Igor Artamonov

https://lnkd.in/gMNUNJ_G The OFAC has ruined ETH. I can no longer view ETH as a real crypto due to blocks being censored by the OFAC. It is no longer distributed in nature but centralized. Remember Ethereum's original pitch – as a network where "code is law," where software replaces middlemen, and where "decentralized" networks can operate outside constraints from "centralized" companies and governments.

Do you know how much simpler #P2P transfers with #Web3 is than this? Lots. Do you know how much #cheaper it is with #Web3? Lots. Do you know how much more #secure it is with #Web3? Lots. We’re talking orders of magnitude lower cost, faster, and more secure with #Web3.

15

3 Comments

Kazakhstan has taken a significant step in regulating the crypto industry, liquidating 36 exchanges with a reported $100M turnover and seizing $4.8M in USDT. This bold move underscores the growing focus on oversight and compliance in the crypto space. #CryptoRegulation #Kazakhstan

1

A new player in the 1000 tokens/sec. SambaNova Systems is a competitor to Groq. They make hardware chips specialised for serving LLMs and have their own LLMs, such as Samba CoE and language-specific models. There are multiple options to try: - The main page https://sambanova.ai/ - The Sambaverse, where they offer a playground to compare two models at the same time: https://lnkd.in/eAHqXkys The Sambaverse offers an API key, but there is no pricing or info about the rate limits, so I asked in their discord channel, and the answer was: "Sambaverse is for proof of concept work, just to test how everything works. When ready to deploy into production, you’ll need dedicated resources provisioned." It is promising to have more companies working on hardware chips to improve the inference speed of LLMs. #ai #llm

29

How could BTCUSD go to $1 Million? As of 1244hrs, 11th of Aug 2024, BTCUSD is $60,983.23 with a total supply of 19,738,290 BTC, equating to a market cap of $1.2 Trillion. If BTCUSD is to go to $1,000,000 it would require an additional inflow of $18.5 Trillion at the above rate. Yes, that is how capital works, like energy. “Energy cannot be created or destroyed, it can only be changed from one form to another.” – Albert Einstein. So then where can we get $18.5 Trillion ? (1) State reserves that are a portfolio of foreign currencies, gold, SDRs and IMF deposits. These reserves help shield the local economy from foreign shocks via currency open market operations and to offset local economic shocks or unexpected budget deficits. The latter is the reason the mathematically challenged, namely, Crypto maximalists, Trump and RFK Jr., claim that Bitcoin could solve America’s out of control debt problem. But the math does not lie, the U.S. only has about $0.2 Trillion of such reserves. Even China only has about $3.4 Trillion while Japan a mere $1.2 Trillion. So this will not work! (2) Central banks manage a portfolio of only government securities bought in secondary markets and are not allowed to participate in primary issuances of such securities neither in other asset classes unless during exigent circumstances such as war or a pandemic. But the math does not lie, the Fed only has a $7.8 Trillion balance sheet of such securities which if not used to fund government securities, the state will default. If the state was to do Quantitative Easing(QE) 5 to acquire BTC, it would need to “print” 3X the total of QE1+QE2+QE3+QE4. So this will not work either! (3) Commercial banks which beyond the mandatory 10% reserve requirement, have assets largely in the form of risk-minimal cash and cash equivalent(CCE) instruments such as government bonds, money market funds, reverse repos and top grade corporate bonds. However, the liabilities side of their balance sheets are mostly the deposits of its clientele, i.e. our monies which cannot be invested in high risk instruments without the explicit approval of the ultimate beneficiary. But the math does not lie, even assuming commercial banks could invest all of their annual net income, the top 10 banks in the world collectively will only have about $0.7 Trillion. That would take a good 25 years! In all, if you still think BTC will go to $1M, then prove it!

12

2 Comments

Before introducing new terms, it might be a good idea to lookup existing catalogues of financial instruments? Just because asset registers (aka tokens) are operated on blockchain now, doesn't mean that the financial instruments didn't exist before already? "ISO 10962, known as Classification of Financial Instruments (CFI), is a six-letter-code used in the financial services industry to classify and describe the structure and function of a financial instrument (in the form of security or contract) as part of the instrument reference data. It is an international standard approved by the International Organization for Standardization (ISO)."

28

5 Comments

How does Payment Tokenization work? How does it guarantee security for sensitive client information? . The diagram below shows the tokenization process in Apple Pay. The PCI Council defines tokenization as "a process by which the primary account number (PAN) is replaced with a surrogate value called a token. The key is that only trusted parties know the true PAN of the client. Other parties, including the merchants, PSPs, and Apple, don’t know client-sensitive information. 🔹 Add a Card to Wallet Steps 1-3: A user adds a credit card to Apple Wallet. The card details including PAN are sent to the Apple Pay server. Apple Pay identifies the issuing bank from PAN and sends card details to the issuing bank. Step 4: The issuing bank requests a Payment Token from TSP (Token Service Provider). A TSP is an entity that registers with EMVCo. Step 5: The TSP validates PAN and generates Payment Token, Payment-Token-Key (public key). It also vaults PAN for lookups. Steps 6 - 8: The issuing bank receives tokens from TSP and generates CVV-key. These keys are returned to the Apple Pay server, and then to the Apple Wallet. Step 9: Apple Pay provisions the tokens onto the SE (Secure Element), which is a secure chip on iPhone to store sensitive client information. The Payment Token stored on SE is called DAN (Device Account Number). 🔹 Pay with the Card Step 1: Merchat setup - The merchant that supports Apple Pay gets certificates from payment processors. Step 2: The user authenticates themselves using biometrics. Step 3: When the user pays for a product, the information stored in SE is sent to the merchant’s PoS. Steps 4 - 6: The transaction is sent to the payment processors with an encrypted Payment Token. The payment processors decrypt the information and send it to the acquiring bank, which then sends it to the card network. Step 7: The card network looks up PAN in the PAN vault, and decrypts dynamic cryptogram. Step 8: The card network sends PAN, dynamic CVV, and the transaction to the issuing bank. Steps 9 - 10: The issuing bank validates and authorizes the transaction, and sends the authorized response to the card network. 🔹 Security - Apple Pay doesn’t store PAN on iPhone or Apple Pay servers. - DAN (Payment Token) only resides on SE. - Only the TSPs know about the mapping from Payment Token to PAN. — 👉 If you like my FinTech posts, please follow Hua Li and click the 🔔 on my profile to receive updates. hashtag #systemdesign hashtag #payments hashtag #fintech.

37

3 Comments

#DeFi protocols are like building blocks - you can put them together and build sophisticated, structured products. They are also permissionless and instantaneous, which enables all sorts of experimentation and automation. Index Coop leverage tokens - like ETH2x - are a great example of this: #Aave, #Uniswap, and #IndexProtocol working together seamlessly under the hood. #Composability is a killer feature for #crypto. You are probably underestimating the impact open innovation will have on the future of finance.

21

All our GTM strategies are based on market statistics from blockchain-based web2 solutions like #Arbitrum One, #Circle CCTP, #Wormhole, #Avalanche, etc. But something goes wrong in this area, and we understand why. Only 97 users used the #StarkNet #Bridge yesterday. Can you imagine that? SWAPping 100,000 #Ethereum-based #USDT to #BSC-based USDT costs 224 USDT in the #Synapse and 10 USDT in the #Across. And 0 USDT in any #CEX. Why do all influencers and KOLs keep silent about #crosschain data and asset transfer drama that drives us into total slavery to #web2 solutions? It's not only about cross-chain transfer but about the whole #web3 world. Perhaps, they don't understand that the market needs pure web3 cross-chain solutions that most numbers of #blockchain foundations will support. Perhaps we should stop thinking only about one-time personal profit and start thinking about long-term market profit that adds value to all of us. Just think about it. Web2 ruining web3 and all that we have now #Bitcoin. Let's add value to web3 together, from now on. https://lnkd.in/d6P3UxK9

2

How do you decide which provider to use for open-source AI models? If you want to use Llama3 70B as an example, there are many providers now where you can pay per 1M tokens. However, choosing the optimum one would depend on multiple factors, such as: - Max tokens - Price. - Latency - Integration (whether it is a direct integration or via your preferred LLM gateway platform). One way to simplify the decision is to look at the OpenRouter ranking to see which models are trending within a week or a month, then go to the model status to see all the providers, max output, input/output pricing, latency and throughput. #llm #ai #opensource

11

Permissionless Ethereum networks are suffering from validators using MEV (Maximal Extractable Value) bots to front-run user transactions, leading to financial losses or at least negative impacts on users’ trades. Recently, in the third consultation paper on MiCA, ESMA has clearly declared this behavior as market abuse (page 10, item 19). Market participants will be required to monitor and report MEV instances, adding extra operational burdens to honest actors. At SWIAT, we use a permissioned Ethereum network with a reputable set of validators operating within a clear governance framework. Additionally, a network coordinator monitors and enforces the governance terms. This relieves regulated entities from the burden and simplifies the risk assessment of the infrastructure, which is crucial for scaling blockchain use cases in the regulated environment. Until all responsibilities can be clearly encoded into technology, regulation, governance, and policymaking will remain essential! #blockchain #ethereum #mev #abuse #esma

21

9 Comments

Federal Reserve Board - Cross Border Venture Capital and Reverse Technology Flow Source: https://lnkd.in/gBAy2xK9 #crossborder #venture #capital #reverse #technology #flow Credits: Ufuk Akcigit Sina Teoman Ates Josh Lerner Rick Townsend Yulia Zhestkova Grigsby Highlights: We find that as a country increases its knowledge of a technology class relative to the U.S., it also becomes less likely to invest in a U.S. startup innovating in that technology class. Overall, these results show that foreign countries tend to invest in U.S. startups that specialize in technological areas where they are behind the U.S., and these investments appear to be strategic 'responses' to address the technology gap, at least partially. Next, we seek to understand whether the cross-border investments into U.S. startups foster innovation and knowledge spillovers back to the financing country. The results suggest that there are benefits from these investments in the form of knowledge creation and spillovers back to the financing country. Thus, the patterns are consistent with the idea that investing countries learn from U.S. startup investments. Follow-on analysis also brings out that investing foreign firms increase their citations to the patents of the U.S. startups in which they invest after the investment event, providing more direct evidence of learning for the investing firm itself. Overall, the results indicate that both the investing firm and also other firms from the investing country sharing the same technology benefit from increased spillovers following an investment event. In technology classes that are more reliant on such knowledge, spillovers could be expected to be stronger. Our analysis illustrates that this conjecture is indeed true, as the effects are found to be larger in high-basicness classes. Similarly, knowledge flows appear to be greater in classes that contain patents subject to a secrecy order from the federal government. These patterns of patenting and citations suggest the real knowledge transfer is taking place, and cross-border investments are a means to access that knowledge. Finally, we examine the consequences of these investments. While we cannot demonstrate causality directly, more foreign investments in firms specializing in a technology class are associated with greater subsequent patenting by U.S. startups in this class. These results are at least consistent with the idea that such investments ease capital constraints for the recipient firms, as hypothesized. As such, our empirical analysis reveals that foreign CVC investments constitute another mode of technology transfer between countries, with the recipient firms of these investments also benefiting from them in terms of additional capital

10

IMHO: A token is just a good old book-entry that happens to be on blockchain/DLT. Blockchain/DLT is just a shared ledger that happens to get its integrity maintained by a peer-to-peer network instead of a trusted third party. How the book-entries are interpreted gives them a meaning. Is it a bearer bond? Registered bond? Apples? Pears? So if we use ERC-20 to represent a bearer bond, then it's still a bearer bond that happens to be represented as a book-entry on a blockchain using the ERC-20 standard. So why bothering calling it a token. It's a bearer bond. ERC-20, ERC-721, UTXO, Runes, etc., for me these are just different technical representation of the transaction and account book. That said calling ERC-20 a *token* standard is extremely misleading. Any accountants out there to give it a better name? 😀

18

13 Comments

🎯 Zero-knowledge proofs(ZKPs) at the infrastructure level 📍 ZKPs disrupt the Blockchain at the infrastructure level: zkEVM, Mina protocol, zkVM(Joltn SP1, RISC Zero, etc), zk rollup(Scroll, Aztec, etc), ZK Coprocessor(Lagrange, Axiom, etc), zero-knowledge proofs verification layer(Aligned), ZK Proof Aggregation(Electron, Nebra), Compute network for zero-knowledge proofs(Gevulot), Proof of storage(Filecoin network), zkML(EZKL), etc. Will ZK Eat The Modular Stack? https://lnkd.in/d65mHzmD 📍 There are different ZK proof protocol implementations (plonky3, halo2, etc.) and ZK proof recursion and aggregation approaches, such as the folding scheme. We have different approaches to developing ZK-based computer programs: zkVMs, low-level ZK circuit programming(Halo2/Circom), and high-level ZK application programming(o1js, noir). 📍 Today, most innovation in ZK is at the infrastructure level(performance, maturity, hardware acceleration). This is the nature of things because we’re at an early stage of the technology(we have the first practical and in-production ZK system: the first ZK era). Over time, we will have more performed and matured ZK infrastructure, tools, languages, and libraries. ZKP is transforming the current blockchain infrastructure to be adopted at scale and at a low cost(trust-minimization).

33

2 Comments