Let me preface this by stating that it is incredibly important for Bitcoin today, to consider and plan around the problems of the far future. The long-term feasibility of Bitcoin has a direct impact on the people investing in it today, because if Bitcoin has flaws in 100 or 200 years, it undermines it's strength as an immutable, unchanging, sound money today. And due to this, super long term obstacles shouldn't be viewed a 'cross that bridge' scenario. I wrote about some of these in my BTC 2048 Op-ed, but figured that these are interesting enough to warrant their own post.
I split a handful of these potential obstacles into short term (<10 years), medium term (<50 years) and long term (100+ years). A few of these have already have extensive discussions surrounding them, and obviously all of this is just an opinion and meant to spark discussion.
Short Term
Scalability to the Masses
This is the one obstacle that we most often see discussed on this subreddit. Let's quickly review the math as to how we got here. Bitcoin's base layer enables approximately 600,000 transactions per day. That equates to just over 200 million transactions per year - if each block is used optimally. Even if we assume every individual is using the second layers for all of their transactions, people eventually need to settle on the base chain. This number is incompatible with a population of 8 billion people trying to load open/close a lightning channel. Even more so as humanity's population continues to grow.
L2s will be the life-blood of a scalable Bitcoin as a result. With 99.99%+ of all transactions occurring on them. There are many different types of L2s that may solve these issues, from a spectrum of centralization. The only certainty is that they will be necessary.
In order for them to succeed, the User-friendliness, accessibility, and verifiability/trustworthiness of these L2s will need to be scrutinized and improved. This is the one obstacle mentioned here that already has a lot of work underway, and it is great to see.
Medium Term
Fusion-Powered Mining
Fusion power, while discussed for decades, finally seems to be making some real progress across the globe. With China most recently operating a record net-positive for 1066 seconds. Eventually, it will be an economically feasible source of essentially unlimited energy for a infinitesimally smaller amount of capital. Currently, energy prices can range from $0.03 to $0.30 per kWh, which is the main cost driver for Bitcoin mining around the world. What does this look like if a kWh of usage energy can be obtained for near zero?
The economic framework of Bitcoin mining may shift from one based on energy acquisition, to hardware acquisition. The operators of these fusion reactors may become the gatekeepers to mining, as it is not economically feasible to mine without access to their unlimited energy.
Markets for obsolete miners may skyrocket, because even an Antminer S9 has use if the input is nothing - or they may be recycled to manufacture newer, more efficient miners if the bottleneck is strong enough on the hardware.
Quantum Computing Break
SHA-256 is incredibly secure, but if we are doing a thought experiment potentially decades out, we would be ignorant to assume it is infallible - especially with the rise of compute and quantum computing coming in the decades ahead. The chance that SHA-256 becomes brute forceable is non-zero. What does this world look like?
In the event that computers begin to come close to the compute necessary, it would be hard pressed to believe that all active Bitcoin would not move to more secure attack resistant addresses. Blackrock, MSTR, whales, and retail would have the ability to move their coins to a safer location far before an attack was imminent on them. This however, does not account for lost coins. Which could account for millions of coins that would remain on susceptible addresses.
Lets dismiss for a moment the thought that once SHA-256 becomes susceptible to attacks, all Bitcoin in these addresses would be instantly drained. It would be more akin to today's mining, except the mining would be for lost Bitcoin.
At first, the largest dormant addresses containing >100 BTC would be attacked, and it would take potentially years with massive amounts of computing power to obtain the key. It would find an economic equilibrium, just like mining, where the amount of energy and hardware spent on compute would make it just profitable enough to be worth brute forcing the private key of an address.
This would, in a sense, potentially provide a 'new' inflow of BTC for the markets, supplementing the lack of newly minted Bitcoin coming from miners.
Long Term
Multi-planetary Usage of Bitcoin
There is now a path forward with the rise of privatized space travel, to colonizing other planets. Let's skip ahead one or two centuries, to a future in which we now have a sizeable colony on Mars, and potentially beyond.
One of the great aspects of Bitcoin is that the 10 minute block-time allows blocks to propagate across the entire globe relatively instantly, with all nodes in sync. But on larger scales that interplanetary distances entertain, the speed of light becomes an important factor that can not be ignored. Even the closest planet, Mars, is an entire 12 light-minutes away from Earth on average. That's essentially an entire block behind. And forget about the distances to farther planets or even stars.
This presents an obstacle for the two major aspects of Bitcoin; transacting and mining.
Transacting Bitcoin offers a few problems in this scenario. Imagine a scenario where 10,000 satoshis are held in a given address. Now imagine two people hold the private keys to this address, one on earth, and one on Mars. It would be difficult to prevent a double-spend attack if each planet keeps it's own record of transactions. Would we require all Mars transactions to require multiple confirmations? Is a solution like this a band-aid patch to future space travel? We can not expect a far future civilization in the Proxima Centauri System, 4 Light Years Away, to require 50,000 confirmations on their transactions for example.
One potential solution to this is to have separate transaction layers for each subsequent planet. Imagine one 'batch' transaction from earth that balances the ledger each day. It either load's up Mars' BTC balance, or offloads some of it back to Earth - Like opening or closing a lightning channel. Once that BTC has landed in that layer, Mars is able to transact within their layer until they wish to settle back onto the base chain on Earth. A still present issue with this is the potential increased centralization that forms.
Mining seems to be essentially impossible for anyone located outside of Earth. How will your hashes be accepted if you are an entire block behind? You inherently are locked out of verifying transactions because the speed of light prevents you from receiving transactions fast enough. Unless mining becomes localized to planetary chains; each planet mining their own transactions. This creates issues of their own though as fragmenting hashrate in this manner enables the opportunity for 51% attacks if the hardware is transported.
