Full-text search across article titles and summaries. Refine with sector, year, content type, and reading-time filters.
When we are zoomed in on one specific thing, we often end up completely missing the holistic picture. The challenge with blind spots is that, unless you are aware of them, they are not visible. To identify blind spots, we can start by asking the hard questions, seeking missing information, getting outside opinions, and developing a trusted support network.
Sponsored
for developers like you: front-end, back-end, product - you name it. And here’s the best part…
This article discusses React tricks learned from maintaining the open-source React library “Wouter”. Some tricks mentioned include using React.cloneElement for component composition, optimizing component rendering with React.memo, using useState for resource initialization, and stable callbacks with useEvent. The article also touches on managing external state changes with useSyncExternalStore and addressing tearing issues with hashchange events in React.
This article covers web development trends in 2023. The year showed the continued dominance of React, the rise of Astro as a popular framework, Vercel's success with its AI SDK, the growth of Cloud Development Environments (CDEs), steady progress in WebAssembly, and a renaissance in the open web. Generative AI is bound to help speed up web development even more in 2024.
Leave work slightly unfinished in order to reach a productive “flow state” easier the next day. Learn keyboard and mouse shortcuts, maintain a list of common commands and references, and learn to say “no” more often. This includes saying “no” both in terms of choosing technology and tasks that align with your goals and priorities
Mentoring is valuable for professional growth. A few different approaches to finding mentors include seeking internal mentors at large companies, considering paid coaches for engineering leaders at smaller companies, and connecting with experienced individuals both in-person and online. Mentoring doesn't have to be formal and structured, as even informal mentorship can be valuable and initiated by simply asking someone you admire for their time and advice.
Sponsored
As your codebase grows, maintaining quality can become a challenge. Vercel’s Developer Experience Platform’s [11] new features, designed with larger teams in mind, include automated issue detection, static analysis, responsibility tracking, and a collaborative dashboard.
Bytebase provides advanced database DevOps and CI/CD for engineering teams. It provides a unified process to manage, query, secure, and govern all of a team’s databases from one place.
Baseline is a new feature introduced this year that tells you when a specific web platform feature (such as CSS or JS) can be considered safe to use in a web browser. This post highlights various features that became part of Baseline in 2023, including size container queries, new color spaces and functions, compression streams, offscreen canvas, and module preload.
Swift's inclusion in Apple's operating systems has benefits such as not requiring apps to embed Swift libraries. However, it also creates challenges such as the inability to remove or change APIs without affecting existing apps. Alternatives to Apple’s approach, such as library deduplication, explicit "polyfill" support, and explicit backwards-deployment support, improve the developer experience and help manage the evolution of Swift in the OS.
A browser’s URL/search bar is an expensive piece of real estate. Google pays Firefox approximately $400 million and Safari receives between $15-20 billion for Google to be the default search engine. This article floats the idea that creating a popular browser and selling the default search preference could be a viable business strategy given the lucrative nature of the browser's URL/search bar.
Teams need to become more product-oriented. This means aligning non-product managers to think like product managers.
A detailed retrospective of open-source LLMs this year. This year has been fast-paced for AI and there are new architectures that have appeared that may replace the Transformer.
There are two types of work: harmlessly-interruptible work and Do Not Disturb work, and interruptions can significantly impact the latter. This means that private offices are useful for programmers to be productive.
Sponsored
Between GitHub, Jira/Linear, and Slack, engineering teams create a lot of data. But measuring the wrong things can destroy motivation and productivity. Swarmia [5] has identified key metrics [5] that are more signal than noise (including DORA and SPACE) — and makes them available to engineering leaders, managers, and teams. No leaderboards, no individual performance data.
Meta's second-generation Ray-Ban glasses can interpret scenes and make judgments using generative AI. A new update is rolling out a feature that uses the on-glasses cameras to look at images and interpret them. The update also brings Bing-powered search to the glasses. This article contains some examples of how the feature can be used.
Apple's Vision Pro will redefine our relationship with visual media and content, creating many new possibilities for photographers and filmmakers. Spatial video is a mixed-reality video format that allows videos to record the depth and spatial information of a scene, creating a more immersive three-dimensional experience. It is able to create super immersive photos and videos that look amazing on the Vision Pro. It will not be long before creators start thinking of using the technology to tell a story.
Scientists at the US National Ignition Facility announced that they had achieved ignition in December 2022 after more than a decade of effort. The team can now replicate the feat again and again. It has achieved ignition in four out of its last six attempts. The latest series of experiments features a 7% boost in laser energy, which in theory should lead to larger yields. The team now plans to focus on changes to produce a more symmetrical implosion.
Japan recently approved a COVID-19 vaccine that triggers the body to produce an enzyme that makes copies of its antigen mRNA. It is the first fully approved self-amplifying RNA vaccine. The technology allows smaller doses of the vaccine to achieve the same level of efficacy as current vaccines with cheaper shots, shorter manufacturing times, and fewer side effects. Self-amplifying RNA vaccines could prove to be a powerful weapon in the fight against cancer.
This guide shares strategies and tactics for getting better results from large language models. The methods can sometimes be combined for greater effect. Some examples only work with the most capable models. Strategies for getting better results include writing clearer instructions, providing reference texts, splitting complex tasks into simpler subtasks, and using external tools.
Engineering management has changed a lot over the past decade. Management seems a lot less attractive these days. However, engineering managers are incredibly important and still needed. Managers are responsible for the systems and structural support that enable engineers to deliver products and outcomes. This article lists reasons for engineers to become managers. One of the main reasons is that becoming a manager can make you better at life and relationships - the role changes who you are as a person and the relationships you have with others.
This article contains several stories about new developments in AI. These stories cover the competition in the AI space, the state of employment in the industry, AI strategies, Sourcegraph's Cody AI coding assistant being generally available, and more. Engineers who have committed to commenting on their code throughout the years have provided incredibly valuable context to help large language models reason about their code. The article takes a moment to thank those coders for their service in helping out AI.
A team of researchers has proposed a novel framework that treats AI programs as if they were a participant in a psychological study. It has three steps: test the program in a set of experiments examining its inferences, test its understanding of its own way of reasoning, and examine, if possible, the cognitive adequacy of the source code for the program. While current generations of AI programs have come close to passing the Turing Test, the results don't imply that these programs can think and reason like humans. The new framework could be an alternative to the Turing Test for evaluating the intelligence of programs.
Sponsored
Boost your human intelligence with Brilliant’s bite-sized lessons in AI, LLMs, machine learning, and more. Think smarter – unlock possibilities. Try Brilliant free for 30 days.
This post provides advice for those looking at early-stage VC - it will be continually updated with additional inputs to be useful for candidates in the future as well.
Consulting is a smart way to self-finance a startup, but founders need to be serious, disciplined, and relentless - this article presents traps to avoid for consulting companies.
The usual technology paradigms will not serve as an adequate guide for navigating the new world of AI.
There are no known fundamental barriers preventing humans from significantly enhancing adult intelligence with gene editing - many pieces of a viable editing protocol have already been demonstrated independently.
AWS' new lower-latency S3 storage class costs 8 times more than S3 standard - while it is expensive, it might be attractive to new startups as it is a significant time saver.
A framework for understanding other people's hopes and fears about AI.
Sponsored
How can you avoid incomplete, inaccurate responses in your GenAl app? Neo4j knowledge graphs [7] offer the capabilities you need. 🚀 Power LLMS with your company's data using retrieval augmented generation (RAG)
Video understanding is an old field that is just now starting to find success. The main driver of success is quality datasets. This work from Meta includes many sensor data sources and hand poses. It provides baseline performance for existing methods.
Snapchat is introducing new AI-powered features for Snapchat+ subscribers, including an AI image generator that creates images from text prompts and a Dream selfie feature that creates fantastical images with friends. A new AI extend tool allows users to automatically expand and fill in the background of photos. These features are rolling out regionally to over 7 million Snapchat+ subscribers.
Google has launched Gemini Pro, a lightweight version of its advanced AI model Gemini Ultra, for public preview in Vertex AI, offering text and image processing capabilities and customizable contexts for various use cases.
Transformers, as amazing as they are, are often hard to train, with their many instabilities. One of the main issues is when an Attention matrix’s entropy collapses. This work is a way to prevent that with a simple reparametrization.
This project introduces a novel technique for estimating lighting in images, overcoming the limitations of current methods that rely on HDR panorama datasets. By using diffusion models to render chrome balls into standard images, the technique reveals a unique relationship between chrome balls and diffusion noise.
DriveMLM is a new framework that uses large language models to enhance autonomous driving. This system, which combines language decisions with vehicle controls, not only integrates with existing autonomous driving systems, but also shows better performance in simulations.
extracts and transform complex data (webpages, docs, images…) for use with every major vector database and LLM framework [16]. “Unstructured has solved the most difficult part of building an LLM application.” — LangChain Co-founder Harrison Chase. Make your
Sponsored
LLMs work best with clean, curated data. But most of this data is
Coffee, launched last week, helps streamline frontend development with AI by plugging into existing codebases. The project focuses specifically on a first-class DX based on learnings the Coframe team has gathered from their own experience in generating 80%+ of their frontend with AI.
DeepEval is an all-in-one evaluation framework for LLM apps that allows users to easily identify and iterate on unsatisfactory LLM outputs. It currently offers 14+ evaluation metrics such as hallucination, summarization, G-Eval, RAGAS, etc., for users to evaluate entire datasets in a Pytest-like or component-based fashion.