Note: My methodology for evaluating projects may differ from that of most KOLs; I do not discuss the feasibility of projects from a technical perspective. Instead, I analyze whether a project is worth investing in based on investment logic. After all, we are not technicians; we are investors looking for returns.
"Fully Homomorphic Encryption" (FHE) has been recognized for a long time, but I haven't delved into it. As a thesis-oriented investor, I think I can share my perspective and thought process upon encountering this new concept, hoping it can be helpful to everyone.
I am not a cryptography expert, so the first time I heard about Fully Homomorphic Encryption (FHE) was from an article discussing FHE technology details by Portail Ventures around mid-last year. Even at that time, I didn't pay much attention to FHE, but last month, Vitalik reposted his related article from 2020 on Twitter.
At this point, we can divide it into two phases. If you have a deep understanding of cryptography, you can try to grasp the entire technical logic of FHE. However, if you cannot understand those functions, then the key points to focus on in Vitalik's article are:
1. Vitalik's definition of FHE is as follows:
"Fully Homomorphic Encryption has long been regarded as one of the holy grails of cryptography. The prospects for Fully Homomorphic Encryption (FHE) are very bright: it is a type of encryption that allows third parties to compute on encrypted data and obtain encrypted results, which they can return to anyone with the decryption key for the original data, while the third party itself cannot decrypt the data or results."
In layman's terms: FHE is a type of encryption that allows third parties to participate in computations and obtain results, but the third parties cannot decrypt.
2. The article uses the email example:
A simple example: suppose you have a set of emails and want to use a third-party spam filter to check if they are spam. The spam filter wants its algorithms to be private: either the spam filter provider wishes to keep its source code closed, or it relies on a very large database that it does not want to disclose, as that would make attacks easier, or both. However, you care about the privacy of your data and do not want to upload unencrypted emails to a third party.
The original text has a diagram, but the English version is not easy for everyone to understand, so I have translated it into Chinese for better comprehension.
At this point, the essence of Vitalik's article has been distilled, which is of core significance for investment. The subsequent technical principles and a plethora of mathematical models are what technical personnel need to understand when designing products. If you understand, that's great; if not, it doesn't matter. Many may think this is very important, but it is actually quite trivial.
It should be understood that with new concepts, we need to comprehend what kind of technology they can bring, what products this technology can generate, and what market those products will have.
Thus, we enter into the commercialization research of FHE, which includes two inferences: 'application scenarios' and 'market scale.' If you want to take a shortcut, you can directly Google search and then verify whether the proposed application scenarios have potential.
Here is a lazy way to list things. Normally, you would define what problems FHE can solve based on an understanding of the future of the entire industry. However, writing that out would be too lengthy. Let's start with a straightforward outline, and we can elaborate on more detailed aspects in the community later.
For example, after searching, one can see the envisioned scenarios for FHE by the well-known venture capital firm Portail Ventures: 1. Private transactions; 2. Private smart contracts; 3. Private information retrieval; 4. MEV extraction; 5. Confidential bidding and auctions; 6. AI applications; 7. Data economy; 8. DAOs and governance; 9. Post-quantum applications.
There are many factors to consider, but the perspective we should adopt is: what does the industry lack, and can it solve that problem? Then we can discuss other aspects.
Many of the envisioned scenarios here have yet to find a viable route, such as private transactions in full-chain gaming and AI applications. Therefore, diving headfirst into these areas may not be very useful.
Therefore, in my view, FHE, as a foundational infrastructure technology for the industry, tends to focus more on enhancing software services and third-party supplements.
For example, providing modular supplements to zero-knowledge proofs to improve privacy; or assisting third-party services to expand their business scope. The resulting products from this are three main categories: plugins/hardware types, application chains, and technical facilities.
There is a significant problem here: for this type of project, it is impossible to estimate profitability because there are currently no applications that can provide commercial activities. So, it's very likely that they will follow the old path of public chains: VCs clustering together to push valuations and dumping on the secondary market.
It is worth noting that Fully Homomorphic Encryption is not new; it has several applications in traditional fields, thus there is a greater chance of creating usable products. However, it is not clear at this point, so projects need to be studied specifically.
If you still want to take a shortcut, you can directly Google 'What are the ecological projects of FHE?' I just started searching, and sure enough, the main projects are the types mentioned above.
Hardware & Plugin Types
For example, Ingonyama specifically designs processors for advanced cryptographic applications, providing underlying technical support for FHE and ZKP. They raised $20 million in November 2023 and another $20 million this January.
Another chip company, Chain Reaction, specializes in developing blockchain chips and has stated that it will release an FHE chip this year. Last year, it raised $70 million, with total funding reaching $115 million.
Another company providing CaaS (Software as a Service) is Cysic, which raised $7 million in seed funding led by Polychain. Additionally, there are innovative companies like Optalysys that research optical technologies.
Infrastructure Types
The leading project for FHE infrastructure is Zama, which specializes in providing FHE technology for blockchain and AI. It completed a $73 million Series A funding round in early March this year, focusing on B2B applications.
The remaining similar products are quite few, with most raising around $3 million, such as PADO, Sybscreen, SherLOCKED, and Fair Math.
For retail investors, the above two categories have almost no participation, but what can be seen is that leading companies in FHE infrastructure and hardware have significant funding, indicating that these products must cover the entire industry and possess a very large market scale.
For retail investors, what matters is whether there are projects issuing tokens and whether there are projects about to issue tokens, as this determines whether we can buy directly or participate in airdrops. For instance, public chains? Indeed, there are.
Public chain projects:
First, the layer2 series:
Fhenix completed a $7 million seed round in September 2023 and is expected to launch its mainnet in 2025.
Shibarium raised $12 million in funding.
Next is the layer1 series:
Inco Network, a privacy-focused L1 public chain, raised $4.5 million this February.
Octra did not mention funding.
Have you noticed that I hardly read through all the extensive technical introductions and principles of each public chain? That stuff is useless; can you achieve commercialization through this? No. Is there any practicality? No. The absence of both implies that strategic losses are inevitable. Then, it leads to the old path of VC speculation in public chains.
Therefore, the core focus is on two aspects: which VCs have invested and how much they invested. This determines whether you can participate in its airdrop with a certain safety net.
The last category is Depin projects:
Arcium, a Depin project in the Solana ecosystem, strategically raised $3.5 million this year, with many top VCs' names visible.
Privasea, Depin+AI raised $5 million in funding this March with participation from well-known institutions.
Cluster Protocol, a Depin computing protocol, did not disclose the amount raised, but it is estimated to be a small sum.
Mind Network is interesting; it is a re-staking platform that has aligned with Depin and also leverages AI. Last June, it raised $2.5 million in seed funding.
Final Summary.
After reviewing the entire sector, concepts are indeed very important, but so far, there hasn't been any impressive innovation based purely on concepts.
In the past, DeFi created liquidity mining, the metaverse gave rise to NFTs and GameFi, and even layer2 projects had airdrops. FHE has not yet produced new innovations, but it may still do so; we can only continue to monitor whether interesting applications emerge.
Besides that, what can we do? Airdrops. ZKS has entered the final phase, and at this point, PHE-related public chain projects present a good opportunity. A quick search on Twitter reveals that hardly anyone is discussing this, indicating significant speculative potential.
Public chain types and Depin types can be pursued while ensuring the security of your computer equipment.
This is the research process for retail investors regarding a new concept. The difficulty does not lie in so-called information searching but in the selection of information. This relies on your thinking system: what do we need to find and discern what is truly important.
#MindNetwork Fully Homomorphic Encryption FHE Reshaping the Future of AI