WAGMAS web3summit.xyz

We're All Gonna Make a Summit! Hype-free web3 community & Summits | Season 2, Episode 1 is coming up in 2023

[WAGMAS web3summit.xyz @Web3_Summit]

WAGMAS NYC is now scheduled for the Fall of 2023.

Stay tuned for more updates at:

https://t.co/yUHssGg9wl

[nicolas, 30 ans @nic_carter]

So the whole tech world is learning about OCP2.0 today via pmarca on Rogan. What happened exactly? For those curious, let's dig into OCP as it pertained to crypto. Quick thread

[Shayne Coplan 🦅 @shayne_coplan]

Polymarket called the ceasefire news days before the media.

To be an informed citizen of the world, you can no longer afford to ignore Polymarket.

[@levelsio @levelsio]

Not your keys

Not your coins

[Ed Newton-Rex @ednewtonrex]

Nathan is (rightly) one of the most highly respected investors in AI, having gone all-in long before the current hype. AI companies would do well to listen to his advice on copyright.

This thoughtful and balanced essay is clear: the AI industry “has a good chance of losing” the fight over copyright, and model builders should be exploring deals rather than burying their heads in the sand.

[Ed Newton-Rex @ednewtonrex]

The 3 major record labels are suing AI music companies Suno and Udio. Here are the two lawsuits in full.

- They accuse Suno & Udio of “willful copyright infringement on an almost unimaginable scale”
- They provide evidence that both companies trained on their music, including outputs that closely resemble their recordings (ABBA, Michael Jackson, Green Day, James Brown, & many more)
- They outline why this is not fair use
- They say this “wholesale theft of… copyrighted recordings threatens the entire music ecosystem and the numerous people it employs”
- They include unknown co-defendants who assisted in copying/scraping
- They demand a jury trial

If you do one thing today, read the full complaints.

🧵 of some of the outputs 👇

Suno: https://t.co/V2rZju4BeK
Udio: https://t.co/RiXFsP1qIQ

[Luiza Jarovsky, PhD @LuizaJarovsky]

🚨BREAKING: @UMG, @capitolrecords, @sonymusic, @warnermusic, @AtlanticRecords, @Rhino_Records & other major record labels sue AI music program Suno AI for copyright infringement. Important quotes:

"Given that the foundation of its business has been to exploit copyrighted sound recordings without permission, Suno has been deliberately evasive about what exactly it has copied. This is unsurprising. After all, to answer that question honestly would be to admit willful copyright infringement on an almost unimaginable scale. Suno’s executives instead speak publicly in exceedingly general terms. For example, one of Suno’s co-founders posted online that Suno’s service trains on a “mix of proprietary and public data,” while another co-founder stated that Suno’s training practices are “fairly in line with what other people are doing.” Piercing the veil of secrecy, an early investor admitted that “if [Suno] had deals with labels when this company got started, I probably wouldn’t have invested in it. I think that they needed to make this product without the constraints." (page 3)

-

"Plaintiffs could have proceeded with this action based solely on eliciting that reasonable inference of copying. Nevertheless, Plaintiffs’ claims are based on much more. In particular, Plaintiffs tested Suno’s product and generated outputs using a series of prompts that pinpoint a particular sound recording by referencing specific subject matter, genre, artist, instruments, vocal style, and the like. Suno’s service repeatedly generated outputs that closely matched the targeted copyrighted sound recording, which means that Suno copied those copyrighted sound recordings to include in its training data. In addition, the public has observed (and Plaintiffs have confirmed) that even less targeted prompts can cause Suno’s product to generate outputs that resemble specific recording artists and specific copyrighted recordings. Such outputs are clear evidence that Suno trained its model on Plaintiffs’ copyrighted sound recordings." (page 5)

-

"At its core, this case is about ensuring that copyright continues to incentivize human invention and imagination, as it has for centuries. Achieving this end does not require stunting technological innovation, but it does require that Suno adhere to copyright law and respect the creators whose works allow it to function in the first place. Plaintiffs bring this action seeking an injunction and damages commensurate with the scope of Suno’s massive and ongoing infringement. (page 6)

➡️Read the full lawsuit below.

➡️To stay up to date with the latest developments in AI policy, regulation, and lawsuits, subscribe to my newsletter (link below)

[Laura Shin @laurashin]

How Coinbase Might Bring 1 Billion People Onchain Using Its Smart Wallet

On @Unchained_pod, @LukeYoungblood talks about the 🔥 new Coinbase smart wallet:

🛠️ How smart wallets differ from traditional wallets
💸 How @coinbase is subsidizing gas fees to onboard users
🌍 Why @MoonwellDeFi focuses on emerging markets
👀 Coinbase’s Magic Spend feature for seamless transactions

Timestamps:
01:40 How smart wallets differ from traditional wallets and embedded wallets
04:06 Why Luke is excited to be working with Coinbase smart wallet
06:33 What happens if the user loses the device linked to their smart wallet
08:12 How hard it would be for a hacker to try to get access to the assets in this smart wallet
09:36 How Coinbase is initially paying for user gas fees on dapps like Moonwell and other launch partners
12:02 How Coinbase’s Magic Stand feature enables users to transact onchain straight from their Coinbase accounts
14:24 How Coinbase might keep paying gas fees for some dapps even after the initial launch period
16:38 Why the smart wallet is also accessible via the web, and not just through an app
19:50 Why Moonwell has focused on lending and borrowing
18:03 Why Moonwell chose to build on Base as opposed to, say, Solana
21:24 Moonwell’s plans to grow
23:00 How having access to Coinbase’s user base changes Moonwell’s strategy for attracting users

[Bankless @Bankless]

BlackRock is ensuring that Real World Assets could be the year's biggest crypto narrative

This is what's getting tokenized + the standout projects to keep an eye on 👇

Are real-world assets crypto’s next frontier?

Real World Assets (RWAs) are the talk of the town. Big names like @BlackRock are getting into RWAs, @tether is rolling out its own platform, and the numbers are impressive too – RWA tokens reached a record market cap of $2.7 billion in February.

With so many major catalysts, the RWA narrative is expected to take center stage in the second half of 2024.

Today’s story gives a rundown of the RWA landscape, highlighting the different types of real world assets being tokenized and the standout projects to keep an eye on.

Let’s dive in! 👇

What's Next for RWAs?
RWAs are one of the fastest-growing sectors in crypto. Many see them as a way to tap into the world's trillions in assets, fueling the growth of the entire industry.

Larry Fink, the CEO of BlackRock, has called RWAs "the next generation for markets”. @BCG predicts that by 2030, turning these assets into tokens could unlock a $16 trillion opportunity.

However, until now, only one asset has been successfully tokenized and fully integrated into the crypto ecosystem: fiat currency in the form of stablecoins.

Stablecoins are the first, the largest, and the most established RWA. They’ve found product market fit in crypto, see strong demand for different services, and are a fundamental part of every crypto ecosystem.

But what's beyond stablecoins in the RWA space?

In recent years, we've seen a growing trend of various real-world assets being tokenized and brought onchain. Let's check out some of the most popular types.

Commodities, Equities & Funds
Commodities like gold, silver, and crude oil are commonly traded on various exchanges around the world. These natural resources can be tokenized to represent a stake in the actual commodity, much like stablecoins do for fiat currency.

So far, precious metals, especially gold, have gained the most traction in crypto as RWAs. Tokens like PAX Gold (PAXG) and Tether Gold (XAUT) are leading the charge, with gold-backed RWAs comprising 83% of the commodity token market cap, as reported by @coingecko.

This dominance of gold indicates just how nascent the RWA sector is. However, several projects are experimenting with different commodities. For instance, the @Uranium3o8 project has introduced a token pegged to the price of a pound of U3O8 uranium compound 👀

As the RWA tokenization space matures, we might see tokens for other commodities like crude oil and even crops like corn. The thesis is that more global trading will shift onto the blockchain in the future.

Similar to commodities, stocks, and mutual funds can also be tokenized. These assets are mainstays in the traditional finance market, but their adoption in crypto has been slow, largely due to regulatory hurdles.

Complying with laws across jurisdictions is tough, with many projects needing licenses and facing restrictions, like excluding users from certain countries or meeting strict KYC and AML criteria.

Despite these challenges, some projects like @SwarmMarkets and @BackedFi have navigated the regulatory maze, allowing onchain trading of global stocks and funds, like $COIN and $NVDA from the U.S. markets, and index funds like the Core S&P 500, among others.

Treasuries
Treasuries refer to tokenized government debt instruments. Traditionally, these instruments are secure, yield-generating assets as they are issued by governments.

Following the COVID-19 pandemic, treasury rates, which were historically low, saw a rise as the Federal Reserve adjusted its monetary policy to the shifting economy. By October 2023, short-term treasury yields had climbed from near zero to approximately 5.4%.

This uptick in rates spurred the launch of projects tokenizing U.S. treasuries, with some notable examples being:

Franklin Templeton — @FTI_US launched the Franklin OnChain U.S. Government Money Fund (FOBXX) in 2021, the first U.S.-registered fund on a public blockchain. It offers a 5.11% yield and has a $365 million market cap, ranking it among the largest onchain treasury products.

BlackRock — launched the BlackRock USD Institutional Digital Liquidity Fund ($BUIDL) in March 2024 on Ethereum. It currently leads the onchain treasury fund market with over $375 million in assets under management.

Ondo — @OndoFinance launched the Ondo Short-Term U.S. Government Treasuries (OUSG), which provides access to short-term U.S. Treasuries with a 4.68% yield and a market cap around $140 million. A significant portion of OUSG is invested in BlackRock's BUIDL. Ondo also offers the USDY yield-bearing stablecoin, with a market cap exceeding $120 million.

This category has seen substantial growth as treasury yields have become more appealing with rising interest rates. Other notable projects in this space include @superstatefunds, @maplefinance, Backed, @OpenEden_Labs, and more.

Onchain Private Credit
Private credit involves lending by financial institutions to businesses through debt instruments, essentially loans.

In the context of the RWA sector in crypto, these loans are tokenized through credit protocols, allowing lenders to extend capital to these institutions in exchange for yield.

In traditional finance, private credit is a massive $1.6 trillion market, and it's slowly carving out a significant niche in crypto.

Crypto credit protocols have already tokenized over $4.4 billion in loans, with more than $600 million currently loaned out to real-world businesses, generating returns for onchain lenders.

For onchain investors, private credit presents an attractive proposition due to its higher yield potential. For example, lending stablecoins through a protocol like @centrifuge can yield an average APY of 8.7%, surpassing the 4-5% APY typically found on platforms like AAVE, though it comes with increased risk.

Key players in the private credit space include:

- Centrifuge
- Maple
- @goldfinch_fi
- @Credix_finance
- @TrueFiDAO
- @homecoinfinance
- @ribbonfinance

Real Estate
The real estate category within RWA focuses on tokenizing physical properties like residential houses, land, commercial buildings, and infrastructure projects.

Real estate stands as the world's biggest asset class. But traditionally, real estate investment requires significant capital due to the high cost of properties. Making real estate tradable onchain by tokenization introduces a novel investment paradigm, enhancing accessibility, enabling fractional ownership, and potentially increasing liquidity.

Nonetheless, real estate's inherent illiquidity has tempered the pace of its onchain adoption. The protracted nature of real estate transactions and the small buyer pool make it challenging to align sellers with buyers onchain, especially given how the sector has operated on legacy systems traditionally.

Projects like @RealTPlatform are striving to inject liquidity into the market by simplifying property fractionalization, thus allowing sellers to easily divide their assets and enabling buyers to acquire tokenized shares.

Additionally, platforms like @Parcl allow for speculation on the value of real estate across various locations, such as different U.S. cities, through their onchain trading mechanisms.

All these initiatives will make the real estate market more liquid in the long term.

Closing Thoughts
The concept of RWAs promises to bring a new level of global access and liquidity to traditional assets like real estate, commodities, and debt. Should the bold predictions for RWAs materialize, the tokenization of these assets may well redefine how the world trades these assets and in turn, bring more people onchain.

[Chomba Bupe @ChombaBupe]

The problem with generative AI is that it never solves problems of its own to share that experience with others, it needs tons of data to copy from & pretend to have the experience to share with prompters.

That's why getting information from real humans is irreplaceable.

[Luiza Jarovsky, PhD @LuizaJarovsky]

🚨 AI Policy Alert: The German Federal Office for Information Security publishes the report "Generative AI Models - Opportunities and Risks for Industry and Authorities." Quotes & comments:

"LLMs are trained based on huge text corpora. The origin of these texts and their quality are generally not fully verified due to the large amount of data. Therefore, personal or copyrighted data, as well as texts with questionable, false, or discriminatory content (e.g., disinformation, propaganda, or hate messages), may be included in the training set. When generating outputs, these contents may appear in these outputs either verbatim or slightly altered (Weidinger, et al., 2022). Imbalances in the training data can also lead to biases in the model" (page 9)

-

"If individual data points are disproportionately present in the training data, there is a risk that the model cannot adequately learn the desired data distribution and, depending on the extent, tends to produce repetitive, one-sided, or incoherent outputs (known as model collapse). It is expected that this problem will increasingly occur in the future, as LLM-generated data becomes more available on the internet and is used to train new LLMs (Shumailov, et al., 2023). This could lead to self-reinforcing effects, which is particularly critical in cases where texts with abuse potential have been generated, or when a bias in text data becomes entrenched. This happens, for example, as more and more relevant texts are produced and used again for training new models, which in turn generate a multitude of texts (Bender, et al., 2021)." (page 10)

-

"The high linguistic quality of the model outputs, combined with user-friendly access via APIs and the enormous flexibility of responses from currently popular LLMs, makes it easier for criminals to misuse the models for a targeted generation of misinformation (De Angelis, et al., 2023), propaganda texts, hate messages, product reviews, or posts for social media."

➡️ According to the report, special attention should be given to the following aspects:
➵ Raising awareness of users;
➵ Testing;
➵ Handling sensitive data;
➵ Establishing transparency;
➵ Auditing of inputs and outputs;
➵ Paying attention to (indirect) prompt injections;
➵ Selection and management of training data;
➵ Developing practical expertise.

➡️ Of the dozens of AI reports published lately, this one is especially detailed regarding AI-related risk and potential countermeasures.

➡️The document is a must-read for people developing AI or working on AI policymaking and regulation, especially pages 8-28.

➡️ Link to the @BSI_Bund report below.

➡️ For more information on AI policy and regulation, subscribe to my weekly newsletter (link in bio).

[Ed Newton-Rex @ednewtonrex]

A reminder that the general public don’t buy the argument that, just because humans can learn from publicly available work, AI models should be able to.

In the largest public poll on this question I’ve seen (2,052 people in the US), when people were asked how AI companies should be able to train models, ‘on any text or images that are publicly available’ was almost the least popular answer. But it was by far the most popular answer when people were asked how humans should be able to learn.

Source: https://t.co/ln7POqQeH6

[BioTone ZKI @AutoPilotCyber]

Post-PKI is real.

The economic benefits of a Quantum Leap in cyber _security_ are irresistible.

Ask us how you can start to enjoy them!

[Ed Newton-Rex @ednewtonrex]

Generative AI products usually collect ratings and feedback on output from users (👍/👎 etc.). This is generally called preference data. They use it to improve their models, and it makes a big difference.

If their models are trained on copyrighted work without permission, this means they are generating valuable preference data via what many consider to be copyright infringement.

If you train a model on copyrighted work without permission, you shouldn’t be able to use *any* downstream data you get from this. As we regulate models that exploit copyrighted work, we need to ensure we don’t neglect downstream data (preference data, synthetic data) that is acquired via this exploitation.

[Ed Newton-Rex @ednewtonrex]

Artistwashing: when you solicit positive comments about your generative AI model from a handful of creators, while training on people's work without permission/payment.

[b/acc, context platform engineer @AccBalanced]

@chiefaioffice Best example yet that content, both professional and individual, has tangible value 💰

@sharetoCLICK empowers everyone to sell licenses to their personal content.

Do what you love
Own what you do
Right where you are

[b/acc, context platform engineer @AccBalanced]

I’m applying this to first-principles encryption. PKI and asymmetric encryption was appropriate for the 1980s. No more.

Today PKI is an unmitigated mess of risk via excessive complexity and misplaced trust. PKI defeats zero trust.

Cleaner simpler lighter-weight ubiquitous zero knowledge symmetric keys, over any cipher, is the way:

[b/acc, context platform engineer @AccBalanced]

@UK_Daniel_Card Yes. But I’m feeling young again by creating the 21st century version of this that’s Post-PKI, via Autonomous Key Management (AKM).

Zero-knowledge symmetric cryptography for all. Already Post-Post-Quantum 😎

[b/acc, context platform engineer @AccBalanced]

Looking for an engineering job at the next cyber unicorn? #PostPKI
https://t.co/ykFej0ifoR

[b/acc, context platform engineer @AccBalanced]

@HackingDave More incentive to usher in the Post-PKI era. Keys to the kingdom are not a cyber liability when they are autonomous, zero-knowledge, ephemeral and quantum random:

[BioTone ZKI @AutoPilotCyber]

Why transition cryptography twice before the end of the decade, when you can skip another asymmetric encryption transition (Kyber) and graduate directly to accessible symmetric-only cryptography for all?

The post-PKI era is here.

https://t.co/uWEaLkDXPG