pay out my team even in the absence of my payment in. But they had built a business that accrued the resources and implemented the redundancy necessary to absorb a shock that neither of us anticipated.
Success and scale offer other benefits than durability: the accrual of leeway.
This is financial and non-financial. The goodwill they accrued made me a lot less annoyed when our rates went up a few percent the following year.
AI will uncover endless vulnerabilities in the process of unlocking endless new productivity, technologies, and services.
It will be tempting to scrap what is old, to rebuild anew, to move to a modern service, to drop The Old Guard. This may be the correct risk-adjusted move in many scenarios.
But remember that you’re trading a demonstrated history of survival for something that has yet to establish its own. You’re back to finding out if it will beat the odds.
And when the worst case happens, will it have its own backstop, or will it be waiting for another brave soul to raise its hand?