Facta non verba.

Imo, indeed, ‘convenience’ and ‘better UX’ in terms of crypto products do not mean ‘better tech with abstraction’. The key is bridging the gap between the two points: what users subconsciously expect & how the product actually works out.

Will try to bring the takeaways in eng

WE DEFINITELY NEED TOSS building crypto product!!! (Toss is the biggest fintech comp in Korea) They’ll definitely gonna nail it. If you are Korean and resonate that we need better UX for crypto, you’ll definitely know what I mean. Btw it’s a bit sad that their conference does not support English subtitles… — SIMPLICITY23

There’s lots of ways to be, as a person. And some people express their deep appreciation in different ways. But one of the ways that I believe people express their appreciation to the rest of humanity is to make something wonderful and put it out there. And you never meet the people. You never shake their hands. You never hear their story or tell yours. But somehow, in the act of making something with a great deal of care and love, something’s transmitted there. And it’s a way of expressing to the rest of our species our deep appreciation. So we need to be true to who we are and remember what’s really important to us. —Steve, 2007 https://book.stevejobsarchive.com/

Bump — some of my highlights for detailed breakdown

Brief overview on the McHenry-Thompson bill: An effort at creating a comprehensive regulatory framework for crypto Short thread here Detailed breakdown here Bulk of it was focused on the 'decentralization' aspect of crypto To this, here's some intriguing thoughts by @lex_node on how the intricate balance might be difficult to achieve. Link: https://twitter.com/JBSDC/status/1664791786038935552 Author: @JBSDC

"Just a friendly reminder, countries are going to do what's best for their citizens. US Law, theoretically, is the same on AI training data. If the West is going to appropriate Japanese culture for training data, we really shouldn't be surprised if Japan decides to return the favor."

Japan Goes All In: Copyright Doesn’t Apply To AI Training - Japan's gov't has decided not to enforce copyrights on data used in AI training to boost its development - Japan plans to become a global leader in AI, despite resistance from artists who fear this could devalue their work - It could potentially increase Japan's GDP by 50% or more - It signals Japan's desire for a reciprocal exchange of data with the West for AI development - Their stance could significantly influence global discussions on AI regulations, showcasing a major economy advocating for unrestricted AI research and development

Financial benefit doesn't stay for a long time. But technological benefit does. We should figure out how DLT tech should be intergrated in the age of AI for the industry's sake.

Some of these might indeed come to fruition, while others might not. But imo it will take a considerable amount of time to see these models in action (a.k.a. 'decentralized' versions of GC and RLHF model). While we all acknowledge their necessity, I don't believe that focusing too much on the technical side and emphasizing the degree of 'decentralization' from the outset will aid in making these models operational (as we've seen through past couple attempts).

The Convergence of Crypto and AI: Four Key Intersections 1. The "AirBnB for Graphics Cards" Model - There is a massive demand for high-performance graphics cards, leading to a market akin to an "AirBnB for graphics cards" - Challenges including the varying capabilities of different graphics cards, latency issues in a decentralized environment, and verification problems - i.e. Render Network, Akash, BitTensor, Gensyn, Prodia, and Together 2. Token-Incentivized Reinforcement Learning from Human Feedback (RLHF) - Token incentivization is proposed as a model for RLHF, especially for domain-specific models and workers with high income outside of RLHF work - Potential industries including medicine, law, engineering, architecture, finance, scientific research, education, and environmental sciences - i.e. Hivemapper 3. Zero-Knowledge Machine Learning (zkML) - With zkML, blockchains can know of events outside of the chain, enabling them to programmatically move value based on real-world states - Oracles partly solve this problem, but some data needs to be computed before going to the chain, which is a task suited for zkML - i.e. Modulus Labs 4. Authenticity in the Age of Deep Fakes - Public key cryptography, along with a public record that maps public keys to real-world identities, can help maintain authenticity and trust in digital media - Two configurations of this include embedded hardware w/ devices incorporating native hardware-based signing features and user-controlled software w/ sw providers integrating public key cryptography mechanisms - i.e. Solana Labs' Saga phone, software tools like Photoshop, Octane, and Stable Diffusion

"The most effective AI solutions are the ones that perfect the data collection process — both in terms of quantity and quality of information received. We envision a future where there are Incentivised Data Marketplaces. Decentralized computing protocols, such as Filecoin and dFinity, and distributed GPU rendering protocols, such as Render Network, are leading examples of projects that embody a common theme. With the advent of distributed AI protocols, parts of the ecosystem can be tokenized while leveraging incentives in exchange for user engagement and the sharing of data to further enhance the AI-based model. This symbiotic relationship between users, AI, and tokens holds immense potential to revolutionize the industry." - from Top 10 Trends in 2023: An Overview by Hashed

And now everyone is talking about the intersection of AI and crypto. I don't think the 1B+ adoption of crypto could be solely done with blockchain or DLT tech in broader perspective. It could rather act as a great baseline & infrastructure of next generation of blooming applications enabled by better user interface; ensuring better user protection with better sense of data sovereignty and guaranteed ownership. That's why I don't think they are not just like oil and water.

🎉 Celebrating the 2nd Anniversary of Facta non verba. 🎉 Beginning as a Korean channel primarily covering global startups (especially on SaaS and crypto) and along with real estate, it has come a long way. A year ago, I've transitioned it into an English channel to serve global audiences with extensive coverage of crypto (and some AI as the broader theme of the next internet). Posted 4K+ msgs combined for the last 2yrs. - Today, Facta non verba. is a melting pot of hundreds of industry pioneers, innovators, and enthusiasts from the fields of crypto, startups, and finance. Its reach extends mainly across South Korea, Singapore, SEA, and India (and even includes a growing contingent from the US & Europe). - Thank you for being a part of my exciting journey. Onwards to another year of growth, exploration, and discovery! 🚀

🎉 Celebrating the 2nd Anniversary of Facta non verba. 🎉 Beginning as a Korean channel primarily covering global startups (especially on SaaS and crypto) and along with real estate, it has come a long way. A year ago, I've transitioned it into an English channel to serve global audiences with extensive coverage of crypto (and some AI as the broader theme of the next internet). Posted 4K+ msgs combined for the last 2yrs. - Today, Facta non verba. is a melting pot of hundreds of industry pioneers, innovators, and enthusiasts from the fields of crypto, startups, and finance. Its reach extends mainly across South Korea, Singapore, SEA, and India (and even includes a growing contingent from the US & Europe). - Thank you for being a part of my exciting journey. Onwards to another year of growth, exploration, and discovery! 🚀

Interesting POV of Gosselin on Rollup Sequencers. Rollup Sequencers Are Centralized — And That’s Fine "Ethereum has been designed as a maximally decentralized layer-1 with relatively little economic activity on the base layer. Its goal is to settle data without what he describes as “contention” — demand for settling on a specific position — which takes place inside layer-2s instead." (...) “At the end of the day, yes, you’re going to have trade-offs in these different execution environments, but ultimately, app developers just want an interface for connecting and automatically deploying their services. These shared sequencers or decentralized block builders, cross-chain bridges, are all in the same game of trying to build and provide those services."

Feel the articles from Frontier Research is a true gem; on my list to go over whenever I get free (As some of you might know, Frontier Research is led by Stephane Gosselin, Co-Founder of Flashbots)

A well written article on DEX design, what areas of their current mechanics are good and what can be improved on. A good read to understand DEXs from the ground-up Article link Article Summary DEX's Value Proposition · Mechanisms to permissionlessly create markets for any asset hold substantial value Current DEX Landscape · AMMs have grown to be particularly useful for creating markets on pairs where there is insufficient demand for professional market makers to participate · CLOBs remain the venue of choice for trading volume on high demand pairs · > 60% of the DEX volume is still driven by MEV, including CEX-DEX arbitrage Characteristics of a Good Exchange · Trust, Best prices, Fairness, Speed, Availability, Information, Deep liquidity, Good yield To which, DEXs offer: · Decentralization (Trust & Security) · Open source settlement mechanisms (Fairness) · Open exchange histories (Transparency) Struggles of DEXs DEXs struggle to provide the following · Reliably good prices · Good yields for LPs · Prevent MEV to maintain fair execution Advantages of AMMs · Low liquidity requirements: AMM always gives a price · Passive liquidity: Anyone can LP · Simplicity: AMM has less storage and compute cost · No gatekeepers: Listing of any project Disadvantages of AMMs · High gas fees · Stale prices: Prices on AMM moves through trades and relies on arbitrageurs for price discovery · Loss vs rebalancing (LVR): AMMs are passive and lag in price updates, causing LPs to lose out to arbitrageurs. Characteristics of LVR: Permanent, Increases with volatility, Depends on price discovery level · Difficult to attract liquidity: Due to LVR causing LPs to lose out · Value extraction: Frontrun, sandwich etc. · Fragmented liquidity: Because of the multiple pools · Price and inclusion uncertainty · Fixed spread The following are some areas that can be improved on and the mechanics to go around them 1. Fixing High Gas Fees · Cheaper blockspace: L2s have lower costs · CoWs (Coincidence of Wants): Assets swapped P2P between traders who are trading complementary pairs concurrently. Example: CowSwap · Compute off-chain, verify on-chain 2. Fixing Stale Prices · Request for Quote (RFQ): Buy directly from market maker to get access to prices and liquidity both on and off chain. More gas efficient as they bypass pool routing. Examples: Hashflow and Airswap · Just-in-time (JIT) Liquidity: MMs take on less risk if they set the price after user submits a trade, allowing them to quote better tighter prices · Lower DEX Fees: Keeps prices up-to-date through arbitrage. One solution is to set prices with oracles 3. Fixing LVR · Oracle-based Pricing: Oracle updates prices on AMM quickly and accurately to reduce price difference with market. This makes arbitrages unprofitable, minimizing toxic flow and essentially LVR (less loss, less need for rebalancing) · Incentivized Delay: Allows DEX to separate toxic from non-toxic flow and adjust fees accordingly · Active Liquidity Management: Concentrated liquidity increases capital efficiency for LPs. Example: Maverick · Dynamic Spread & Volatility Oracles: AMMs can adjust fees dynamically based on current market volatility. Charge a higher fee for higher volatility to reduce the profitability of arbitrages 4. Fixing Vulnerable Settlement (MEV) · Private Submission: Privacy RPCs do not route through public mempools, preventing frontruns and sandwiching · Batch Auctions: Make prices fair. Batched orders execute at the same price, reduce frontrunning and sandwiching. Also adds latency that discourages toxic order flow · Dynamic Slippage Tolerance: DEXs can predict the right slippage with volatility and depth oracles · Make All LPs JITs LPs: Determine prices after user signs txns

Bump for the ppl who want details for MiCA of EU

Tooooo massive... but here's some of the comments what I've highlighted. - It's very rare to see two Chairs collaborate like this. In fact, it basically only happens for very marquee legislation. — link - So, anon, the fact that these two members decided to work together on this bill is very significant. It shows they want to get this passed, are aligned on the goal and the substance, and everyone should take this legislation seriously. — link - This is a significant safe harbor, protecting companies granting of tokens to a network from getting tripped up by securities laws. This could apply to airdrops too, finally letting them be more used in the U.S. — link - Under this section, many tokens will qualify as digital commodities and subject to CFTC laws and regulations rather than SEC laws and regulations, especially if they are part of a decentralized network. — link - Rather than just dividing the crypto world between the CFTC and the SEC, the bill is creating a detailed process for industry and government to do so. That’s actually a pretty elegant solution! — link - Any firm that has filed a provisional registration and is in compliance will not face SEC or CFTC enforcement action for incorrectly listing a digital asset as a digital commodity or failing to register with the CFTC. That’s a BFD. — link