Reading Digest, July #17
Hey there, my fantastic readers! I hope you’re ready for another thrilling edition of my daily reading digest. If you’re new here, buckle up and get ready for a wild ride through the captivating world of online content. And if you’re a regular, thank you for your continued support — it means the world to me!
Today’s digest is a true smorgasbord of fascinating topics, ranging from the potential impact of AI missteps on global peace and security to the unveiling of the world’s first meltdown-proof nuclear reactor in China. We’ll explore the latest developments in AI, including the release of Meta’s biggest ‘open’ AI model yet, the arrival of the first GPT-4-class AI model anyone can download, and how AI is being used to read slide decks.
But that’s not all — we’ve got some juicy pieces on the world of tech and business, including how Airbnb smoothly upgrades React, the economic mind of JD Vance, and how 77% of employees report AI has increased workloads and hampered productivity. We’ll also take a closer look at the confusion among officials during the deadly attempt on Trump, as revealed by rooftop bodycam video.
For the curious minds out there, we’ll dive into the world of marketing and advertising, exploring the debate between driving paid traffic to a landing page vs. a product page and the five planning pitfalls GTM teams should avoid. We’ll also explore the latest research in AI, including fast grammar-structured generation, explicit bidirectional interaction with face geometry for head pose estimation, and interpretable survival analysis via deep ReLU networks.
But that’s just the tip of the iceberg, my friends. From the art inspired by democracy and the despots who overthrew it to the hidden limits of superintelligence and why it might never happen, this digest has something for everyone. We’ll even take a closer look at how the FTC is launching a probe into how companies use data to tailor what each customer pays and the myth of Macintosh and Xerox PARC.
So, grab your favorite beverage, get comfortable, and join me on this thrilling journey through the world of online content. I can’t wait to hear your thoughts and reactions in the comments below!
Happy reading, my incredible friends!
The Demon Slayers, by Sam Kestenbaum
The article explores the rise of the “Demon Slayers” movement, a group of evangelical Christian pastors and ministers who have gained popularity through their focus on exorcism and deliverance from demonic spirits. It examines the media-savvy approach of the movement, led by pastor Greg Locke, and their efforts to stage large-scale exorcism events and produce feature-length films on the topic. The article delves into the historical and cultural context that has contributed to the growing interest in the supernatural and demonic among some evangelical Christians, as well as the tensions and challenges faced by the Demon Slayers as they seek to maintain their popularity and expand their media empire.
AI Missteps Could Unravel Global Peace and Security
The article discusses how civilian advances in artificial intelligence (AI) could have serious consequences for international peace and security, and how AI practitioners can play critical roles in mitigating these risks. It highlights the need for AI education to include courses on the societal impact of technology, responsible innovation, AI ethics, and governance.
Hong Kong major sanction evasion route, alleges report
The article discusses how Hong Kong’s government and local businesses are undermining international sanctions by facilitating the transfer of restricted and sensitive technology to sanctioned regimes like Russia, Iran, and North Korea.
CrowdStrike was the best ad campaign for the Mac…and cost Apple nothing
The article discusses how the recent CrowdStrike outage, which caused significant disruptions globally, inadvertently served as a great advertisement for Apple’s Mac computers. It highlights how Macs were unaffected by the outage, leading to memes and humor around Mac users remaining unperturbed while the rest of the world faced chaos. The article also suggests that the CrowdStrike incident could sway some average consumers towards purchasing Macs over PCs, especially given Microsoft’s statement about the security limitations of its operating system compared to Apple’s.
TLDR: 1 year of building with LLMs — D-Squared
The article discusses the key lessons learned by six practitioners who have been building with large language models (LLMs) for a year. It covers tactical, operational, and strategic lessons, with a focus on the tactical aspects.
AI companies promised to self-regulate one year ago. What’s changed?
The article discusses the progress made by seven leading AI companies (Amazon, Anthropic, Google, Inflection, Meta, Microsoft, and OpenAI) on the voluntary commitments they made with the White House a year ago to develop AI in a safe and trustworthy way.
Why The Llama 3.1 Announcement Is Huge — Tim Kellogg
The article discusses the announcement of Meta’s new large language model (LLM), Llama 3.1 405B, and the accompanying letter by Mark Zuckerberg on the benefits of open-source AI for developers, Meta, and the world. The article highlights four key reasons why this is a significant moment: data sovereignty, cost savings, independence, and customizability.
The first GPT-4-class AI model anyone can download has arrived: Llama 405B
The article discusses the release of a new AI language model called Llama 3.1 405B by Meta (Facebook), which is potentially the first openly available large language model (LLM) that rivals the capabilities of top AI models like GPT-4 and Claude 3.5 Sonnet. The article explores the implications of this release, including the debate around the use of the term “open source” and the potential impact on the AI industry.
Meta releases its biggest ‘open’ AI model yet | TechCrunch
The article discusses Meta’s latest open-source AI model, Llama 3.1 405B, which is the company’s largest model to date with 405 billion parameters. The article covers the key details about the model, including its capabilities, training data, and licensing, as well as Meta’s broader strategy around generative AI.
Chat BCG: Can AI Read Your Slide Deck?
The article evaluates the accuracy of two large language models, GPT-4o and Gemini Flash-1.5, in reading and interpreting data from labeled and unlabeled charts. It aims to assess whether these advanced multimodal models can perform well on specific “reading and estimation” tasks, particularly in the context of visual charts in business decks.
How Airbnb Smoothly Upgrades React
The article discusses Airbnb’s recent upgrade of its frontend from React 16 to React 18, the current major version of React. It describes the challenges and solutions involved in this upgrade, including:
- Creating the React Upgrade System to enable progressive rollout and testing of the upgrade across Airbnb’s web monorepo
- Using module aliasing and environment targeting to handle the two different React versions
- Leveraging Airbnb’s comprehensive test suite to ensure the safety of the upgrade
- Controlling the rollout of the upgrade across different product surfaces
The article also shares lessons learned and plans for future upgrades using the React Upgrade System.
U.S. would defeat China in a war over Taiwan, Joints Chiefs head says
The article discusses the U.S. military’s confidence in defeating China in a potential conflict over Taiwan, as stated by the Chairman of the Joint Chiefs of Staff. It also covers the evolving nature of warfare and the need for a mix of traditional military force and asymmetric warfare capabilities.
Is X Throttling Kamala Harris’ Account? Key Democrat Claims ‘Time-Sensitive Censorship Issue’
The article discusses Rep. Jerry Nadler’s call for the House Judiciary Committee to investigate Elon Musk’s X (formerly Twitter) regarding the apparent censorship of Vice President Kamala Harris’ campaign account. Nadler claims this is a “serious and time-sensitive censorship issue” as Harris is at the start of her presidential campaign.
Freedom of City award winner criticises disability access
The article discusses the case of Anna Landre, a wheelchair-using activist, researcher and consultant, who was awarded the Freedom of the City of London for her contributions to computer science and disability campaigning, but was unable to collect the award on stage due to the lack of an accessible ramp. The article highlights the irony of being granted the Freedom of the City but not the freedom of the room, and the systemic issue of inaccessibility that disabled people often face.
The economic mind of JD Vance
The article discusses the selection of Sen. JD Vance, R-Ohio, as the Republican vice presidential candidate and his economic policy views, which depart from traditional Republican conservatism.
77% Of Employees Report AI Has Increased Workloads And Hampered Productivity, Study Finds
The article discusses the rise of AI and its impact on the workforce, highlighting the disconnect between the high expectations of managers and the actual experiences of employees using AI. It explores how AI is increasing workloads, hampering productivity, and contributing to employee burnout, despite the optimistic expectations of AI’s benefits.
Rooftop bodycam video shows confusion among officials, rifle shooter used in deadly attempt on Trump
The article discusses the bodycam footage of the shooting incident that targeted former President Trump during a campaign rally in Butler, Pennsylvania. The footage shows the body of the shooter, Thomas Matthew Crooks, who was shot dead by snipers on the rooftop of a building. The article also mentions the resignation of Secret Service Director Kimberly Cheatle amid criticism over the security lapses that allowed Crooks to get a clear shot at Trump.
Netanya-Who? Washington gives Israeli leader the cold shoulder
The article discusses Israeli Prime Minister Benjamin Netanyahu’s visit to Washington, D.C., which has received relatively little attention on Capitol Hill compared to the busy news cycle.
Kamala Harris holds first campaign rally as Biden prepares to address nation: Live
The article discusses the 2024 US presidential election, focusing on Vice President Kamala Harris’ campaign launch and her attacks on former President Donald Trump, as well as the potential impact of President Biden’s decision to not run for re-election.
We’re building nuclear spaceships again — this time for real
The article discusses the development of nuclear thermal rockets (NTRs) for use in space exploration and intercontinental ballistic missiles (ICBMs), focusing on the Phoebus 2A reactor and the challenges of creating a flying nuclear reactor.
World’s first meltdown-proof nuclear reactor unveiled in China
The article discusses the world’s first meltdown-proof nuclear reactor developed by researchers at Tsinghua University in China. The reactor, called the High-Temperature Gas-Cooled Reactor Pebble-Bed Module (HTR-PM), has a capacity of 105 MW and has been in development since 2016, with commercial operations beginning in December 2023.
5 Planning Pitfalls GTM Teams Should Avoid
The article discusses the importance of adopting a comprehensive and integrated Go-To-Market (GTM) strategy, rather than relying solely on numerical analysis and calculations. It highlights five common biases that can hinder the GTM planning process, including “Mathism”, “Incrementalism”, “Departmentalism”, “Growthism”, and “Idealism”. The article introduces the GTM Operating System, an 8-pillar framework developed by GTM Partners to provide organizations with clarity and alignment in their strategic planning and execution process. It also emphasizes the importance of Clarity, Alignment, and Teamwork (CAT) in overcoming internal misalignment, which is identified as the biggest roadblock to reaching business goals.
Driving Paid Traffic To A Landing Page vs. Product Page: What Is Better?
The article discusses the differences between product detail pages (PDPs) and landing pages, and when to use each in the customer journey. It highlights the benefits of using landing pages for marketing campaigns and provides best practices for creating effective landing pages.
How fast can grammar-structured generation be?
The article discusses the use of structured generation to improve the results obtained from large language models (LLMs). It focuses on ensuring that the costs of structured generation are low enough for widespread use, while also maintaining flexibility in the structure being imposed.
Three Archetypes of AI Application Startups
The article discusses the three main archetypes of AI applications that have emerged in the technology industry over the past year: AI copilots, AI colleagues, and AI-native services. It provides examples and explains the key characteristics of each archetype.
Using LLMs for Evaluation
The article discusses the use of large language models (LLMs) as evaluators for other LLMs, a technique commonly referred to as “LLM-as-a-Judge”. It covers the motivation behind this approach, the different setups and prompting strategies used, the strengths and limitations of LLM-based evaluations, and various analyses and insights from related research papers.
6DoF Head Pose Estimation through Explicit Bidirectional Interaction with Face Geometry
This study addresses the challenge of estimating head translations within the context of six-degrees-of-freedom (6DoF) head pose estimation, placing emphasis on this aspect over the more commonly studied head rotations. The authors propose a novel approach called the head Translation, Rotation, and face Geometry network (TRG), which stands out for its explicit bidirectional interaction structure that leverages the complementary relationship between face geometry and head translation. The authors also develop a strategy for estimating bounding box correction parameters and a technique for aligning landmarks to the image, which demonstrate superior performance in 6DoF head pose estimation tasks. Extensive experiments on ARKitFace and BIWI datasets confirm that the proposed method outperforms current state-of-the-art techniques.
SurvReLU: Inherently Interpretable Survival Analysis via Deep ReLU Networks
The paper proposes a deep rectified linear unit (ReLU) network called SurvReLU that bridges the gap between deep survival models and traditional tree-based survival models. SurvReLU can achieve interpretability like a tree-based model while maintaining the representational power of a neural network. The key contributions are:
- Establishing an explicit connection between ReLU networks and tree-based survival models by showing that ReLU networks can partition the input space into locally homogeneous regions like a tree.
- Introducing a statistically-driven method to dynamically optimize the topology of the ReLU network for survival analysis, enabling automatic pruning of the resulting tree structure.
- Demonstrating that SurvReLU can be optimized end-to-end with flexible loss functions, including both continuous-time and discrete-time survival losses.
Experiments on both simulated and real-world datasets show that SurvReLU achieves competitive performance compared to previous deep and tree-based survival models, while providing better interpretability.
Bucketed Ranking-based Losses for Efficient Training of Object Detectors
The paper focuses on improving the efficiency of ranking-based loss functions, such as Average Precision (AP) Loss and Rank & Sort (RS) Loss, for training object detectors. These loss functions outperform widely used score-based losses, but have high time and space complexities due to the need for pairwise comparisons among positive and negative predictions.
AssistantBench: Can Web Agents Solve Realistic and Time-Consuming Tasks?
The article introduces AssistantBench, a new benchmark for evaluating the ability of web agents to solve realistic and time-consuming tasks. The benchmark contains diverse tasks covering various scenarios and domains, requiring agents to browse the web, identify relevant information, and synthesize outputs. The article also proposes a new web agent called SeePlanAct (SPA) that outperforms existing state-of-the-art web agents on the benchmark. The results show that AssistantBench is challenging for current systems, with no model reaching an accuracy of more than 25%. The article provides a detailed analysis of the errors made by different models, highlighting the limitations of closed-book models, retrieval-augmented models, and web agents in solving these types of tasks.
An Update on our Make Designs Feature | Figma Blog
The article discusses the launch and temporary rollback of Figma’s new AI-powered design feature called “Make Designs”, which allows users to generate UI design drafts based on design system components and a simple prompt. The article explains the technical details of how the feature works, the issues that were encountered, and Figma’s plans to improve the feature going forward.
Coups, crackdowns and chaos: art inspired by democracy — and the despots who overthrew it
The article discusses a new art exhibition in Athens, Greece that explores the theme of democracy through the lens of the country’s history with military dictatorships in the 20th century. The exhibition examines how artists in Greece, Portugal, and Spain responded to and depicted the suppression of civil liberties and the eventual restoration of democracy in their respective countries.
What My Clients Told Me About Their Wives
The article explores the shifting allegiances and perspectives of a sex worker as she navigates her relationships with her clients and their unseen wives.
Mark Zuckerberg — Open Source AI is the Path Forward In the early… | Facebook
The article discusses the benefits of open source AI and why Meta is committed to releasing Llama, an open source AI model, as the industry standard.
Vibes Are a Legitimate Economic Indicator
The article discusses Kyla Scanlon, a 27-year-old economics commentator who has gained popularity for producing smart and accessible explainers on economic topics. It focuses on her recently published book “In This Economy?: How Money and Markets Really Work”, which aims to help the layperson understand the underlying forces driving the current economic landscape.
YouTubification is coming for the media
The article discusses the growing influence of YouTube as a major podcast platform, and how this “YouTubification” of media is impacting the podcast industry. It explores how the success and approach of YouTube creators like MrBeast and Steven Bartlett are shaping the future of podcasting, and raises concerns about the implications for traditional media and journalism.
Disruption Comes to Google
The article discusses the dynamic nature of the tech industry and the challenges faced by established companies like Google in the face of disruptive technologies like LLM-based answer engines. It examines Google’s dominant search business model and the potential threats posed by the rise of tools like ChatGPT, Perplexity, and Microsoft’s Copilot.
People Don’t Read Online — They Scan. This Is How to Write for Them
The article discusses how people read and scan online content, and provides tips for writers to adapt their content to these reading patterns.
We Need To Rethink The Myth Of Macintosh And Xerox PARC
The article discusses the history of Xerox’s research and development efforts, particularly the creation of the Palo Alto Research Center (PARC) and its role in developing groundbreaking technologies that later influenced the personal computing industry. It challenges the common narrative of Xerox’s “blunder” in failing to capitalize on these innovations, arguing that PARC was a strategic and successful investment that helped save the company.
Power Play
The article discusses the challenges and opportunities surrounding the increasing role of batteries in our daily lives as cities electrify. It explores various battery ownership models, the supply chain issues, and the need to make battery technology more visible and connected to people’s daily lives.
Your OKRs Aren’t OKRs
The article discusses the common challenges teams face when setting Objectives and Key Results (OKRs) and provides guidance on how to effectively implement OKRs.
Work From Office
The article discusses the polarizing nature of remote work, its benefits and drawbacks, and its broader implications for the economy and society. It explores the shift from the traditional office-based work model to remote work, the impact on different socioeconomic groups, and the potential opportunities and challenges it presents.
The Hidden Limits of SuperIntelligence & Why It Might Never Happen
The article discusses the limitations and challenges of achieving true artificial general intelligence (AGI) and conscious machines. It presents arguments from various researchers and philosophers against the notion that AI can become conscious and superintelligent like humans.
FTC launches probe into how companies use data to tailor what each customer pays
The article discusses the Federal Trade Commission’s (FTC) inquiry into how companies use consumer data to set individualized pricing for products and services, a practice known as “surveillance pricing.” The FTC has ordered eight companies, including major firms like Mastercard, JPMorgan Chase, and Accenture, to explain how they gather and use consumer data to determine pricing, which could undermine consumer privacy and marketplace competition.
Our website: https://aili.app
Follow us on X (Twitter): https://x.com/aili_app
Join our discord channel: https://discord.gg/CQtysdQfDM
