Search the archive

PRAGMATIC DRAG AND DROP

A performance-focused drag-and-drop library that can be used to power experiences on any frontend stack.

EVERYTHING I KNOW ABOUT THE XZ BACKDOOR

A sophisticated security attack compromised the xz compression library, a critical dependency of OpenSSH. The backdoor, introduced by a seemingly helpful new maintainer, allowed attackers to gain unauthorized access to affected servers. The backdoor was discovered by Andres Freund, a software engineer at Microsoft, when he found that SSH logins were using a lot of CPU power compared to normal.

CAN DEMIS HASSABIS SAVE GOOGLE?

Demis Hassabis, the founder of DeepMind, is now tasked with keeping Google in the lead of the AI race. To succeed, Hassabis must translate DeepMind's research into tangible product improvements and convince Google to embrace potentially disruptive changes. If he is successful, DeepMind's focus on reinforcement learning could position the company well as chatbots evolve into autonomous agents that can take action in the real world.

HOW GITHUB REPLACED SOURCEFORGE AS THE DOMINANT CODE HOSTING PLATFORM

Before GitHub, sites like SourceForge and Google Code existed for hosting open-source software, but they lacked the social features and focus on Git that makes GitHub so useful. GitHub's focus on making collaboration seamless and social catapulted it to its dominant position. However, GitHub's market share might decline in the future as the market for code development tools matures and engineers demand more specialized tools for closed-source development.

GUESS MY RGB

A fun game using RGB sliders to try to match the background color of the website.

CODE SMELL 244 - INCOMPLETE ERROR INFORMATION

Error messages should contain information needed to identify and fix the problem.

TOP 9 WEBSITE PERFORMANCE METRICS YOU CANNOT IGNORE

To optimize website performance, focus on minimizing load time, time to first byte, request count, DOMContentLoaded, time to above-the-fold load, first contentful paint, page size, round trip time, and render-blocking resources.

I TESTED CLAUDE 3 OPUS ON REAL GITHUB ISSUES

Claude 3 Opus is able to generate a partially correct code patch in minutes.

🆓 Try Veracode DAST free for 14 days

SAM BANKMAN-FRIED SENTENCED TO 25 YEARS FOR FTX FRAUD

FTX co-founder Sam Bankman-Fried has been sentenced to a total of 300 months in prison for seven counts of conspiracy and fraud charges. He has also been ordered to forfeit more than $11 billion, including property. The sentence is shorter than the 40 to 50 years requested by the prosecutors, but longer than the six and a half years his lawyers asked for and far less than the maximum of 110 years. Bankman-Fried has been residing in custody since August 2023, where he has been subjected to multiple extortion attempts.

FACEBOOK LET NETFLIX SEE USER DMS, QUIT STREAMING TO KEEP NETFLIX HAPPY

Facebook Watch, announced in 2017, never really took off. Recently revealed documents show the company's plans for the platform and how it was confident in its ability to compete with Netflix. Mark Zuckerberg abruptly announced a change in strategy in May 2018, suddenly slashing the budget for the platform. He claimed the budget changes were based on his knowledge of the company's strategy and financial outlook. Months later, Facebook and Netflix announced a new data-sharing agreement that involved Netflix reaching $200 million in advertising spend by 2019. The companies also had a data-sharing agreement that may have allowed Netflix to access Facebook users' private messages.

HOW WE’LL REACH A 1 TRILLION TRANSISTOR GPU

Recent achievements in AI have been due to three factors: innovation in efficient machine-learning algorithms, the availability of massive amounts of data to train neural networks, and progress in energy-efficient computing through the advancement of semiconductor technology. The industry will need much more powerful devices to continue at its current pace. For the past 50 years, the industry has focused on shrinking the transistor, but now, there are many more possibilities. Future GPUs will not be restricted to the sizes and form factors of current devices - integrated AI systems will be composed of as many energy-efficient transistors as is practical.

MEMORIES ARE MADE BY BREAKING DNA — AND FIXING IT

A study in mice shows that memories are formed from breaks in DNA. These breaks occur due to a rush of electrical activity and are repaired using an inflammatory response. The findings suggest a mechanism for faulty memory in people with neurodegenerative diseases.

VALKEY: THE OPEN SOURCE ALTERNATIVE TO REDIS BACKED BY AWS, GOOGLE, ORACLE

Valkey is a high-performance data structure server that primarily serves key/value workloads. It is an open-source fork of the popular Redis data store. The project started as Redis Labs, the company behind the original Redis codebase, changed Redis to more restrictive licensing. The Valkey fork has assembled an impressive roster of maintainers and contributors and has wide-ranging industry support.

SPICE

Spice is a runtime that makes querying data by SQL across one or more data sources simple and fast. It provides developers with a unified SQL query interface that can locally materialize, accelerate, and query data tables sourced from any database, data warehouse, or data lake. Spice makes it easy to build data-driven and data-intensive applications. It leverages industry-leading technologies like Apache DataFusion, Apache Arrow, Apache Arrow Flight, and DuckDB.

A FEW THOUGHTS ON THE DOJ’S ANTITRUST CASE AGAINST APPLE

None of the things mentioned in the DOJ's case against Apple should be a surprise to anyone who has been following in the Apple space for years. People have been hoping the company would change or tweak its ways in many of these areas. While Apple's actions may be disagreeable, they may not be illegal. The DOJ is fighting an uphill battle with its case, but it will hopefully apply some pressure on Apple to loosen its grip on some things.

GOOGLE SAYS APPLE IS BRINGING RCS TO THE IPHONE IN ‘FALL OF 2024’

The Android website has added a new landing page for Google Messages which notes that RCS is coming to the iPhone in the fall of 2024. It is not clear whether RCS will arrive with the initial release of iOS 18 or in an update in the last few months of the year. The updated landing page also notes that animated emoji reactions are coming soon to RCS. Screenshots of the landing page are available in the article.

PYPI HALTED NEW USERS AND PROJECTS WHILE IT FENDED OFF SUPPLY-CHAIN ATTACK

New project creation was halted on PyPI for around 10 hours while the site dealt with an onslaught of package uploads that executed malicious code on any device that installed them.

4.5 MILLION TIMES FASTER INTERNET? ASTON UNIVERSITY MAKES IT POSSIBLE

Researchers at Aston University achieved an internet data transmission speed of 301 terabits per second by utilizing wavelength bands that have not been previously used in fiber-optic systems.

MEET HUME’S EMPATHIC VOICE INTERFACE

Hume can understand peoples' tone of voice and use that to inform its generated voice language.

GOOGLE NOW INDEXING BITCOIN’S BLOCKCHAIN TO SHOW WALLET BALANCES

Google Search can now display balances for wallets on the Bitcoin, Arbitrum, Avalanche, Fantom, Optimism, and Polygon networks.

MOJO IS OPEN SOURCE!

Modular has shared its full commit history and open sourced Mojo, a programming language that combines Python's syntax and ecosystem with systems programming and metaprogramming features.

MY LIST OF CHALLENGING SOFTWARE PROJECTS SOME PROGRAMMERS SHOULD TRY

A list of project ideas for those interested in exploring new areas of knowledge - most of the ideas may not be relevant to today's job market.

AI21 RELEASES JAMBA LANGUAGE MODEL

Mamba is a style of model that is supposed to beat Transformers in efficiency while matching performance. Jamba is a novel variant that includes MoE layers. It can run at 1.6k tokens per second with a context length of 128k tokens. It achieves 67% on the MMLU benchmark. The weights are available.

HUME INTRODUCES EMPATHIC VOICE INTERFACE

Hume has released the Empathic Voice Interface (EVI), the first conversational AI with emotional intelligence.

GOOGLE STARTS TESTING AI OVERVIEWS FROM SGE IN MAIN GOOGLE SEARCH INTERFACE

Google is piloting AI-generated overviews in search results for complex queries with a subset of U.S. users without requiring opt-in to its Search Generative Experience. This feedback-driven test aims to enhance user experience and may impact site traffic dynamics due to changes in ad placements and organic search visibility.

1 BIT, POST TRAINING, QUANTIZATION

1 Bit language models are exciting. This work shows how to quantize the Linear layers of a language model without sacrificing performance. This can result in a 70B model running on consumer GPUs.

LONG FORM FACTUALITY IN LARGE LANGUAGE MODELS

You can use language models to generate factual information. Google has released a dataset and benchmarks to show how well each model performs. This report gives insight into how to improve a model’s factuality and shows language models winning against human annotators in a majority of cases.

IMPROVING MODEL ADAPTATION

CoDA is a new approach to Unsupervised Domain Adaptation (UDA). It helps AI models better adapt to unlabeled, challenging environments by learning from differences at both the scene and image levels.

real-time data available for retrieval. Watch the demo

USING AI TO IDENTIFY SOURCES OF CONFUSION IN YOUR CONTENT

This approach pinpoints the exact content that triggers user questions online, from lectures to news articles. This method, known as backtracing, aims to help content creators refine their work by identifying and understanding the causes of confusion, curiosity, or emotional reactions.

HIGH-RESOLUTION LARGE LANGUAGE-VISION ASSISTANT

Researchers have developed a new method, Mixture-of-Resolution Adaptation (MRA), which improves AI's ability to recognize details in images.

3D MEDICAL IMAGING DATASET

CT-RATE is a dataset that marries 3D medical imaging with textual reports. CT-CLIP is a versatile AI framework optimized for these images.

EXPONENTIAL GROWTH AND AI ESCAPE VELOCITY

AI Futurist Ray Kurzweil invented the law of accelerating returns, which states that the rate of progress is increasing at an exponential rate over time. In a recent discussion, Kurzweil explored various topics including prospects that are only getting better, the direction in which the AI economy is headed, human relationships with AIs, longevity escape velocity, and much more.

PLENTIFUL, HIGH-PAYING JOBS IN THE AGE OF AI

AI experts are exploring automating human tasks, leading to concerns about job loss and wage declines. However, the view that AI advancement will inevitably result in human job obsolescence may not be correct. Humans may retain employment in an AI-dominated future due to constraints like computing power and opportunity costs, although this scenario is not guaranteed.

COFRAME LIVING IMAGES (PRODUCT LAUNCH)

Coframe’s generative A/B testing tool allows you to optimize images on your website in minutes with generative AI.

META IS ADDING AI TO ITS RAY-BAN SMART GLASSES

Meta is introducing AI features to its Ray-Ban smart glasses, including translation, and the identification of objects, animals, and monuments, with activation through voice commands.

GOOGLE BRINGING GEMINI NANO TO PIXEL 8

The Pixel 8 will receive Gemini Nano in a developer preview to enhance features like Summarize in Recorder and Gboard Smart Reply.

SAM BANKMAN-FRIED SENTENCED TO 25 YEARS IN PRISON FOR ORCHESTRATING FTX FRAUD

FTX co-founder Sam Bankman-Fried has been sentenced to 25 years in prison on seven fraud and conspiracy charges related to the collapse of the FTX exchange. He was also fined $11 billion, around the same amount that was lost by the exchange and the Alameda trading firm. The judge recommended he serve his sentence in a medium or low-security federal prison near the Bay Area to facilitate family visitation.

>$1B IN US TREASURY NOTES TOKENIZED ON PUBLIC BLOCKCHAINS

The market for tokenized U.S. government securities has surged since January 2023, surpassing $1 billion. The data reflects a significant uptick in demand and adoption of blockchain technology for traditional financial assets, with tokenized Treasuries offering a modern, blockchain-based approach to trading government bonds.

BLOBSTREAM LIVE ON BASE MAINNET

Blobstream is now live on the Base Mainnet, marking an evolution in onchain development. Developers can now create high-throughput blockspaces, comparable to smart contracts, using Celestia's modular data availability (DA) layer. This innovation removes the need for alternative Layer 1s or centralized DA solutions, allowing for seamless deployment within the same ecosystem. Developed by Succinct Labs, Blobstream acts as a data attestation bridge, ensuring rollup data integrity on Celestia through DA attestations and Merkle proofs and facilitating the scaling of Ethereum's rollup frameworks.

DEGEN LAUNCHING L3 ON BASE

Degen Chain is a pioneering Layer 3 (L3) platform designed for the /degen community. It is built upon Arbitrum Orbit with Base for settlement and AnyTrust for data availability. Degen Chain uses $DEGEN as its native gas token, making it one of the first community tokens to operate with its own L3 infrastructure. This innovation, offering ultra-low-cost transactions and new scalable solutions for the EVM ecosystem, represents a significant step in onchain internet evolution.

ETHEREUM HAS BLOBS. WHERE DO WE GO FROM HERE?

In this post, Vitalik Buterin discusses the success of blobs in bringing Ethereum from a zero-to-one problem to a one-to-N problem. Scaling Ethereum will have incremental progress and will see blobs get bigger and rollups more efficient at using them. Security standards for rollups need to become stricter as only Arbitrum has reached stage 1 of decentralization. While protocol development is decelerating, it is accelerating on the application layer. Buterin concludes by urging developers to build for the Ethereum of the 2020s.

ON WHAT IS DECENTRALIZED

This thread explores the current state of decentralization, emphasizing the contrast between permissionless blockchains like Ethereum and websites like OpenSea. Any website can choose whether or not to display an NFT, but they cannot modify the NFT itself. Most Layer 2s are still fairly centralized and controlled by a few parties. Decentralization will likely improve in the coming years.

ETHEREUM ‘BLOBSCRIPTIONS’ LIFT OFF AND IT’S TAKING BLOB FEES WITH IT

Blobs, the new storage type for Ethereum L2s introduced in the March 13 Dencun upgrade, have been overrun by the inscriptions mania that previously congested several other chains earlier this year. BlobScriptions caused blob gas fees to spike to around $18, though it has since fallen to around $1.20. The fees should return to normal levels as the true value of blobs becomes evident after this price discovery phase.

A BRIEF HISTORY OF MEMECOINS

Memecoins blend internet culture with digital currency. Tracing back to the early days of cryptocurrency, with each cycle introducing new iterations, they gained momentum through early Bitcoin alternatives, ICO crazes, DeFi summer, and meme stocks, evolving to include NFTs which brought visual representation to these tokens. Memecoins, such as Dogecoin and Shiba Inu, have often been regarded as cultural tokens, gaining value from their associated memes and community engagement. Despite their potential for high returns, they carry risks of volatility and regulatory uncertainty.

PRISMA FINANCE SUFFERS $12 MILLION EXPLOIT

The PRISMA token crashed 30% after the exploit - $110 million was withdrawn from the protocol following the news.

SCROLL AIRDROP GUIDE

This thread explains why the Scroll Layer 2 is worth farming and provides an overview of the activities to do and dApps to interact with to potentially become eligible for its eventual airdrop.