I've noticed many people discussing the Lindy Effect in relation to lending protocol risks. There seems to be significant misunderstanding, with two distinct types of Lindy emerging:

→ The true Lindy Effect (which CT should discuss)
→ The experienced Lindy Effect (which CT actually discusses)

First, what’s the lindy effect?

The Lindy Effect proposes that non-perishable entities (ideas, technologies, books, cultural phenomena) have longer life expectancies as they age. Essentially, the longer something exists, the more likely it is to continue existing.

The Lindy Effect concept originated from a 1964 article by Albert Goldman, discussing comedians' longevity in relation to their career length. It was later refined by thinkers like Benoit Mandelbrot and 
Nassim Taleb.

The largest onchain marketplace. Buy and sell crypto on Ethereum and 10+ other chains.

wizard building public financial infra @morpho

The Flaw

When a protocol is upgraded, the true Lindy Effect should reset to zero. Theseus’ paradox does not apply here. Indeed, the Lindy Effect applies only to non-perishable entities, but once a contract is updated, the previous version has perished, and a new entity has been created. Yet, most users assume the protocol's core logic remains unchanged, leaving their perception of the experienced Lindy Effect unaffected.

Upgradeable protocols benefit from this cognitive bias, causing users to overestimate their safety. This misconception becomes even more pronounced upgrades swiftly fix critical issues thanks to effective bug bounty programs. While the incident may be quickly forgotten, the new code could have introduced unforeseen vulnerabilities.

In summary, only immutable protocols like Uniswap or Morpho are truly "lindyable" objects. Applying this principle blindly to regularly upgraded contracts can be dangerous. Upgradeable contracts accrue Lindy, but only from their last upgrade.

Now that everyone understands the term, how does it apply to DeFi and onchain protocols?

Applied to DeFi, the Lindy Effect suggests that the longer a protocol operates without major incidents, the less likely it is to be hacked. For example, @uniswap v2 has been running since 2020 without facing any significant issues. Based on this, we can reasonably expect it to continue operating safely for the next few years.

This principle works well for immutable protocols like Uniswap. However, it becomes problematic when applied to upgradeable protocols such as Lido, Aave or Compound. These protocols have undergone multiple upgrades over the past year, either for improvements or to fix major bugs discovered in their codebase.

adding to list of threads that made me smarter

Isn’t the point of building and focussing so much on infra before consumer apps to essentially get rid of all the non-lindyable infra naturally & then later on follow the same process w consumer dapps?

Every hack makes the entire defi space stronger because it’s all public.