AWS Preps ‘Bastion’ Cloud Service for Advertisers

Amazon Web Services is preparing to unveil a cloud service to help companies improve the way they target ads to potential customers without violating data privacy laws, according to three people with knowledge of the product.

The move comes as advertisers try to recover from, and get ahead of, Apple’s and Google’s restrictions on their ability to track consumers online.

The new AWS service, Bastion, is known in tech industry parlance as a data clean room. It lets multiple companies pool data they have on existing or potential customers without any of them being able to view the entire pool.

The idea is to protect the identity of customers, both for privacy reasons and out of competitive concerns. In these metaphorical rooms, companies pooling data—for instance, a retailer like Target and a streaming service that sells ads, such as HBO Max or Hulu—could see how much overlap there is between their respective customers and use that to determine whether Target should target new or repeat customers through ads on the streaming services.

The Ethereum Foundation has revealed that the long-awaited merge will be completed between Sept. 10 and 20.

This means that the historic switch from a proof-of-work to a proof-of-stake blockchain could take place a few days earlier than previously advertised.

It comes after a series of repeated delays — and it's worth remembering that this milestone was meant to have been achieved back in 2019.

All of this has meant those who have staked Ether so they can run a validator node have faced a long wait while their crypto was locked up in the deposit contract.

Confirming that the merge is imminent, the Ethereum Foundation declared:
"Following years of hard work, Ethereum's Proof-of-Stake upgrade is finally here! The successful upgrade of all public testnets is now complete, and the merge has been scheduled for the Ethereum mainnet."

In other developments, the foundation has also announced that it is quadrupling the bug bounties that are rewarded to people who discover vulnerabilities before the merge.

All of this means that those who uncover critical bugs could end up receiving $1 million.

‘Reality Pro’, ‘Reality One’ and ‘Reality Processor’ - a set of shell companies that are linked to Apple has secretly trademarked potential names for Apple’s upcoming AR/VR headset.

Trademark filings suggest that Apple may be staking claim to potential names for its highly anticipated mixed-reality headset, part of the tech giant’s push into its first new product category in years. 

Applications were filed in the US, EU, UK, Canada, Australia, New Zealand, Saudi Arabia, Costa Rica and Uruguay for the names “Reality One,” “Reality Pro” and “Reality Processor.”

Though Apple itself didn’t make the filings, they follow a pattern that the iPhone maker has used in the past—including relying on law firms that the company has previously enlisted to lock down brands.

Reality One, Reality Pro and Reality Processor trademarks.Source: US Trademark and Patent Office
Apple’s headset is expected to combine virtual and augmented reality technology and vault the company into closer competition with Meta Platforms Inc., the leading provider of VR gear. It’s been seven years since the company last went after a new hardware category with the Apple Watch.

The new trademarks are all registered to a shell corporation named Immersive Health Solutions LLC that was incorporated in February. That company itself was registered by another Delaware shell corporation, the Corporation Trust Co., typically used for filings by firms looking to avoid detection. The RealityOS trademark used that same firm.
In order to file the trademark paperwork, a process that began in Canada with an original application in February and in several other countries in August, the company behind the trademarks relied on several big-name and boutique law firms in each operating nation.
In the US, Canada and New Zealand, the filer enlisted law firms that Apple has used in the past to either register trademarks or for other matters. In New Zealand, for instance, the law firm Simpson Grierson was used for the “Reality” filings. Apple relied on that same firm to file the corporate name Apple Sales New Zealand. 

Samsung slates crypto exchange launch in South Korea for 2023

Samsung’s investment arm is eyeing the launch of its own cryptocurrency platform, South Korean media reported on Monday.

Samsung Securities is one of 7 domestic brokerages looking to start operating a crypto exchange in the first half of next year. Talks with local authorities over securing licensing are said to be in advanced stages.

Mirae Asset Securities and five other domestic companies are also planning to launch their own platforms.

Its previous attempts to launch an exchange were stalled last year when the business was reportedly unable to hire the right professionals to run it.

But the company has continued to keep pace with developments in the digital asset industry.

Earlier this month, Samsung Securities was one of three financial institutions in South Korea to partner with the country’s largest exchange, Bithumb. The tie-up means Samsung Securities customers can check the status of crypto investments through Samsung’s own app.

Samsung Securities was also the first to launch a global blockchain exchange-traded fund (ETF) in Asia.

Samsung has been working on blockchain projects across several parts of its operations in the last couple of years.

In May, the technology giant partnered with Nifty Gateway to add NFT functionality to its smart TVs. The release of a new smartphone in February was also accompanied by a commemorative NFT.

It also took its first steps into the Metaverse in January with the opening of a virtual store in Decentraland.
Meanwhile, its venture funding arm, Samsung Next, has invested in numerous blockchain-related companies, including Axie Infinity, the Bored Ape Yacht Club, FTX, and SuperRare.

Pfizer Ventures has applied to VitaDAO's Institutional Genesis Raise proposing to contribute $500,000

Pfizer plans to support VitaDAO beyond the capital contribution with:

• Access to dealflow
• Ability to support incubation and commercialization efforts of the intellectual property (IP)
• Ability to raise awareness for our mission and grow our community

⚡️Y Combinator Names Venture Capitalist Garry Tan As Its Next President

The storied Bay Area-based accelerator has chosen venture capitalist and Y Combinator alum Garry Tan to serve as the fourth leader in its seventeen-year history. Come January, Tan will replace president Geoff Ralston, who took the mantle in 2019.

When he takes over the top job at Y Combinator, Tan will step away from a full-time role at Initialized Capital, the firm he created with Reddit cofounder Alexis Ohanian in 2012. An early investor in cryptocurrency exchange Coinbase, grocery delivery service Instacart and supply chain software unicorn Flexport, Tan has appeared on the Forbes Midas List for the past four years, most recently ranking at No. 28. He debuted at No. 4 on the inaugural Forbes Midas Seed List in 2022.

Meta now lets you post NFTs and other digital collectibles to Facebook

Users of Facebook can now post their digital collectibles and NFTs to their accounts on the platform. 

Meta, the parent company behind social media platforms Facebook and Instagram, announced Monday that it allowing users to add digital wallets and virtual assets to Facebook. The firm's move now brings NFTs to Facebook's 2.9 billion users, according to Statista.

This update means that users can post their NFTs across both Facebook and Instagram, Meta tweeted. 

Before today's move, Meta had only allowed users to post their digital collectibles on Instagram.

Instagram's digital asset integration expanded across 100 countries and included the Ethereum, Polygon and Flow blockchains. 

U.S. Federal Reserve to launch 'Fednow' instant payments system between May and July 2023

Seven Seven Six, the venture capital firm led by Reddit co-founder Alexis Ohanian, plans to raise $177.6 million for a new fund focused on investments in cryptocurrency tokens.

The new fund, named Kryptós, will be the two-year-old firm’s first focused exclusively on crypto.

It will seek to take advantage of the recent monthslong sell-off in cryptocurrencies like bitcoin and ethereum, which has extended to the digital tokens issued by blockchain startups.

“Now is potentially a unique opportunity to invest in strong founders at a discount,” according to the fundraising materials. The firm expects to launch the fund in October, according to the materials.

TSMC says 3D chip stacking packaging techs as vital as cutting-edge chip production. 3D stacking capacity will 20x by 2026. It will build 23 plants, more than 5 each year, from 2020 to 2023, compared with 2 on average in 2017-19.

Bitcoin’s rising correlation with the tech-heavy Nasdaq 100 since November has been mirrored by falling relative performance.

The BTC to Nasdaq ratio has been declining since the start of the year. This suggests that BTC is clearly perceived as a riskier asset

Japan’s regulator proposes tax breaks for investors, crypto

Japan’s financial regulator proposed easing corporate tax rules for crypto assets as well as lighter levies for individual stock investors in support of Prime Minister Fumio Kishida’s efforts to reinvigorate the economy.

Companies should be exempted from paying taxes for paper gains on crypto coins that they hold after issuing them, the regulator proposed in its annual tax-code.

The Financial Services Agency also called for boosting a program that gives tax breaks to individual investors.

The moves support Kishida’s “New Capitalism” vision, which seeks to boost the world’s third-largest economy. He has pledged to double the wealth of households while offering support to help the country’s so-called Web3 businesses grow.

Crypto lobbying groups have been calling for changes, saying high corporate taxes have raised the bar for launching projects in Japan, causing some companies to relocate to Singapore and elsewhere.

Currently, profit from cryptocurrency holdings, including unrealized gains, is subject to corporate tax of about 30%.

For retail investors, the FSA wants to expand a tax break initiative known as the Nippon Individual Savings Account, by raising its investment limits and making the program permanent. Under NISA, individuals can have some of their investment gains and dividends exempt from capital-gains tax over a period of time.
 
The move marks the latest in years of efforts to prod individuals to put their savings to productive use, such as investing in stocks for the benefit of the broader economy. Japanese households hold roughly half of their 2 quadrillion yen ($14.5 trillion) financial assets in cash and deposits, Bank of Japan data show.
The tax panel of Japan’s ruling Liberal Democratic Party makes decisions near the end of the year after reviewing proposals from government offices

When Peter Thiel said "AI is communist," it was true that AI benefited from extreme centralization and access to insanely expensive resources.

BUT after 1 week of explosive Stable Diffusion open-source innovation, it's clear centralized AI is no longer the only game in town.

Why self-supervised learning is a medical AI game-changer

Self-supervised learning has been a fast-rising trend in artificial intelligence (AI) over the past couple of years, as researchers seek to take advantage of large-scale unannotated data to develop better machine learning models. 

In 2020, Yann Lecun, Meta’s chief AI scientist, said supervised learning, which entails training an AI model on a labeled data set, would play a diminishing role as self-supervised learning came into wider use. 

“Most of what we learn as humans and most of what animals learn is in a self-supervised mode, not a reinforcement mode,” he told a virtual session audience during the International Conference on Learning Representation (ICLR) 2020. And in a 2021 Meta blog post, LeCun explained that self-supervised learning “obtains supervisory signals from the data itself, often leveraging the underlying structure in the data.”

Because of that, it can make use of a “variety of supervisory signals across co-occurring modalities (e.g., video and audio) and across large datasets — all without relying on labels.” 

Growing use of self-supervised learning in medicine

Those advantages have led to the notable growing use of self-supervised learning in healthcare and medicine, thanks to the vast amount of unstructured data available in that industry – including electronic health records and datasets of medical images, bioelectrical signals, and sequences and structures of genes and proteins. Previously, the development of medical applications of machine learning had required manual annotation of data, often by medical experts. 

This was a bottleneck to progress, said Pranav Rajpurkar, assistant professor of biomedical informatics at Harvard Medical School. Rajpurkar leads a research lab focused on deep learning for label-efficient medical image interpretation, clinician-AI collaboration design, and open benchmark curation. 

Highlighting self-supervised learning

In a recent review paper in Nature Biomedical Engineering, Rajpurkar, along with cardiologist, scientist and author Eric Topol and student researcher Rayan Krishnan, highlighted self-supervised methods and models used in medicine and healthcare, as well as and promising applications of self-supervised learning for the development of models leveraging multimodal datasets, and the challenges in collecting unbiased data for their training. 

The paper, Rajpurkar said, was aimed at “communicating the opportunities and challenges that underlie this the shift in paradigm we’re going to see over the upcoming years in many applications of AI, most certainly including medicine.” 

With self-supervised learning, Rajpurkar explained that he, “… can learn about a certain data source, whether that’s a medical image or signal, by using unlabeled data. That allows me a great starting point to do any task I care about within medicine and beyond without actually collecting large labeled datasets.”

Self-supervised learning across modalities

For example, another soon-to-be-published paper showed that even with zero-annotated examples of diseases on chest X-rays, Rajpurkar’s team was actually able to detect diseases on chest X-rays and classify them nearly at the level of radiologists across a variety of pathologies.  

Momentum to apply methods

These possibilities have created momentum toward developing and applying self-supervised learning methods in medicine and healthcare, and likely for other industries that also have the ability to collect data at scale, said Rajpurkar, especially those industries that don’t have the sensitivity associated with medical data. 

Synthesizing theories of human language with Bayesian program induction

A fantastic piece of work by Kevin Ellis.

70 datasets from 58 diverse languages, synthesizing human-interpretable models for core aspects of each language’s morpho-phonology

"Binance looks to bolster lead position in crypto market with free trading. Binance has a 55% trading market share in cash trading, according to CryptoCompare. Its nearest rivals, Coinbase and FTX, have shares of less than 10%."

SEC and CFTC want help building new crypto reporting regime for hedge funds — link

Moore's L2 law: every year Ethereum's L2's 10x their throughput

Qualcomm just announced a multiyear partnership with Meta to develop custom VR chips for future Quest headsets.

Qualcomm CEO Cristiano Amon and Meta CEO Mark Zuckerberg announced the news during a press conference at Berlin's IFA tech show.

Qualcomm's already made the chips in Meta's VR headsets going back to the Oculus Go, and announced a partnership with Microsoft to make next-gen chips for its future AR smart glasses January's CES in Las Vegas.


The Qualcomm partnership promises "multiple generations of premium devices and experiences powered by custom VR platforms in the years to come," according to a statement from Qualcomm. 

In a comment from Mark Zuckerberg provided by Qualcomm, the partnership will bring "customized virtual reality chipsets -- powered by Snapdragon XR platforms and technology -- for our future roadmap of Quest products." 

When asked for further comment on the partnership via email, a Qualcomm representative clarified that the new customized chips will be powered by Qualcomm's Snapdragon XR platform, but won't use Snapdragon Spaces, an AR software toolkit that Qualcomm's building to bridge phones and AR glasses. Instead, these future products will use Meta's Presence Platform for VR.

According to Qualcomm, the chips could show up in other, non-Meta devices, too. "The VR chipsets that are being designed as part of this collaboration are not exclusive to Meta, but we're excited to work together on a deeper engineering level that's a first for both companies. It's this type of collaboration that's foundational for the metaverse and we can't wait to show you what we'll build together."
Although Meta is working on future AR glasses of its own, Qualcomm hasn't confirmed any involvement in that part of Meta's metaverse roadmap: "At this time, we are strictly only talking about VR."

Coinbase Venture’s POV on the Web3 Developer Stack.

Web3 developer stack

As Web3 gains momentum, there’s an opportunity to invest in more robust developer tools and infrastructure that will unlock the next stage of Web3 innovation and adoption.

This landscape is a framework for thinking about the Web3 developer stack as it continues to evolve.

The Web3 Developer Stack as a nascent ecosystem of promising companies that empower developers with infrastructure and tooling to build on top of crypto network primitives.

Today, Web3 developers represent <1% of the global software developer community (~31.1M global software devs as per SlashDataHQ) given a fragmented developer experience and skepticism on blockchain's long-term viability.

Electric Capital compiled a report which shows that the number of monthly active Web3 developers reached all-time highs in 2021 (*Note: the report doesn’t count work on proprietary projects).

Compared to Web2, building in Web3 requires specialized expertise to interface with complex infrastructure and commonly involves redundant processes.

That said, the tooling that will help onboard the next 1M+ web3 developers is rapidly improving.

Coinbase segmented the Web3 developer stack into the following four layers:

-Protocols
-Infrastructure
-Developer Tools
-Application Enablement

Protocols are the underlying main blockchain architecture and settlement layer

-Layer 1
-Layer 2
-Bridges & Swaps

The infrastructure layer consists of foundational building blocks for creating Web3 apps on top of crypto network primitives

-Node Providers
-Wallet/Key Management
-Identity
-Decentralized Compute
-Decentralized Storage
-Oracles
-Interoperability Protocols

Developer Tools consist of platforms and services that are highly reliable at doing specific tasks across the Web3 developer journey

-Frameworks & IDEs
-Low-code / No-code
-Index & Query
-Test, Simulate & Monitor
-Secure & Audit
-Messaging
-Analytics

The application enablement layer consists of point solutions that provide user-facing applications access to Web3 use case primitives

(i.e, dev tooling for NFTs, DeFi, DAOs, Metaverse/GameFi).

While the framework and layers we highlighted will likely remain unchanged, we continue to see new developer tooling primitives emerge and expect the entire stack to evolve dramatically in the coming years