Creating the theoretical best possible money (analyzing Elon Musk's thinking in Lex Fridman's podcast)
Creating the theoretical best possible money (analyzing Elon Musk's thinking in Lex Fridman's podcast)
Digging into the core of what a perfect form of money would look like and applying this to crypto.
This article is mostly paraphrasing Elon Musk's words from his podcast with Lex Fridman, and then applying it to crypto practically to arrive at the theoretical best form of money.
Cryptocurrency is an interesting approach to reducing the error in the database that is called money. Right now the money system, for practical purposes, is a bunch of heterogeneous mainframes running on COBOL, in batch mode.
This is no exaggeration. There is a fascinating article about it here.
The Bank of New York Mellon in 2012 found it had 112,500 individual COBOL programs, constituting almost 350 million lines.
The financial world essentially still runs on COBOL.
The government effectively has editing privileges on the money database. They use those privileges to create more money whenever they want. This increases the error in the database that is money.
I think money should really be viewed through the lens of information theory, kind of like an internet connection. What’s the bandwidth, total bit rate, what is the latency, jitter, packet drop, errors in the network communication.
Applying this to crypto specifically, the ideal crypto according to these criteria would have:
High volume throughput (bandwidth/bitrate)
Fast and predictable confirmation times (latency/jitter)
No dropped payments (packet drop/errors)
There is also error in money that arises from governments diluting or debasing the money supply as basically a pernicious form of taxation.
How does this form of error lead to a hidden form of taxation?
Let's say there is a total money supply of $1,000. You hold $10, or 1% of supply.
The central bank now adds $100 to the money supply. You still own $10, but the total has increased to $1,100. You own 0.9%, rather than 1%.
You've effectively lost 10% of your wealth, but would likely never be aware.
Extrapolate this on a longer term and you'll see how it adds up to your $10 gradually becoming worth far less over time. This is what is meant by dilution/debasement.
Think of money like an information system.
The question then becomes: what system, from an information theory standpoint, allows an economy to function best?
Cryptocurrency takes us to the 21st century in terms of the actual systems that allow you to do transactions, to store wealth. It’s an attempt to solve both technological error and debasement error.
Think of money in that way, as a database for resource allocation throughout both space and time. What would be the most effective and efficient version of that database?
There is a fundamental issue with for example Bitcoin in that in that its transaction volume is very limited. Its latency, the latency for a properly confirmed transaction is too long, much longer than you’d like.
It’s actually not great from a transaction volume standpoint, or from a latency standpoint. What you want to have is a high transaction volume capability. You want to have low costs of doing a transaction.
Putting numbers on it, Bitcoin's maximum throughput on its base layer maxes out at ~7 transactions per second.
These transactions take a while to settle - average confirmation time yesterday was 79 minutes with an average fee of $3.16 per transaction.
When Bitcoin started, in 2008, internet connections were much worse than they are today. Having a small block size and a long synchronization time made sense in 2008, but 2021 or fast forward 10 years, and it’s comically low.
What can we conclude from Elon's thoughts about ideal currency?
Our ideal form of money should have the following characteristics.
High and scalable throughput
(Consistently) quick transactions
No dropped payments
Low fees
We can also conclude that he is of the opinion that Bitcoin is not the answer for obvious reasons, and that "Doge is fundamentally better than anything else he's seen". So how does Doge do on these metrics?
~33 transactions per second, difficult to increase further
A few minutes average confirmation time (increases with increased usage)
No dropped payments
Average tx fee $0.27 yesterday, increases with increased usage
At a quick glance, Doge is indeed better for payments than Bitcoin is. However, as Elon says, it's helpful to boil things down to first principles. So what are perfect characteristics in terms of currency, reasoning from first principles?
Infinitely scalable throughput
Instant confirmation, at light-speed
Never a dropped payment
$0 fees
Do we have something that approaches this form of perfect currency? I would argue that we come quite close, with Nano.
Scalable throughput. Improve the bandwidth/hardware of the nodes running the Nano network and maximum throughput automatically increases.
Nano's median global confirmation time is 313 milliseconds, 99th percentile at 634 milliseconds.
Nano has done 140 million transactions without ever a dropped transaction, doublespend, or rollback.
$0 fees, forever, by design. The sum total fees of the 140 million transactions done so far is $0, the sum total for the next 140 billion will be $0.
Nano also has 0 dilution/debasement. It's not possible to create new Nano. It's 100% predictable and fixed.
An additional important factor that Elon did not outright mention is (energy) efficiency. Nano transactions are so efficient that if the network were processing ~7,000 transactions per second, it could still be powered by one single wind turbine.
Does Elon's statement of "Doge being fundamentally better than anything else he's seen" make sense?
Yes. He might not have heard of Nano yet. It's likely, given that the foundation developing Nano has so far focused on the protocol, rather than marketing.
Is Nano perfect? No. Software almost never is. Tesla engineers built their own C compiler for maximum efficiency. Apply this to Nano and I'm sure we could further optimize throughput and speed up transactions a bit more. There is a new version of spam resistance and while the theory is great, it's not fully implemented yet.
Fundamentally however, we're approaching perfection. It's hard to beat lightspeed confirmations. It's hard to beat scaling that uses all available resources. It's hard to beat $0 fees forever.
Think of money in that way, as a database for resource allocation throughout both space and time. What would be the most effective and efficient version of that database?
Allocating resources throughout space at lightspeed with $0 loss and 0 error, and throughout time thanks to 0 dilution over time leads to the ultimate form of money.
It means that our database for money allocation can function at maximum efficiency. Not constrained by borders, by transaction fees, by high and variable latency, by unpredictable issuance. Without error, without loss.
This is what makes crypto so exciting to me. We have the chance to create the ultimate form of money. We can take steps back, work up from first principles, and approach perfection.
To anyone else that finds this goal exciting, I would strongly recommend checking out Nano. If you're interested in the future of money, if you're interested in making the world run more efficiently, Nano is a likely candidate. If you're a software engineer, you can help develop this new future. If you're a researcher, you can help develop this new future. If you're a writer, if you're an investor, host a podcast, if you run a business, a charity, or a country, you can usher in this future.
If you're interested in doing so, go down the rabbit hole. Explore, poke holes, be critical. I'm adding some links below for those interested. Many thanks for reading if you got all this way. Would love to hear your thoughts.
For investors wondering why a new form of currency could be interesting for them
For merchants and charities looking to implement Nano transactions
For anyone wondering what Nano's role in the long-term future of crypto might be