All the points made by Justin Drake on Bitcoin and PoW, in order of the video. Most are quotes, some bullets have context added or sentences rearranged to make sense. No opinions are added. I simply wanted to collect all the information that was said about Bitcoin and proof of work together. I knew when first watching this episode that there was so much misinformation being spewed, it's really apparent now to see it all when written out. Not saying everything below is misinformation, since this is just a collection of everything Justin said, but it's pretty clear what his goal was.
- PoW is equity-like in that the miners have constrained voting power in the context of consensus but not in governance
- PoW does not have a social layer to intervene when there is consensus layer censorship
- Bitcoin has both commodity and equity properties
- Miners have outside power
- PoW is proof of capital
- Miners can be regulated in the same way a custodian can be regulated
- A 51% attacker gets to collect all rewards (issuance and fees) because they can mine a chain where they are they only miners winning every block and no one else wins a block
- Honest miners would turn off their miners because they would be spending electricity and not getting income
- Honest miners would liquidate their assets and sell their hardware
- The attacker can buy those miners to add to their hashpower
- The counterattack would take months if not years to organize
- Bitcoin has checkpoints all over the place
- In the Bitcoin Core software there are 12 checkpoints
- If there was an attacker that would build from genesis, the chain would not be valid because the developers have put subjective checkpoints in the code base
- If you've been offline, you need to download the Bitcoin core software and when you're downloading the software you're trusting the code
- Bootnodes: the first time you sync you need to connect to peers so in the code are hard-coded IP addresses
- If all the IP addresses are compromised, you'll be disconnected from the real p2p network and syncing the wrong chain
- Bitcoiners thinking that always syncing from genesis is fantasy
- If there was a reorg of 100 days, the last 100 days was just completely wiped out, do you think the Bitcoin community would accept that? that's totally normal? we're just going to accept the longest chain, even though there were 100 days that were just wiped out?
- Social governance would create a new subjective checkpoint and pin the new chain despite not being the longest
- The last person to come into miner has the advantage
- In PoW, the big fish have unfair advantages
- If you're a retail miner buying one Bitmain mining rig, you're going to get absolutely screwed on the price, you're going to be overpaying by 2x-3x-4x-5x relative to the professional miners buying in bulk
- Because of Moore's Law, you have a last-mover advantage
- The attacker can pick the latest and most bleeding-edge hardware
- For the attacker, the price of electricity is to a large extent irrelevant
- In order to perform a 51% attack it suffices to run it for 1 day or 1 week and then discourage all the honest miners to just turn everything off
- 99% of the cost of an attack is obtaining the mining rigs and infrastructure to run the rigs it's not in the actual operation of the miners and the electricity costs associated
- As a miner, you can go and acquire hardware which is literally in the dump
- You have professional miners who need the price of electricity to be let's say under 10 cents and if the price is over 10 cents they would get rid of that hardware because it's no longer valuable to them but it is valuable for an attacker
- The fact that PoW is inherently unfair, is inherently unequal, it actually creates advantages for an attacker
- Slashing attacker's stake is not available in the context of PoW
- In order to be a profitable miner, you need to operate at scale
- You can't just be a retail miner with just one rig, that doesn't work
- There are centralization factors because there are miners building their own custom firmware, they have their own specialized know-how
- They know how to take the mining rigs that they receive and just assemble them as quicly as possible so that they can be immediately connected
- There's also some centralization factors around the supply chain, what if the fab, TSMC, or global foundaries, was to create their own rig? We've seen that with Intel building their own chips. What if you just cut the middle man? That is a huge centralization factor.
- All Bitcoin mining is industrial because he doesn't know anyone who mines
- In Bitcoin mining, you have 2 classes of people, retail who are going to get absolutely rekt in terms of returns, and they might not even realize that they're getting rekt, and the reason is the price of bitcoin went so high that in dollar terms they're profitable but in bitcoin terms they put in 10 bitcoin as an initial investment and they've got 3 bitcoins back.
- Where if you're a super sophisticated industrial miner, you can sell hashrate futures, you can buy energy futures, you can be really well connected to the fabs, you might even make your own chip.
- Making a chip is at least $10m, if we're talking 3nm chip it might be minimum investment of $30m
- The "do nothing" approach to a 51% wouldn't work out because the price of bitcoin will crash and it will be difficult for honest miners to be profitable
- Miners would amplify their risk by doubling the amount of hardware and risk they have to fight a 51% attack with no income
- Nuclear option isn't satisfactory because you're effectively starting from scratch because the old SHA256 is completely compromised
- When starting with a new algorithm you could have an attacker buy out in terms of capital $5b, $10b, of the new algorithm and completely overwhelm the new form of PoW
- PoW is going to be completely disrupted by quantum
- BTC are non-voting equity tokens
- Governance by rough consensus off-chain through the community. This is how Taproot went through, this is how Segwit went through. Lots of debate and discussion, very chaotic and organic and ultimately the way that these governance decisions are enacted is through users installing software and agreeing that this is the canonical piece of software.
- PoW and PoS from a meme perspective are almost the exact opposite
- PoW issuance and transaction fees which is basically the income that miners have needs to be sold to cover for expenses
- PoW is a liquidity engine where you're adding liquidity to the market by being a forcing function to sell the issuance and transaction fees
- Professional miners will hedge risk because they're interested in a BTC-denominated return which will roughly match their opportunity cost
- As a professional miner, how do you lock in 5%? You use financial products. You use uncertainties, the price of electricity is uncertain. You can just buy electricity futures. The hashrate is uncertain. You can sell hashrate futures. And tada, you've locked in your 5% profit on your BTC, now it's just a matter of executing it over 4 years.
- The large professional miners are very much risk adverse and they're not going to take crazy risk with hundreds of millions of dollars. What they want is risk-minimized return on your bitcoin which roughly matches the opportunity cost.
- PoW is just PoS with extra steps
- The risk that's being introduced with PoW can be removed with financial products
- In PoW, the maintenance costs of $100 of economic security is roughly $100 every single year so the cost is roughly 20x higher than PoS
- Bitcoin has about $10b of economic security, using numbers from the grid purchase from Intel, $50 per THs and there's $200m per THs so you get roughly economic security of $10b
- Bitcoin community doesn't have a backstop power