If you think “AI content” means spamming SEO articles, you're not keeping track. Major brands are using AI-driven content creation to deliver highly targeted and engaging customer experiences, effortlessly.

In this recorded conversation, Contentful and Omni Talk look at how 3 major brands are using AI tools for segmentation, automated content creation, localization, and improving brand consistency.

👉 Watch the video

👉 Start building your own AI-enabled experiences — and take the chaos out of managing and delivering personalized content — with Contentful.

Google's new app, Edge Gallery, allows users to download and use AI models on smartphones. The Android app can be downloaded and installed through a GitHub repository - an iOS version will be available soon. There are several models available, including Gemma 3 and 3n and Alibaba's Qwen 2.5 models. The models can be used without an internet connection once installed.

The Justice Department's antitrust case against Google revealed a heavily redacted document from OpenAI from late 2024 that detailed the company's internal strategy and its goal to create a de facto 'interface to the internet'. While much of the document was blacked out, it was clear how much OpenAI expects ChatGPT to revolutionize the internet. The company planned to spend the first half of 2025 building out ChatGPT as a super assistant and then shift in the second half to generating monetizable demand. One of the company's cornerstone strategies is to build out more data centers to keep up with ChatGPT's ballooning user base.

Waymo cracked a million total paid rides in late 2023. By the end of 2024, it reached five million, and it is on track to blow past 20 million full autonomous trips by the end of the year. The company's service is only available in a few cities right now, but it might be available everywhere sooner than expected. The competition in the industry is just getting started, with Tesla planning to reveal a robotaxi service in the coming weeks.

SpaceX plans to build the biggest structure in the world to house up to 1,000 Starship rockets per year. The gigabay is part of the company's plans to launch multiple rockets per day to reach and colonize Mars. Elon Musk estimates that SpaceX will need to launch up to 2,000 rockets to Mars in every two-year window to carry the cargo needed to set up a colony. He aims to send humans to Mars by 2029.

Datadog AIOps uses Generative AI and Machine Learning to reduce noise, surface key insights from your data, and automate issue responses. Use it to proactively detect issues across your tech stack, augment troubleshooting workflows with contextual insights, and resolve issues quickly with automated RCA. Read the solution brief

This is a report on Claude Code generated by Claude Opus 4 with the help of almost all of the major flagship models. Claude Code is an agentic coding tool with a few new novel components. It has a streaming architecture that handles real-time model responses, tool execution, and UI updates, safety systems that provide security without disrupting workflow, a tool design that bridges AI reasoning and system execution, and prompt engineering that reliably controls complex model behavior. The report covers the foundation of Claude Code's architecture, data structures and the information architecture, control flow and the orchestration engine, tools and the execution engine, and much more.

AWS strives to deliver reliable services that customers can trust completely, which requires maintaining the highest standards of security durability, integrity, and availability. Systems correctness serves as the cornerstone for achieving these priorities. This article looks at the formal methods used across AWS to deliver complex services with high confidence in correctness. It covers more traditional formal approaches such as theorem proving, deductive verification, and model checking, as well as more lightweight semi-formal approaches such as property-based testing, fuzzing, and runtime monitoring.

Millions of young graduates are entering into industries that have little use for their skills, view them as expensive and expendable, and are rapidly phasing out their jobs in favor of artificial intelligence. This emerging crisis for entry-level workers is being fueled in part by rapid advances in AI capabilities. Unemployment for recent college graduates has jumped to an unusually high 5.8% in recent months. While some executives are making a calculated bet on AI systems, there are many risks, for example, the expectation that entry-level jobs are short-lived may lead companies to underinvest in job training, mentorship, and other programs aimed at entry-level workers, leaving those workers unprepared for more senior roles later on.

US efforts to win the tech race against China have so far failed. While White House officials and their proxies want to ban the export of AI microchips and the tools to make them, some outside the administration say that such policies are backfiring by accelerating China's development of its own tech ecosystem. The debate comes down to whether it is better to have China dependent on US tech or if the national security risk of China having US tech is too great. China's well-organized, well-funded, and consistent push for self-sufficiency has lessened its dependence on other countries, and each year it produces a greater proportion of everything it needs.

We also write TLDR AI. Same format, all about LLMs, GenAI, and deep learning. Delivered every weekday. Subscribe for free.

Claude created an SVG for an icon when it couldn't find the icon within the library the developer had provided for it.

It's worth taking a day off once in a while to contemplate where you're going in life.

How an entrepreneur reached $100 per month from their personal projects after two years.

Agentic pair programmers are too fast to collaborate with humans.

Judge Amit Mehta hasn't made it clear which way he is leaning, but he expects to issue a ruling in August.

We tend to infer causality when we observe specific trends and patterns even when the causal relationship may not exist.

Spencer Rascoff co-founded Zillow. His new venture is Pacaso — a tech-enabled marketplace for luxury vacation homes that's already generated $110M+ in gross profits, facilitated $1.1bn+ transactions, and grown to over 2,000 homeowners.

And while Zillow went straight to IPO, Pacaso is offering retail investors a rare opportunity to buy in at $2.80/share. The company has launched a public fundraise to support further global expansion into markets like Europe and the Caribbean.

→ Backed by SoftBank, Greycroft, Maveron

→ Nasdaq ticker "PCSO" already reserved

→ $1.3tn market opportunity

Lock in the $2.80/share price by TOMORROW before it changes

Apple is expected to launch a gaming app alongside the next version of iOS this September. The app will apparently replace Apple's Game Center. It will feature editorial content and let users access games available on the App Store. The app will work similarly to the Xbox app for iPhone. The macOS version of the app will let users access games from outside the App Store.

OpenAI is exploring ways for users to sign into third-party apps using their ChatGPT accounts. ChatGPT currently has roughly 600 million monthly active users. Adding a 'Sign in with ChatGPT' feature could help OpenAI compete with other massive technology companies that help people with a range of online services. OpenAI appears to be interested in integrating the sign-in service with a broad array of companies.

SpaceX's Starship successfully reached orbit last night, but it later began spinning and made an uncontrolled re-entry into the Indian Ocean. It was the smoothest Starship test of the year - the previous two tests ended in explosions. Last night's launch was the first to use a flight-proven Super Heavy booster. The booster had launched and returned during Starship's seventh test flight.

The world's first regenerative cell therapy for treating spinal injuries has been granted approval for a registrational Phase I clinical trial. Chinese biotechnology company XellSmart's allogeneic induced pluripotent stem cells regenerative therapy aims to repair injuries and provide the foundation to regrow all the cells needed to return function to the damaged region. It doesn't require harvesting cells from a patient and is a 'one size fits all' treatment that will work with anyone who has a spinal cord injury. If it passes through the trial phases, the treatment will be easy to manufacture and scale for broad access.

A monorepo is a powerful tool for institutional consistency. They are used to enforce engineering culture and organization standards and build a culture of code sharing. Achieving a productive monorepo is a continuous process as organizations continue to scale and tools begin to break, but it can be worth it if organizations are willing to commit to it. This post looks at some of the tools needed to set up a monorepo and discusses implementation strategies for them to paint a clear picture of what it is like to walk the monorepo path.

Amazon Aurora DSQL is a serverless, distributed relational database service optimized for transactional workloads. The goal of its design was to break up the database into bite-sized chunks with clear interfaces and explicit contracts. Each component would do one thing and do it well but work together to offer all of the features users expect from a database. DSQL was written in Rust, which gave the team the control they needed to avoid tail latency in the core parts of the system, the flexibility to integrate with a C code base like Postgres, and the high-level productivity the team needed to stand up its control plane.

Shiny new technologies emerge every few years promising to make software developers obsolete. However, what actually happens isn't replacement, it's transformation. These technologies that promised to eliminate the need for technical expertise ended up creating entirely new specializations, often at higher salary points than before. The same pattern is playing out in AI-assisted development. The most valuable skill in software isn't writing code, it's architecting systems, and that's one skill AI isn't close to replacing.

Elon Musk had approached Apple with an offer in 2022 after hearing the company was planning to announce a satellite feature with the iPhone 14. His offer was for SpaceX to provide exclusive satellite connectivity to iPhones for 18 months if Apple would pay $5 billion upfront. Musk had also threatened to announce his own satellite feature that would work on iPhones if Apple couldn't come to terms with SpaceX. Apple rejected the offer and Musk made good on his threat - SpaceX announced a partnership with T-Mobile to provide satellite services for users in areas with no reception. The company's satellite efforts will likely remain very limited moving forward as it could trigger the US government to regulate it as a telecommunications carrier.

You've read through this newsletter. Now, take a minute to learn about how Pacaso is capturing the hottest sector in real-estate, and how you can become part of their success - if you act today.

LiveStore is a fully-featured, client-centric data layer with a reactive embedded SQLite database powered by real-time sync.

Cline approaches codebases the same way a senior engineer would - with curiosity, systemic exploration, and the right tools.

The FTC's case will be hard to win because it seeks to unwind mergers from years ago that regulators had approved at the time.

DuckLake is an open Lakehouse format that stores metadata in a catalog database and stores data in Parquet files.

Developers can now grant large language models access to any tool that they can represent as a Python function.

While users own the hardware, the software that's needed to run it is subject to a license agreement that can be revoked.

Apple's rumored Face-ID door lock and smart display are more than just products..

They signal a major push into the $158B smart home market, growing 23% annually.

As Apple doubles down, investors are hunting for the next breakout company and acquisition target, following in the footsteps of Google (acquired Nest, $3.2B) and Amazon (acquired Ring, $1.2B).

History shows: when Apple enters, they go big.

One startup is perfectly positioned to benefit.

With $12M in revenue, distribution in Bestbuy and HomeDepot, and 10 defensive patents, they are built for a breakout.

Shares are $1.90 until 05/30. Join 3000+ investors before the price changes.

Apple plans to introduce a design overhaul for all of its operating systems at WWDC. The updated interface, codenamed 'Solarium', will give the operating systems a more slick and modern appearance. It borrows a lot of elements from visionOS. The UI overhaul will be the biggest for iPhones since iOS 7 in 2013 and for Macs since macOS Big Sur in 2020.

SpaceX has been moving more employees into Starship roles in recent months. The space company has also been making big investments in its complex in Texas for the vehicle and has similar plans for Florida. It is behind schedule on a variant of Starship for a NASA mission to the moon currently set for 2027. SpaceX has been struggling this year, with back-to-back Starship flight tests that ended in explosions.

The number of diseases that the GLP-1 category of drugs seems to be effective in treating is expanding rapidly. The drug could be beneficial to a third to a majority of adults. Many of the proven and potential benefits of the drugs cascade from their effect on obesity, but some are independent of weight loss, possibly from the anti-inflammatory effects of the drugs.

TSMC is working with startup Avicena to produce microLED-based interconnects. The technology is designed to meet the high needs of communication among an increasing number of GPUs in a low-cost, energy-efficient way. AI clusters are facing unprecedented requirements - sooner or later, the copper wires that connect processors and memory within a single AI data rack will have to be replaced with optics. Avicena's approach is very unorthodox, but it is ideal for these short-distance applications.

Delve ran 4 placements across 2 TLDR newsletters and brought in 66 leads, with a meaningful chunk from enterprise. The result: $1M in attributed pipeline and a 52x ROI.

The case study breaks down their strategy, results, and example ads.

Read the Delve case study.

Nix was designed to solve the lack of true reproducibility in software development. While it has been around for two decades, Nix has largely flown under the radar of mainstream development. This is likely due to its steep learning curve. However, the tooling around Nix has matured significantly, so it's now worth giving it a try.

While some engineers can't seem to stop raving about how much of a superpower large language models are for coding, others find it to be an extremely finicky process that requires an extreme amount of coddling to get average results at best. This could be because of the type of work the engineers are doing - AI seems to be productive for greenfield work, but it is pretty useless when it comes to maintaining existing systems, adding more complicated features, or needing to know business domain details. Some engineers find it much quicker and easier to just write code themselves as they always do a better job in the end.

Kodgem built its business around a device that helped people with their posture. Despite the company's growth to $1 million in revenue, there was zero profit for its founders due to several mistakes, including inventory planning disasters, no real financial management, and failed marketplace partnerships. The founders never paid themselves, causing them to almost burn out completely multiple times. They had gotten caught up in the growth story and forgotten the most basic rule of business: you need to make more money than you spend.

The biggest banks in the US are in discussions about whether to team up and issue a joint stablecoin. The talks are in early stages and any final decision would depend on how stablecoins will be legislated. Banks have been bracing for the possibility that stablecoins could become widely adopted under President Trump. Stablecoins could speed up more routine transactions, but there is still some skepticism about the security of the technology and the regulatory implications of getting involved with digital assets.

People make technical decisions, in part, because the technology feels good, comfortable, or familiar - and that's fine.

People use data to make decisions because they think that data is the truest form of objectivity, but bias isn't something that can be easily removed - it's important to remember that data isn't any more objective than we are.

We're building a world where curiosity gets lobotomized at the door - the industry is going to be filled with people who think they're good because of their bots, while people who put in the effort will get chewed up by a system that doesn't value understanding anymore.

This article discusses an approach to lossless video compression that involves repurposing Bloom filters and requires no external data for decompression.

The Dave & Buster's anomaly demonstrates how Apple's BlastDoor mechanism, which was designed to thwart hacking attempts, works as designed.

This post demonstrates the optimal flow for creating new Debian packages.

Authentication isn't just login. It's building secure flows, managing sessions, handling tokens, supporting SSO, and protecting user data. AuthKit provides a fully hosted, customizable UI for sign-up, password resets, MFA, and seamless SSO — production-ready from day one.

But authentication alone doesn't stop abuse. WorkOS Radar adds real-time threat prevention, detecting brute force, credential stuffing, and more. Built-in protections can be enabled instantly — no scripts or custom logic required.

Build secure, scalable authentication from day one with AuthKit and WorkSO Radar.

Nvidia's DGX Cloud Lepton is a new service designed to link artificial intelligence developers with Nvidia's network of cloud providers, including CoreWeave, Lambda, and Crusoe. It provides access to Nvidia's graphics processing units, making those chips more widely accessible to all developers. The platform will have a marketplace of GPU cloud vendors that developers can pick from to train and use their AI models. Developers can choose which cloud provider to work with and even work with multiple cloud providers.

Demis Hassabis, the chief executive of Google DeepMind, has been dreaming of artificial general intelligence (AGI) for years. He was already taking the possibility of AGI seriously when Google acquired DeepMind in 2014. Today, he is among a growing number of tech leaders racing to build AGI. Hassabis and a colleague received the Nobel Prize in Chemistry last year for their work on AlphaFold, an AI system capable of predicting the three-dimensional structures of proteins. This article contains a transcript of an interview with Hassabis about his views on AGI and the future that follows its arrival.

Autonomous robots aren't being used to operate on people yet, but the technology is almost capable of such a feat and more autonomy is on the way. The technology could help hospitals meet the high demand for safety and consistency during surgeries, manage routine tasks, and prevent mistakes. Future robots will use AI to autonomously manage movements and decision-making during surgical tasks, reducing the need for constant human input. With the number of surgeons around the world declining while the number of people who need surgery continues to increase, the need for the technology is clear.

Starfish Neuroscience, a brain-computer interface startup incorporated by Valve co-founder and CEO Gabe Newell, plans to produce its very first brain chip later this year. The company is currently seeking collaborators to open up avenues for the technology. Starfish's goal is to create a smaller and less invasive implant than the competition that can enable simultaneous access to multiple brain regions. Details from the chip's current spec sheet are available in the article. The article also links to a video of a talk from Valve that explains how the technology could enhance gaming.

Will Larson worked at Carta as its CTO for the past two years. This post contains the most important lessons Larson learned in his time at the company, covering topics such as engineering strategy, large language model adoption, multi-dimensional tradeoffs, creating software quality, and more.

Bohan Zhang, a member of OpenAI's Infrastructure team, recently shared the company's best practices with PostgreSQL at the PGConf.dev 2025 Global Developer Conference. The talk offers a glimpse into the database usage of one of the most prominent unicorn companies and demonstrates how PostgreSQL can scale gracefully under massive read loads. OpenAI utilizes managed databases on Azure, employing a classic PostgreSQL primary-replication architecture without sharding. This article covers the content of Zhang's talk along with comments and includes some more details about OpenAI's infrastructure.

As AI spreads through the labor force, many white-collar workers say their work is becoming more routine, less thoughtful, and much faster-paced. Companies seem to be persuaded that AI can increase productivity. Teams are getting smaller, but they're being expected to output the same amount of work by using AI. Managers say that AI can relieve employees of tedious tasks and enable them to perform more interesting work, but for inexperienced workers, the change resembles the shift from artisanal work to factory work in the 19th and 20th centuries.

Large language models are becoming commodities. Many people use them, but few people are developing them from scratch. Developing large language models involves a lot of hardcore computer science and math - people can use them without having to know any of the internals. While it might be helpful to understand some of the internals and math, it's probably best to stick to learning how to get good at using these models and integrating what they do into agentic systems.

We also write TLDR AI. Same format, all about LLMs, GenAI, and deep learning. Delivered every weekday. Subscribe for free.

F2 is a cross-platform command-line tool for batch renaming files and directories quickly and safely.

MCP could inspire a new generation of developers to push for the web to go back towards its natural architecture, programmable through janky specs that everybody adopted for the sheer joy of something fun to hack on.

TrailBase is an application server with sub-millisecond latencies, eliminating the need for dedicated caches.

Rooting for your friends is the best way to improve your career, happiness, and the people you care about.

You should learn from others, but you also need to put your knowledge to the test so you learn enough about your field to advance it.

This post presents things to consider if you are thinking about creating a new file format.

InfraHub Compute now lets private investors buy GPU-accelerated servers plugged into our global API, powering 45+ resellers.

Once exclusive to Big Tech, this first-of-its-kind platform offers hands-off, income-producing assets.

→ Proven 4-year track record→ 3–4% revenue per month→ Real-time monitoring portal→ Open to private investors

What Airbnb is to property owners, InfraHub Compute is to hardware owners.

With AI compute demand soaring toward a $1.8 trillion industry by 2030, capacity is limited and demand is high. Invest early in the AI revolution's backbone.

Learn how to power your portfolio with GPU infrastructure.

OpenAI plans to ship 100 million AI 'companions' that will be capable of being fully aware of users' surroundings and lives. The company aims to create a third core device for people to use on a desk after a computer and mobile device. The device will be unobtrusive and be able to rest in one's pocket or desk. It won't be a phone - one of the design goals is to help wean users from screens.

Anthropic has launched Claude 4, a new group of artificial intelligence models. The company claims that Claude Opus 4 and Claude Sonnet 4 define a new standard when it comes to AI agents. The models can analyze thousands of data sources, execute long-running tasks, write human-quality content, and perform complex actions. Both models can search the web to complete tasks and alternate between reasoning and tool use.

High NA, which stands for high numerical aperture, is the world's most advanced and expensive chipmaking machine. Costing more than $400 million, the machine took ASML nearly a decade to develop. It is made up of four modules that are manufactured around the world and assembled in the Netherlands. Only five of the colossal machines have ever been shipped. This article takes a look at High NA, how ASML plans to deploy the technology, the benefits that High NA provides, how the technology works, the effects that China, tariffs, and US growth will have on the technology, and more. A link to a video showing High NA in action is available at the end of the article.

The Tor Alva, the world's tallest 3D-printed building, has been completed in Switzerland. The structure, which the creators liken to an 'ornate layered cake', sits at an impressive height of 30 meters. It is located in a village that's currently home to just 11 people. The building took around 900 hours of printing to complete using two construction robots: a 3D printer and a machine used to add steel reinforcement. Pictures of Tor Alva are available in the article.

The world's biggest event for AI builders is in SF from June 3-5. Speakers include leaders from OpenAI, DeepMind, Anthropic, and more, click here to get 30% off your ticket.

Anthropic has announced four new capabilities on its API that enable developers to build more powerful AI agents. The code execution tool, MCP connector, Files API, and ability to cache prompts for up to an hour will enable developers to build agents that execute code for advanced data analysis, connect to external systems through MCP servers, store and access files efficiently across sessions, and maintain context for up to 60 minutes without building custom infrastructure. All of the features are now available in public beta on the Anthropic API.

Stripe did an average of 1,145 pull requests per day in 2024 while having less than a minute of API unreliability for the entire year. The company is likely in the top 1% in software delivery performance. It has a very demanding but advanced engineering culture with a relentless focus on continuously shipping value for users. Shipping safely at this velocity and scale implies heavy investment in automated tests, deployments, rollbacks, observability, code ownership, and other areas.

Meta's managers have been instructed to rank more employees as 'below expectations' in performance ratings. This could set the stage for more performance-based cuts. The review process is set to begin on June 16. Conversations between managers and employees on performance are scheduled for between July and August.

While everyone seems to want to replace software engineers, AI just isn't there yet. Not everyone can just dive into a problem with little context and get started - it takes time to understand the problem and ask relevant questions to determine requirements. There is value in being able to say 'I don't know how to do this', but this is less and less likely to happen now that people are falling back to AI for answers when all else fails. People learn by doing something hard - struggle is part of improvement.

According to Apple supply chain analyst Ming-Chi Kuo, the device OpenAI is working on with Apple's former design chief Jony Ive will be slightly larger than Humane's discontinued AI Pin but as compact and elegant as an iPod Shuffle.

Apple's smart glasses are expected to include cameras, microphones, and AI capabilities.

The group is led by the Forest Road Company, a Los Angeles-based investment firm.

Xiaomi's Xring O1 is a custom 3nm chip that will be initially launched in three devices, including the new 14-inch Xiaomi Tablet 7 Ultra.

Developers should be occupying their energy with solving actual problems - to get there, they need to learn how not to read code.

If you're using Postgres 17+ on both the server and client, you can just start with a TLS handshake, making requests much faster.

The distance between teams that maximize AI coding assistants and those that merely dabble is growing every sprint. DX's guide to AI-assisted engineering can help you and your teams close the gap.

Based on data from DX's customer base and interviews with experienced AI users, this guide delivers concrete strategies that actually work. Discover prompting techniques that elicit better responses, the top time-saving use cases for AI, and leadership strategies for encouraging AI adoption.

Get the guide, distribute it internally, and enable teams to ramp up with AI-powered development. Download free from DX

The Tesla Optimus team's latest video demonstration shows the humanoid robot performing a variety of tasks, including household using a broom and vacuum cleaner, tearing a paper towel, stirring a pot of food, opening a cabinet, closing a curtain, and picking up a Model X fore link and placing it on a dolly. The robot completed all these tasks through a single neural network. It learned its actions using data from first-person videos of humans performing similar tasks. The system could pave the way for Optimus to learn and refine new skills quickly and reliably. The video is available in the article.

OpenAI is acquiring Jony Ive's startup, io, in an all-equity deal that values the startup at $6.5 billion. Ive and his design firm, LoveFrom, which will continue to operate independently, will now lead creative and design work at OpenAI. He could help OpenAI directly compete with Apple in the consumer hardware space - Apple has struggled to develop AI features that keep up with the latest technology. io's staff of around 55 engineers, scientists, researchers, physicists, and product development specialists will join OpenAI. They will be tasked with developing AI-powered consumer devices and other projects.

Ryan Williams, a theoretical computer scientist at the Massachusetts Institute of Technology, recently posted a proof online on the relationship between time and memory in computing. The proof, which showed that memory was more powerful than computer scientists had previously believed, established a mathematical procedure for transforming any algorithm, no matter what it does, into a form that uses much less space. It also implies there are calculations that can't be computed in certain amounts of time, offering a new way to attack one of the oldest open problems in computer science.

Robot.com is a Colombian robotic company with a fleet of over 500 robots working in warehouse logistics, advertising displays, security, inspection, and deliveries. The company signed a multi-year partnership with Amazon Web Services last year, catapulting it into the big league of the global service robots market. The steady growth of the company underscores the potential of the robotics industry in Latin America, but it also highlights the challenges of regions with high production costs, limited investor confidence, and minimal public investment.

The world's biggest event for AI builders is in SF from June 3-5. Speakers include leaders from OpenAI, DeepMind, Anthropic, and more, click here to get 30% off your ticket.

Legacy code is full of surprises. Mainframes are chock-full of history, containing wonderful hacks that get around the limitations of systems. This post contains several examples of these hacks within a COBOL system.

LlamaIndex is an ecosystem consisting of more than 650 Python packages, mostly Integrations and Packs. All of these packages share a single GitHub repository, or monorepo. This article discusses LlamaDev, a tool for managing monorepos at scale, and explains the challenges with existing tooling. LlamaDev makes running tests faster and easier, creates clearer logs so contributors know exactly what happened if a check failed, and can be run locally for debugging.

Tesla's head of AI and self-driving, Ashok Elluswamy, recently admitted in an interview that the automaker's autonomous program is lagging a couple of years behind Waymo's. However, Tesla's approach is much cheaper, and its cost advantage will allow the company to scale faster. Tesla produces over a million cars a year, versus Waymo's few hundred units, and it plans to use the same vehicles for its autonomous riding service. The company is still limited to a level 2 advanced driver system, which requires constant supervision from the driver, while Waymo has been providing customers with level 4 autonomous driving rides for years. It plans to start offering level 4 autonomous rides to customers in Austin next month.

TLDR is hiring a curator with experience founding a company, raising money, and growing a team to join our TLDR Founders team. TLDR Founders has over 300,000 subscribers, it's a great way to build a personal brand within the founder community. To apply please send your LinkedIn or resume to [email protected] along with a couple sentences on how you keep up with news for founders!

Google is getting back into the smart glasses game. The company has announced a partnership with Warby Parker - they plan to launch a series of smart glasses built on top of Google's Android XR as soon as next year. The rise of generative AI has allowed Google to revisit the idea of Google Glass, a wearable it launched in 2013 that failed partly due to a technology gap. The company has learned many lessons since the failed launch. Other companies, like Meta and Apple, are also reportedly working on smart glasses.

China welcomes foreign companies and competition but creates conditions where local alternatives can thrive alongside them, and its provinces, despite competing against each other like countries in the EU, all have the ultimate goal of developing a Chinese champion that can successfully compete on the global stage.

Google plans to start testing ads in AI mode, a new AI-powered search feature the company just rolled out to the US.

Spotify's iOS subscriptions have significantly increased in the past two weeks, while sales have remained relatively constant on Android.

Shopify's Storefront Web Components lets developers bring Shopify-powered commerce capabilities to any website.

Microsoft says it has been unable to implement linked-out payments because of Apple's anti-steering policies, which restrict Microsoft's communications to users and impose a high economic cost.

The Feiyang Times building in Shenzhen is known as a place for phone sales and repairs - and also a marketplace for stolen devices.