Making a Bitcoin wallet is incredibly easy.

In fact, it is probably the easiest wallet to make in the world.

It would be a challenge to make a wallet for the traditional banking system in under a year, but making a Bitcoin wallet takes just 20 minutes.

Today I’m going to teach you how.

Before building a wallet, it’s important to know the basics of wallet theory and the bitcoin transaction model.

If this is too basic for you, feel free to skip the next two sections and get right to the code.

Wallet Theory

What is a wallet exactly?

At its core, a wallet (bitcoin or otherwise) provides three main functions:

1. The ability to view your balance

2. The ability to receive money

3. The ability to send money

Depositing is the same as receiving money. Withdrawing is the same as sending money. They are just special cases where you are the sender of the receive action and where you are the receiver of the send action.

Some wallets may complicate this a bit but they are extensions of the core.

In Bitcoin country, wallets really are as simple as this.

Building a wallet will require us to remember these three functions and design around them.

The Bitcoin Transaction Model

The next topic for us to understand is the bitcoin transaction model.

Understanding this will give us the foundation for implementing the three core features of our wallet: view balance, receive and send.

The way that balances are secured on bitcoin is very simple.

Don’t worry if you don’t get it right away, it can be a bit confusing, but rest assured it is simple and it will become second nature to you as time goes on.

In Bitcoin, money resides in virtual locations.

These locations are identified by numbers known as “addresses.”

And they are unlocked by other numbers that act as virtual keys.

These virtual numerical keys are known as “private keys” because you are must keep them private or otherwise anyone can take your money.

Every address has a corresponding private key. They are generated at the same time and have an important mathematical relationship, which we’ll discuss later.

You send money out of your address by using your private key while at the same time making sure not to reveal your key to anyone else.

You do this by performing an action that only your address’s corresponding key could possibly perform: you use the key to create a digital signature around the description of the transaction.

Then you broadcast the transaction to the bitcoin network.

The bitcoin network (A.K.A. the collection of computers running the bitcoin software) checks the transaction to make sure: (a) it is trying to spend funds that are available and (b) it has a signature that corresponds to the hypothetical key that is tied to the address where the funds reside.

How does the network know that the signature corresponds to the address?

This is perhaps the most complicated part to understand, but we will try to get through it as simply as possible.

When a Bitcoin address is created, it is created alongside its corresponding private key. The two numbers are linked mathematically.

One number, the bitcoin address, can be easily derived from the other number, but the reverse computation is not feasible.

This is called a one-way function, because you can only go from “A” to “B” but not from “B” to “A”.

A good analogy would be that it is easy to check that two types of ground meat are the same, and it is easy to grind beef or turkey into ground meat, but it is essentially impossible to turn ground meat into its unground counterpart.

Going from a private key to an address is grinding meat - doable. Generating a signature from a private key is grinding meat - doable. Checking a signature against an address is comparing two types of ground meat - doable. Going from an address to a private key is un-grinding meat - impossible.

So there we have it.

The way to know that a transaction was signed by the correct hypothetical private key - the one that corresponds to the address where the money resides - is to check the signature against the address.

You don’t need to actually know the private key to know that the signature is correct. You just need to verify the signature mathematically against the address.

Creating Keys and Addresses

Now we’re ready to build the foundation of our wallet.

This is the code for wallet generation, or the creation of the private keys that allow you to spend your funds and the addresses that allow you to check your balance and receive funds.

As we mentioned before, the address and the private key are mathematically linked.

The process for creating a Bitcoin address looks like this:

1. Create a large random number that is mathematically infeasible to guess - this is the private key.

2. Use elliptic curve cryptography to calculate a public equivalent of this number - this is the public key (a really large number can be easily raised to a certain exponential power but the resulting number cannot be easily raised to the inverse power).

3. Take the public key and calculate a cryptographic hash of the public key - this is the public key hash.

4. Build an “unencoded Bitcoin address” by using the Bitcoin scripting language to specify a location where funds can be spent under the following condition, expressed in natural language: “funds can be spent from this address when a valid transaction is revealed with a signature produced from a private key that corresponds to a public key that when hashed equals a particular public key hash”.

5. Take the unencoded bitcoin address and encode it into base58 to produce the final Bitcoin address.

Ok so we have our basic steps. Let’s put them into action.

We’re going to be using JavaScript to make this wallet so that it can run in the browser, it can run on the desktop (via desktop app wrappers like Electron), and it can run on mobile (via mobile app wrappers).

In my experience, the best Bitcoin library for JavaScript is BitcoinJS Lib.

First, we go into the command line and install the library (as well as a few of supporting libraries):

npm install bitcoinjs-lib
npm install ecpair bip32 tiny-secp256k1

Next, we import the necessary libraries into our codebase:

import ECPairFactory from 'ecpair'
import * as ecc from 'tiny-secp256k1'
import * as bitcoin from bitcoinjs-lib

After that, we create a key pair:

const ECPair = ECPairFactory(ecc)

const keyPair = ECPair.makeRandom()

const { address } = bitcoin.payments.p2pkh({ pubkey: keyPair.publicKey })

When it comes to *creating* a Bitcoin wallet, this is all the core code that we need!

We can now take this code and incorporate it into a page where users can generate their own Bitcoin wallets and view their respective private keys and bitcoin addresses.

Implementing “View Balance”

Now that we have implemented wallet creation, we can move on to implementing the first of our core wallet features: the“view balance” feature.

We do this simply by issuing an HTTP request to a Bitcoin blockchain API. This is a website that runs a special version of bitcoin node software that keeps track of the balances for all addresses. This in turn means that you can ask it for the balance of any address and it’ll give you the answer.

There are many options out there but the first I would check out is Blockcypher.

Go to the documentation for their balance endpoint and you’ll see that you can lookup addresses with the following path:

/addrs/$ADDRESS/balance

Let’s use the following address as our example address:

const address = "1DEP8i3QJCsomS4BSMY2RpU1upv62aGvhD"

Now we can insert the example address in the path and create a URL that we can use to get information from the API

const url = `https://api.blockcypher.com/v1/btc/main/addrs/${address}/balance`

Now that we have our URL, we can make a call to the API and log the results to the console as follows:

const addressData = await fetch(url).json()
console.log(addressData)

Once you do this, you should see data in your console that looks something like this:

{
  "address": "1DEP8i3QJCsomS4BSMY2RpU1upv62aGvhD",
  "total_received": 4654010,
  "total_sent": 0,
  "balance": 4654010,
  "unconfirmed_balance": 0,
  "final_balance": 4654010,
  "n_tx": 16,
  "unconfirmed_n_tx": 0,
  "final_n_tx": 16
}

Bitcoin uses 8 decimal places for its balances, so to get the balance in Bitcoins for this address, we simply take the “final_balance” figure which is “4654010” and we divide it by 10^8. That gives us:

0.0465401

The address holds approximately 0.0465 Bitcoins.

With a frontend framework like React, we can display the user’s balance in big characters.

<h1>{addressData.final_balance / 10**8}</h1>

Part II

So there you have it. You’ve learned about how Bitcoin wallets work under the hood, you’ve learned how to create a Bitcoin wallet’s address and private key, and you’ve learned how you can display the balance of the Bitcoin wallet to the user.

In part two of this tutorial, we’ll show you how to implement the “Receive” and “Send” functionality of a wallet.

Testing, testing, testing. The foundation of an effective pandemic response.

And yet even today we have so far to go before we can catch up to what our testing capacity should be.

Back in 2020, the necessary reagents (chemicals) and swabs were in short supply, there were bottlenecks on human labor, and there was only so much our processes and institutions could handle.

But can we change this?

Is testing just a domain for biologists?

Can we use a little math and computer science to get to #TestingForAll?

Let's dive in and see if we can.

Testing Today

The standard way to perform a test for SARS-CoV-2 (the specific species of coronavirus that causes the disease known as COVID-19) is to use a process called Reverse Transcription Polymerase Chain Reaction (RT-PCR):

1. Collect a sample from the patient using a respiratory swab

2. Break open the cells on the swab

3. Extract all nucleic acids from the sample (including both human DNA and viral RNA)

4. Convert all RNA into DNA using a "reverse transcription" process

5. Take small DNA fragments (primers) that only bind to the SARS2 genome (like puzzle pieces) and use them to exponentially copy any potentially present SARS2 sequences through a DNA-doubling process known as polymerase chain reaction

6. Measure the amount of DNA in the sample using a machine called a qPCR (quantitative polyermace chain reaction) machine

7. If the amount of DNA is high, then the primers attached to a matching SARS2 genome, and one can conclude with a high probability that the patient is infected

Pretty clever process right?

It actually is remarkable how we can manipulate tiny little machines (molecules) that we cannot see and use them to take measurements and produce other machines.

The problem here, though, is that this process is very resource intensive.

The number one bottleneck is that the qPCR machines can only handle a certain number of samples per run, which takes about an hour. The LightCycler 96, for example, is about $34,000 and can process 96 samples per run. So if we have one of these machines, we can only test 96 patients per hour.

Can we do better? Perhaps math can help?

Let's see.

Efficiency Through Pooling

The testing process we just described is clever, but it is still heavily constrained.

Perhaps it is constrained because we are operating under assumptions that are meant to be challenged.

The first assumption we'll challenge is that each sample has to be tested one at a time.

What if we tested multiple samples at once? Could we save on testing time?

It turns out we can.

This process is known as pooled testing.

The simple version of this is to take a certain number of people, say 10, and then combine their samples together and test them all at once.

If the test comes up positive, then one of the samples is positive and we can proceed with individual testing.

If the test comes up negative, then all of the samples are negative and we don't have to do any further work.

This method makes a lot of sense if the prevalence in the population is very low, because the "all negative" result would be seen fairly often and save a lot of testing capacity, while the "at least one positive, test all again" case would be seen fairly infrequently.

For example, let's say only 0.1% of the population is positive.

If we test the normal way, then we'd need to test 1609 people to get an 80% chance of someone testing positive.

(1 - 0.001)^x = (1 - 0.8)
x = 1609

If we test 10,000 people, then for the first round, we'd have 10,000 tests with the "individual tests" method and 1000 tests for the "pool 10 samples" method.

The "individual tests" method does not require a second round, but the "pool 10 samples" method does, so that we can handle the postive pools and hone in the individuals who are the real positives.

On average, with 0.1% of the population positive, the "pool 10 samples" method would yield 990 "all negative" results and 10 "at least one positive" results.

Thus the second round for the "pool 10 samples" method would require 10 additional tests.

This would mean a total of 1010 tests for the "pool 10 samples method.

Now let's imagine that 25% of the population is positive.

The "individual tests" method would still require 10,000 tests.

The "pool 10 samples" method, however, would yield vastly different results.

There would be an 94% chance, for example, of each batch of 10 having at least one positive caes.

Therefore, the average number of tests required for the "pool 10 samples" method in this case would be:

1000 + 0.94*1000*10 = 10,400

This is even higher than the "individual tests method," and would double the testing time required on average.

Thus, we wouldn't want to use the "pool 10 samples" method if we expect 25% of the population to test positive.

Can we do better?

Smarter Pooling

It turns out that there are even more advanced methods of pooling that we can employ to further increase our testing efficiency.

Next, we will turn to a form of pooling that allows us to avoid some of efficiency destroying traps that standard pooling falls into with higher infection rates.

We'll call this method "two-dimensional pooling."

Here, we'll imagine a chess board, and place each of the samples on the chess board.

The chess board has 8 columns and 8 rows, so we'll create 8 "column pools" by pulling from samples in the same column, and 8 "row pools" by pulling from samples in the same row.

That means the "Column A" pool will have samples from A1, A2, A3, A4, A5, A6, A7 and A8, while the "Row 1" pool will have samples from A1, B1, C1, D1, E2, F1, G1, and H1.

With this method, we should be able to narrow in on the true positive cases when our pools come up positive.

How?

Let's say B, E and F come up positive, while A, C, D, G and H come up negative.

And let's say 2, 4, 6, and 7 come up positive, while 1, 3, 5 and 8 come up negative.

We can visualize this as follows:

There are a few things going on here.

First, each column that comes up negative allows us to eliminate all of the cases in that column.

Second, each row that comes up negative allows us to eliminate all of the cases in that row.

Third, for each column that comes up positive and for each row that comes up positive, we know that *at least one* of the samples is positive. Each of the remaining squares was in one positive column pool and one positive row pool, but we don't know if that square was the sample that contributed in either case. Thus, we do not make any assumptions about these square and we put question marks over them.

Next, we perform a second round of testing on all of the squares that were in a positive row and positive column, and we determine precisely which samples are positive of the ones that remained.

From this we can conclude that B4, E2, F6, and F7 are positive, while the rest are negative.

That gives us a 6.25% positive rate.

Higher Order Pooling

Incredibly, pooling can get even more efficient still.

In the previous section, we explored how we can perform two-dimensional pooling.

Why can't we add even more dimensions?

It turns out we can.

Let's take that group of 64 samples and instead of using a two-dimensional grid of 8x8, we'll use a three-dimensional cube of 4x4x4 and can perform only 12 tests.

Taking this further, if we have 1024 samples, we can use an 10-dimensional hypercube of 2x2x2x2x2x2x2x2x2x2 and perform just 20 tests (or equivalently, we can use a 5 dimensional hypercube of 4x4x4x4x4, which also comes out to 20).

In this way, the number of tests we must perform is 2 times the log of the number of samples we need to test.

number_of_tests = 2 * log2(number_of_samples)

It's also important to note that we can perform all of the tests at once, saving an enormous amount of time.

Once we have the results, we can treat these bits as the data input to a data structure known in computer science as a bloom filter.

For example, let's say we use the 2x2x2x2 method for 16 samples.

We can number the samples 0 through 15 (#0 for the 1st and #15 for the last or 16th), and perform 2+2+2+2=8 tests on all the samples and label them A through H.

Test A will contain the first half of the samples (#0 through #7) while B will contain the second half of the samples (#8 through #15).

Tests C and D will contain another half of the samples, split up in a different way. C will contain the first quarter of the samples and the third quarter of the samples (#4 through #7), while D will contain the second quarter of the samples and the fourth quarter of the samples (#12 through #15). We will repeat this progressive splitting until we get to H.

Note that a quick way of doing this sample batching computationally is to convert each number to bits. For example, the first sample (#0) will be expressed as the 4-digit binary number 0000. The last sample (#15) will be expressed as the 4-digit binary number 1111.

Tests will be peformed on all samples that match a given bit:

Test A = all samples where the leftmost bit is 0

Test B = all samples where the leftmost bit is 1

Test C = all samples where the second highest bit is 0

Test D = all samples where the second highest bit is 1

Test E = all samples where the third highest bit is 0

Test F = all samples where the third highest bit is 1

Test G = all samples where the rightmost bit is 0

Test H = all samples where the rightmost bit is 1

Now imagine the tests come up as follows:

A = ➕

B = ➕

C = ➖

D = ➕

E = ➖

F = ➕

G = ➖

H = ➕

From here, we take each sample's binary representation and check it against the results.

The first sample is 0000, so it was included in tests A, C, E, G.

Since C is negative, 0000 is also negative, but more generally, all samples with 0 in the second digit (0000, 0001, 0010, 0011, 1000, 1001, 1010, and 1011).

Since E is negative, all remaining samples with 0 in the third digit are negative (0100, 0101, 1100, and 1101).

And since G is negative, all remaining samples with 0 in the fourth digit are negative (0110, 1110).

That leaves two samples left: 1111 (#15 AKA the 16th sample) and 0111 (#7 AKA the 8th sample).

These two samples are the likely positives.

It's important to note that this method does in fact increase the false rate, particularly as true positives exceed 1%.

Therefore, we may want to do another round of testing on the likely positives, which will only require us to do 2 more tests.

Millions at a Time with Barcoding

As if that's not good enough, we're going to take testing to an even higher level.

Imagine we could test a million samples all at the same time, in a single reaction.

Sound impossible?

There's actually a way to do testing that allows us to attach a barcode to each sample, then test the sample, and get a readout of all of the barcodes that are a match.

This may sound too good to be true, but it isn't.

This methodology has not yet proven to be as reliable as RT-PCR, which is the current industry standard, but it is a lot more promising in the long term because if its potential to massively parallelize testing.

Here's how the method works.

First, you take each DNA sample and attach a unique barcode to the end of it. The barcode is very specific sequence of DNA that can be distinguished from all other DNA, so when it is read, it is clear which sample is being read alongside it.

Second, you mix all of the samples together in one batch.

Third, you take the batched sample and you run it through a sequencing machine.

The sequencer will read all of the DNA segments in the batch and will output all of the sequences contained within.

Fourth, you analyze the DNA sequences and whenever you see a barcode, you mark that sequence as a particular patient sample. If the DNA attached to the barcode looks like the virus you are looking for, then that patient is positive.

Conclusion

Well there you have it.

We successfully demonstrated the massive parallelization of testing with mathematics and computer science.

Keep in mind that a lot of this is theoretical, and there are studies still being done to prove the accuracy of some of the more advanced methods, including sequencing-based testing.

And this would be a great time to shout out all of the incredible biologists and engineers working on next generation testing for coronavirus and other diseases.

Hopefully you got a glimpse of what can be accomplished when you combine powerful concepts from biology, mathematics, and computer science.

We are all cyborgs.

The phone you hold in your hand is an extension of your brain.

Just like computers have caches and memory, which allow them to quickly access information, they also have hard drives and remote databases, which allow them to access larger quantities information in a slower way.

With us, we have our internal brains, which allow us to quickly access information in our human memory, and then we have our external brains, which allow us to tap into vast expanses of information all throughout the internet.

We have become cybernetic beings and we hardly even notice it.

The Evolution of the Human Cyborg

Our phones may have made us modestly cybernetic beings, but the future holds even more incredible ways in which we will become cyborgs.

One of these ways is something I like to refer to as the cybernetic immune system.

Our bodies have a natural human immune system that takes care of us and combats invaders.

It has served us well for millions of years of our human evolution and for hundreds of millions of years of our mammalian evolution.

But our organic immune system is fallible. It can still fail us and disease can overtake us.

The good news is that we are a species that has the capacity to wield technology to greatly augment our abilities.

With technology, we can augment our natural immune system.

We can build an external auxiliary immune system that protects us when our natural immune system will not.

How Collectives Protect Individuals

Animals are built to promote their own continued existence.

The animal immune system acts throughout the body to ward off invaders and protect the organism.

The fight or flight response is behavioral code that ensures the organism does its best to escape danger or neutralize any given threat.

But these systems and responses aren't limited to an individual's self-preserving actions.

They can also act on a group level.

For example, if ants detect that one of their kin is infected with a parasite, they will carry the infected member away from the colony in order to protect the collective.

Humans can do this too.

When a crazed trucker was barrelling down the highway and trying to ram cars off the road, I received an alert from local law enforcement warning me about it and telling me to get off of the road.

This allowed me to stay out of harm's way without having to personally witness the threat.

And the law enforcement response was able to neutralize the threat on behalf of the collective society.

We can do the same with disease.

The Human Immune System

Here's the way the immune system works:

When the body is exposed to a pathogen, the immune system starts a nanoscopic R&D program to develop specialized machines (antibodies) that can fight it.

This process can work spectacularly.

But of course, it can take weeks to months to start to work and the system can be outsmarted.

Biotechnology as Immune System Augmentation

The good news is that we have incredible innovations that allow us to pool our resources as a species.

We can take the learnings derived from one person's infection and apply them to the rest of the population.

Our first line of biotechnological defense is the vaccine.

We can produce vaccines by killing viruses (inactivated vaccines), weakening viruses (live-attenuated vaccines), or "photocopying" viruses (subunit and nucleic acid vaccines) and then exposing individuals to these safer versions before the actual viruses reach them.

Our second line of biotechnological defense is the antibody.

We can take the antibodies produced by one person who successfully fought off the infection, analyze them, mass produce them and give them to others as a treatment, protecting the individuals before they even witness the threat.

The Cybernetic Immune System

The next level of biotechnological evolution will come when we are able to fully automate the process of vaccine and therapeutic discovery, manufacturing and delivery.

Our bodies have an automated R&D and manufacturing program.

Why couldn't we have one as a species?

In the future, we will have a single cybernetic implant that rests on our arm and protects us from nanoscopic invaders.

The cybernetic immune system.

This machine will constantly monitor our bloodstream for viruses.

It will sequence them and log them to a personal database.

If multiple individuals fall dangerously ill with never-before-seen symptoms, the public health authorities will be alerted and each person will receive a request to share their viral measurements.

Once measurements have been uploaded and processed, if a genuine new public health threat is detected, all other individuals with cybernetic implants will receive a notification about the threat.

They will receive information about the threat and will be able to make an informed opinion about whether they want to protect themselves from it.

The device's app will show a new vaccine that they can download and "install" into their bodies.

All that is required for this vaccine to be manufactured on the person's implant and delivered into the person's arm is the sequence of the virus.

Once the sequence has been downloaded and once the person has started the installation, the cybernetic implant will begin producing the vaccine and then will deliver it into the person's bloodstream, forever protecting them from the viral threat.

Science Fiction or Reality?

The crazy part about this vision is that it isn't as far off as one might think, at least from a scientific research perspective.

The first part of the vision, which simply detects viruses in the bloodstream and relays the sequences to the network, is the closest part and may only be a decade or two out.

The second part of the vision, which involves manufacturing a vaccine on the person's arm, will take more time - likely multiple decades.

The technology required for Vision I, real-time personal sequencing, involves sequencers that are (1) compact (2) reusable (3) inexpensive.

The current most compact and economical DNA sequencer on the market is the MinION by Oxford Nanopore Technologies, which is about the size of a wallet and could theoretically be attached to an arm as a strap. With the Flongle adapter, one can perform up to 2GB of sequencing for just $90 in a single reaction.

To close the gap, we'd need a sequencer that could run continuously for a much lower cost per run. But that is extremely achievable. Only advancements in engineering are required. No fundamental leap in research is necessary!

The technology required for Vision II, real-time personal vaccine development, involves (a) a demonstration that safe vaccines can be reliably developed from nucleic acid sequences alone (b) synthesizers that are compact, reusable and inexpensive (c) a device/module that can transform a sequence into an injectable vaccine on a person's arm.

Regarding real-time development of nucleic acid vaccines, Moderna already has a platform to produce "just-in-time" RNA-based vaccines. With simply the knowledge of a virus's RNA sequence, Moderna can crank out a new vaccine in a matter of weeks to even days. The greater challenge for vaccines is proving they are safe and they work.

That said, as we go through more and more clinical trials and perfect our vaccination process as a species, we will get to the point where we don't even need trials for each sequence. We'll be able to have suffient confidence in a given process for manufacturing and delivering the vaccine, and we'll be able to rest assured that any given vaccine that is produced with this process will be safe and effective.

Regarding synthesizers, we still have a much longer way to go before they are as compact and affordable as sequencers. Synthesis is still done off-site at a high cost using very large machines. The industry standard for sequencing relies upon a synthesis method called phosphoramidite synthesis, while new research is being done into a different process known as enzymatic synthesis. In the future, breakthroughs in chemistry and engineering should allow us to move beyond our constraints in machine size and cost.

All of this will not be easy and it will not happen overnight.

It will probably take decades.

But it will happen, and it will allow us to escape the evolutionary cat-and-mouse game of pathogen and host.

Cybernetic immunity.

Voting and protests—that’s the political energy that you see. But there’s also a form of political energy that you don’t see. This is the energy that builds up between elections and protests. It is always there. It is simply waiting to be expressed and released.

We can use an analogy from physics. In physics, energy is never created or destroyed. It is only converted from one form to another. The same rule applies to political energy. There is visible political energy, which we call “kinetic political energy” because it is expressed as political activity, and then there is invisible political energy, which we call “potential political energy”, because it is stored and has the potential to produce political activity.

Why is this concept important? If we believe that political energy can be destroyed then we may be less likely to compromise and more likely to overpower the political opinions of our rivals. If we believe, however, that doing so will only create a temporary win and will contribute to a buildup of energy that will later be turned against us, then we may have added incentive to come to the table and compromise to make more parties happy, even if we have majority control and don’t need to do so.

How do we know this is true? Each midterm election in the US tends to be won by the party that lost in the major/presidential election that occurred 2 years prior. And these elections are generally won not because of swayed opinions but rather increased turnout. That means the more you make the other side angry and feel like they didn’t get a win, the more likely they are to turnout in the next election and revert the result.

Let’s realize that political energy can be stored and released again but not created or destroyed. And let us act accordingly. Accepting this fact will help us become a more united and less polarizing society.

One of the questions I’ve been wrestling with is what is going on with trust in the world. It feels like every day there is another institution that disappoints us and causes us to lose our trust in it, from government to the media to even scientific institutions. Not wanting to allow my biases and perceptions to pollute my conclusions, I decided to collect some data.

Trust by Institution

To start, here’s an excellent chart from Nature News (repackaging the data from a survey by Pew Research Center) that shows the institutions and groups that U.S. adults trust the most and the least:

As you can see, scientists and the military come out out top, and trust in these groups has increased since 2016. In fact, I find it curious that trust in every major group increased between 2016 and 2019. While elected officials, the news media, and religious leaders have relatively low rates of trust among the public, trust in them is actually moderately rising (at least according to this data from PEW).

Here’s another chart from Scientific American that shows trust in various institutions over time:

Trust in science has mostly been steady and then risen in recent years, while trust in the military has gone up steadily and trust in medicine, education and the press have all declined.

Trust on Covid-19

A more recent analysis was performed by YouGov to explore who people around the world trust most when it comes to covid:

Healthcare professionals top the list in almost every country, with trust rates on Covid-19 generally in the range of 80-95%. Trust in government and the media is mixed around the world and varies from country to country (around 40% in the USA for each). Friends and family, meanwhile, are consistently trusted around the world when it comes to Covid-19 (between 60% and 90%).

Trust in the Media

As we explore the lack of trust in the media in the United States, we see it has been declining over time (according to this survey by Gallup):

We also see that this decline in trust in the media has been driven by Republicans:

Of course, people across the aisle tend to trust the media they consume even despite distrusting the media at large (according to research by Fortune):

Trust in Government

When it comes to trust in the government, this has also been declining steadily over time (according to this survey by Pew Research Center).

Unpacking this, we can see how Americans view the performance of the federal government across various functions:

People believe the government is competent at (1) keeping the country safe from terrorism (2) responding to natural disasters and (3) ensuring safe food and medicine, but they do not believe the government is doing a good job at (1) managing the immigration system (2) helping people out of poverty (3) effectively handling public health threats.

Trust in Science

Moving over to science, this category appears to be performing the best of them all.

For one, skepticism in science has declined for the first time in three years (according to this report by 3M).

Additionally, a strong majority agree that (1) scientists are critical for our future well-being (2) science brings hope for the future (3) science will make life better in the next 5 years:

These are all heartwarming results, and frankly, results that I wouldn’t have expected.

Trust in Religion

In terms of religion, research from Barna shows that athiesm has increased from 5-7% among the last four generations to 13% among Generation Z.

According to Pew, meanwhile, self-identified Christianity has decreased from 84% in the Silent Generation to just 49% of Millenials, while unaffiliation has grown from 10% in the Silent Generation to 40% of Millenials.

This data indicates somewhat of a decline in religiosity and trust in religion.

Trust in Celebrities and Executives

Investigating trust in other groups produced some interesting results.

This question from 3M Research echoes other research showing high trust in scientists and health professionals, but it also shows that only 10% of people have trust in celebrities and corporate executives:

Why are we placing so much value on celebrity endorsements, again?

Trust is Earned

So what can we do to increase trust?

I truly believe that trust is earned.

For example, trust is highest in countries with high GDP per capita and low inequality (as shown by analysis from Our World in Data). This implies that as people prosper together, they have greater faith in one another, while as people diverge massively in opportunity and outcome, they lose faith in one another and in a society that doesn’t serve them.

Additionally, analysis by Pew Research Center suggests that when people in scientific professions take responsibility for their mistakes, are transparent about their conflicts of interest, and provide fair and accurate information, they earn even more trust:

This lesson could be just as equally applied to the media, to elected officials, to intelligence agencies, and to regulatory bodies. The more that these groups admit responsibility for mistakes, are transparent with the public, and provide fair, accurate and unbiased information, the more that people will trust them.

Let’s learn from this data on institutional trust. And rather than attempting to increase trust superficially, let us see how we can reach greater levels of trust by earning it.

Our democracy is broken. But I don’t have to tell you that—you can just feel it.

We’re living in an era of unprecedented tribalism, polarization and political dysfunction. Politicians regularly choose party over country. Extreme views are on the rise. Misinformation reigns supreme over truth. And more than ever before, people justify using violence to advance political goals.

Too much conservatism is not the problem we face, and neither is too much liberalism. No, the problems we face are born out of the two-party political system, where voters are trapped in a matrix of limited options.

Independents are forced to choose between the lesser of two evils, and Democrats and Republicans alike must must pander to extreme views within their parties, rather than building a nationwide coalition on issues the vast majority of the country support.

Further, there are good indications that a “left vs right” country isn’t who we naturally are, but rather what we have become as a result of our political environment. See how a strong majority of Americans believe the existing parties do an inadequate job of representing the people and thus a third party is needed.

In fact, there is evidence that “left and right are entirely tribal designations and have no unifying philosophy or principle.”

In my writing to follow, I’m going to present a solution to our tribalism—a way to fix our political environment as a means to fixing our democracy. And I’m going to try to convince you:

1. that this is *the* solution

2. that this solution is one that works by convincing people that it works

3. that there are enough of us who want this change to happen that it *can* become a reality—all we need to do is beat the same drum.

So what is the solution?

In short, it is to change the way we vote from single-choice voting, where you choose one candidate, to ranked voting, where you rank the candidates.

Whether you are familiar with ranked voting or not, read on for:

1. Why this is not only beneficial to our political system, but it is the single most important change that we need to make

2. Precisely how this seemingly monumental change can actually be set into motion

The System

To understand what we need to do we need to describe our political system. We’ll call this The System.

Elections are held each year in all 50 US States (major elections every two years). In each of these elections, voters go to the voting both and choose from one of two options: Democrat or Republican. Yes, there are often more choices but they are essentially non-choices, given that everyone knows they will never win.

As a result, ideological D’s vote for D candidates, ideological R’s vote for R candidates, and even most of the independents, who outnumber D’s and R’s, end up voting for either D’s or R’s.

After the election is over, people tend to subconsciously identify more with the side they ended up voting for. They consume more media on that side and they get sucked into one ideological side or the other.

Each of us are told there is no middle ground, that we must choose a side, that voting independently is letting the worst party win, and that a non-binary ideology is simply futile. Media frames issues in “left vs right” language, and non-duopoly parties are routinely kicked off the ballot.

Thus, the two parties are strengthened even more and their strength protects their duopoly. They are in a “stable equilibrium,” or a state of the world that tends to return to its original position when disturbed, much like a marble in the bottom of a bowl.

The parties appear to be enemies but they are actually allies. They need each other. They feed off of each other. If one party gets too strong, the other experiences a surge of enthusiasm and funding to provide balance. All the while keeping other parties out.

At the same time, I want to stress neither Republican nor Democratic politicians are to blame. Rather, they are victims of the two-party system, just as we all are. Just as voters conform their views and behaviors to the two-party system, so must politicians.

The Root Cause

Before we get to a solution, we must understand why this duopoly occurs and persists.

The root cause for this duopoly is simple: the single-choice voting system.

Duverger’s Law states that a single-choice voting system marginalizes smaller parties, generally resulting in a two-party system, while voting systems where you can select multiple candidates tend to favor multiple parties.

This means that creating new parties is futile unless you first change the voting system, as under the existing system a third party will always be marginalized.

In fact, every single time a new major party emerged throughout the history of the US, the new party displaced one of the two existing major parties.

The Solution

Now that we understand the root cause, the fix should be straightforward.

We must switch to a voting system where voters indicate preferences for multiple candidates.

If we make this switch, non-duopoly parties will no longer be marginalized, and new parties will be able to emerge.

There are two major types of voting systems that fit this criteria:

1. Ranked voting systems

2. Approval voting systems

Both are excellent, but we’ll focus on ranked voting because it has had the most real world deployment to date.

See Nicky Case’s interactive voting system guide for a brilliant visualization of different voting systems and how they directly influence the party landscape.

The Strategy

OK, so you may be convinced that ranked voting is the solution to our political woes.

But you’re like “so what if this never gets implemented in the first place.”

And you’re totally right. Let’s talk about how we actually make this happen.

So far, over 20 cities have implemented ranked voting (including New York City and San Francisco) as well as two states (Maine and Alaska).

The key to expanding on this success is to repeat what has worked in the past:

First, in each of the cities and states that implemented ranked voting, a referendum was put to the voters to see if they would like to adopt the new voting system. In Santa Fe, 65% of voters approved ranked voting. In New York City, an overwhelming 74% approved.

Second, referendums for ranked voting won’t pass if voters aren’t convinced that ranked voting is an improvement over the status quo. We need to spread the word about how single-choice voting is at the core of our political mess and how ranked voting is the answer.

Third, ranked voting actually needs to be properly implemented. If the implementation of ranked voting is a disaster, for example with confusing forms or unreliable tallying, then it will be reasonable for voters to push for a return to the status quo.

Conclusion

So there we have it. In summary:

1. We’re trapped in a political system defined by “left vs right” that favors tribalism and political division.

2. Ranked voting is the opportunity for us to escape from our tribalist politics.

3. We can make ranked voting happen by spreading the word, pushing for referendums, and making sure ranked voting systems are properly implemented.

So what do you think? Do you support ranked voting? Have a different take on the problem or solution?

Let me know in the comments below or send me a tweet at @ryaneshea.

In these pages, I plan to write about how we can leverage technology and culture to build a better future for humankind.

To me, humankind’s potential is without limits.

Our progress and our prosperity as a species is due to our ability to solve problems for ourselves and one another.

This requires advancements on two fronts:

1. Science, engineering and technology

2. Governance, civil liberties, economics, and public policy

The spirit of this publication is one of optimism, possibilism, and progress.

If this philosophy captivates your imagination, then join us, read on and subscribe.

Welcome to the Infinite Frontier.

Subscribe now