Google plans to unveil the Pixel 10 series on August 20, but a new leak may have already revealed everything about the devices, along with details on the Pixel Watch 4 and the Pixel Buds 2a. The leaked images show the devices and their various color options and spec sheets. The Pixel 10 series may ditch physical SIM cards in the US, but the Pixel 10 Pro Fold will likely hold out and retain its physical SIM tray. The leaked images are available in the article.
Large language models are currently good tools for certain jobs, but they still have many shortcomings that haven't been fixed since their inception, and these problems don't look to be solvable with the current approach.
POML is a novel markup language that provides a systematic way to organize prompt components, integrate diverse data types, and manage presentation variations, addressing common challenges in prompt development and empowering developers to create more sophisticated and reliable AI applications.
In a live webinar next Thursday [4], Sapient Insights will share their latest research showing HR leaders aren't just supporting tech decisions—they're shaping them. From AI governance to workforce alignment, HR is playing a key role in enterprise tech strategy.
OpenAI has announced GPT-5 in three variants: GPT-5 Pro, GPT-5 mini, and GPT-5 nano. Some of the models will be available across all ChatGPT, including for free users. OpenAI claims the new model family comes with reduced confabulations, improved coding capabilities, and a new approach to handling sensitive requests. Free users now have access to a simulated reasoning AI model. GPT-5 Pro is replacing o3-pro in ChatGPT for those subscriber tiers with access to it.
Meta's TBD Lab houses many of the researchers the company lured away from rival labs. It is spearheading Work on the newest version of Meta's large language model, Llama. The TBD Lab will work alongside Meta's other AI teams on a variety of projects. Meta has to date poached at least 18 researchers from OpenAI and a number from Google.
The US Food and Drug Administration has approved an eye drop that improves near vision in adults with presbyopia, a condition that causes near-sighted vision loss. The aceclidine-based eye drops offer relief from blurry near vision for up to 10 hours. They work by gently shrinking the pupil of the eye, creating a pinhole effect that helps bring nearby objects into sharper focus. The eyedrops do not significantly affect the eye's focusing muscles, so using them doesn't blur distance vision or cause myopic shift.
Eli Lilly has developed a new pill that causes obese people to lose significant amounts of weight. The drug targets GLP-1, potentially increasing access to GLP-1 medications due to price and ease of application. Its side effect profile is similar to that of the injectable GLP-1 drugs. Eli Lilly plans to file for approval for the drug by the end of the year.
GPT-5 rarely screws up and generally feels competent. The model used in ChatGPT is a hybrid that switches between different models. GPT-5 is available as three models in the API: regular, mini, and nano. Each can be run at one of four reasoning levels. This post looks at the key characteristics of the models, how they are priced, and what can be learned from the GPT-5 system card.
Ignore the claims about vibe coding or that you don't need to know how to write code, and instead look for ways to augment what you do. Experiment with AI and discover what value you can extract - there's a lot of grift. Learn for yourself what works and what doesn't. AI in its current form won't ever replace hands-on-keyboards, and it won't replace engineering.
Orchid is a company that lets parents genetically test embryos before pregnancy. Parents can choose their future baby for as little as $2,500. This article features a transcript of an interview with the company's founder, where they debate the scientific, moral, and ethical implications of designing a 'healthy' child and what we lose in separating reproduction from sex. With Orchid's technology, parents in the future will not have to roll the dice with their child's health.
Intel's Chief Executive Lip-Bu Tan and some Intel directors disagree on topics as central as whether the company should stay in the manufacturing business or exit it entirely. Tan recently tried to raise new capital to acquire an artificial intelligence company, but efforts have been stalled by people on the board. Despite Trump unexpectedly calling for Tan's ouster, the company is so far standing by Tan, saying that the board and Tan are all deeply committed to advancing US national and economic security interests. While Intel's chip factories are a money loser, some view them as politically important because they help secure the US semiconductor supply chain.
Tesla's Dojo team has lost 20 workers to a recently formed startup, - remaining Dojo workers will be reassigned to other data center and compute projects within Tesla.
The National Highway Traffic Safety Administration has cleared Zoox to deploy its fleet of steering wheel-less electric pods for demonstration purposes.
There's big investor money in project management. That's a big incentive to make bloated, barely-useable products — designed to tick boxes in a procurement manager's checklist.
OpenAI is in early talks about a potential sale of stock for current and former employees at a value of about $500 billion. Investors have approached OpenAI about buying some of the employee shares. If the deal goes ahead, OpenAI's on-paper price would rise by roughly two-thirds. The company expects to reach 700 million active weekly users this week, up from 500 million at the end of March.
Instagram's new map feature lets users share their most recent active location with others and discover location-based content. The feature also allows users to leave short messages on the map for others to see. Location sharing is off by default in Instagram Map, and users' locations only show up when they open the app, so it doesn't provide real-time location updates. The new feature has launched in the US and will be available globally soon.
One of the most iconic problems in computer science is finding the shortest path from a specific starting point in a network to every other point. A team of researchers has devised a new algorithm that solves the problem without sorting, and it runs faster than any algorithm that does, including Dijkstra's algorithm. The new algorithm's runtime breaks the previous sorting barrier, and it can likely be streamlined to run even faster. This article tells the story of how the algorithm was developed.
Fusion projects such as the International Thermonuclear Experimental Reactor (ITER) rely on 316LN or JK2LB cryogenic stainless steels, which hold up well at liquid-helium temperatures, but their yield strength limits the magnetic field of ITER at 11.8 tesla and forces a large, expensive machine design. Chinese developers created CHSN01 (China High-Strength, Low-Temperature Steel No. 1) to withstand 20 Tesla fields and 1.3 GPa combined electromagnetic stress while still stretching about 30% before breaking. It keeps those properties after 60,000 on/off cycles, the lifetime workload of China's Burning-Plasma Experimental Superconducting Tokamak (BEST). CHSN01 can be used in MRI scanners, particle accelerators, maglev trains, and even quantum-computing dilution refrigerators - swapping in a stronger and tougher steel could shrink magnet footprints and extend service intervals across the board.
Join thousands of fellow developers, engineers, and technologists for hands-on labs, open source tooling, and real-world certifications at IBM TechXchange 2025 [10].
The concept of lockfiles is unnecessary, and it complicates things for no reason. Maven, a 20-year-old Java library ecosystem, has never once needed a lock file. Dependency managers can and are working without lockfiles just the same.
comptime.ts is a simple TypeScript compiler that enables compile-time evaluation of expressions marked with 'comptime'. It can help optimize code by moving computations from run time to compile time. comptime should not be used for complex runtime logic, side effects, or non-deterministic operations. It is useful for computing constant values, generating static content, and optimizing performance-critical code.
The world of tech is changing: long-term employees are reporting a completely different work environment from what they experienced years ago. Pay and bonuses are declining, as well as other perks, like a relaxed atmosphere and occasional fun office activities like Nerf gun fights. Tech behemoths have aged into large bureaucracies. While many of them still provide free food and pay well, they have little compunction about cutting jobs, ordering mandatory office attendance, or clamping down on employee debate.
The two most important companies of the last two decades of tech were Apple and Amazon, specifically AWS. Apple created the smartphone paradigm with the iPhone, and AWS' cloud platform enabled a universe of applications that ran everywhere. Microsoft and Nokia failed in the smartphone era because they both deluded themselves into thinking that their previous domination was an advantage. AI is a paradigm shift, just like mobile was. While the winners of the previous paradigm are not automatically fundamentally handicapped because of their previous successes, they need to be careful about making the wrong choices for years before it's too late.
Upgrade your health with Simple—an AI-powered weight loss app trusted by 18M+. No diets, no fluff. Just smart, science-backed coaching built for busy people. TLDR readers get 50% off → Take the
Large language models can be used to turn short, simple content into something long and seemingly profound - this should make us consider why we find ourselves inflating content.
Apple is highly likely to introduce a new Apple TV, with a newer chip, support for Wi-Fi 7, and an Apple-designed Bluetooth and Wi-Fi chip, later this year.
Manual testing on mobile devices is TOO SLOW AND TOO LIMITED. It forces teams to CUT RELEASES A WEEK EARLY just to test before submitting them to app stores. And without broad device coverage, issues slip through.
Anthropic's new model, Opus 4.1, is more capable at coding, research, and data analysis and better at fielding complex multistep problems than previous versions. It can better navigate large codebases and make more precise modifications to code. Opus 4.1 scores two percentage points higher than its predecessor on SWE-Bench Verified, a popular coding evaluation benchmark. Anthropic is currently finalizing a deal to raise as much as $5 billion in a new funding round at a valuation of $170 billion.
OpenAI's long-awaited open weight models have been released. The 120B model achieves near-party with OpenAI o4-mini on core reasoning benchmarks while running efficiently on a single 80 GB GPU. The 20B model delivers similar results to OpenAI o3-mini on common benchmarks and can run on edge devices with just 16 GB of memory. Both models are mixture-of-experts. It is likely that OpenAI now offers the best available open weights models.