Samsung Galaxy leaks 📱, Meta's $3.5B Ray-Ban bet 👓, AI impacts dev hiring 👨‍💻

TLDR 2025-07-09

Jul 09, 2025

Samsung's event spoiled by massive last-minute leak (2 minute read)

A reliable leaker has shared a whole bunch of marketing materials for Samsung's Galaxy Unpacked event (which kicks off today at 10 AM ET) containing details and images of the company's upcoming devices. The Z Fold 7 will measure 4.2 millimeters when unfolded and the Z Flip 7 will measure 6.5 millimeters when unfolded. None of the marketing materials highlight an S Pen on the Z Fold 7, suggesting that the device may not support an S Pen at all. The leaked material also contains images and specs for new Galaxy Watches, which have upgraded storage, displays, and batteries.

Meta Invests $3.5 Billion in World's Largest Eye-Wear Maker in AI Glasses Push (3 minute read)

Meta has bought a minority stake in EssilorLuxottica. It acquired just under 3% of the Ray-Ban maker, a stake worth around $3.5 billion at the current market price. Meta is considering further investment that could build the stake to around 5% over time, though those plans could still change. The deal provides EssilorLuxottica with a deeper presence in the tech world.

🚀Science & Futuristic Technology

China jumps ahead in the race to achieve a new kind of reuse in space (9 minute read)

China likely completed the first high-altitude attempt at orbital refueling sometime last week. Two of its satellites appeared to dock together in geosynchronous orbit, coming together and then becoming indistinguishable as a single object. Besides refueling satellites in space, the technology could also be used to approach, capture, and disable other satellites. The US Space Force plans to perform the first refueling of a US military asset in orbit as soon as next year.

CRISPR breakthrough allows scientists to edit multiple genes simultaneously (2 minute read)

Scientists at ETH Zurich have developed a new CRISPR method that can modify dozens of genes simultaneously. It can be used to systematically modify entire gene networks in a single step. The technique uses the lesser-known Cas12a enzyme, which is slightly more precise in its ability to identify targeted genes and can handle shorter RNA molecules compared to the traditional Cas9 enzyme. It enables scientists to explore broad genetic interactions.

💻Programming, Design & Data Science

Introducing OpenCLI (2 minute read)

The OpenCLI specification defines a standard, platform and language-agnostic interface to command-line applications that allows both humans and computers to understand how a tool should be invoked without access to source code or documentation. With the growing interest in MCP (Model Context Protocol) and command-line automation, there's a huge potential in standardizing how command-line applications are described. The specification is still a draft and incomplete. It is maintained under the Specter.Console GitHub organization.

Gemini Nano in Chrome 137: notes for AI Engineers (5 minute read)

Gemini Nano is starting to be shipped unflagged in Chrome and will likely be shipped fully unflagged by the end of the year. There are a few APIs available for commonly used patterns, but the main one engineers will be interested in is the Prompt API, which is the most flexible and open-ended one. This article goes through how to set up and use Gemini Nano. It covers the basic important things and pitfalls to be aware of when using the API.

🎁Miscellaneous

Tech Philosophy and AI Opportunity (28 minute read)

While AI is hailed as the route to abundance, the most important financial outcomes have been about scarcity. One of these scarce resources is AI talent. As the work needed to be done is the same across the various companies bidding for talent, it makes sense that the team that the best researchers play for is the team that pays the most. The teams that are destined to lose are the ones who can't or won't offer enough money or mission.

CTOs Reveal How AI Changed Software Developer Hiring in 2025 (18 minute read)

While AI evangelists promise the technology will make everyone become a '10x developer', the reality is that companies are drowning in AI-generated code that looks perfect but breaks in production. The most valuable developers aren't the ones churning out AI code, but the ones who can think their way out of the problems AI creates. Engineering leaders want developers who can look at AI-generated code and figure out what's wrong with it instead of just accepting whatever the AI produces. AI was supposed to make coding easier, but it's actually making the thinking parts of development more valuable than ever.

TLDR Newsletter

Discussion about this post