Cloudflare has developed a new feature to safeguard GraphQL APIs from harmful queries that could lead to a denial of service. The protection mechanism involves the use of a custom-built GraphQL query parser, which provides detailed insights into requests and enables the application of various security settings. This approach effectively manages the unique challenges posed by the single endpoint and highly flexible query structure of GraphQL APIs.
Becoming a senior developer and a better engineer often comes through not just experience, but also through evolving the way you think about and approach problems. This article lays out 9 non-technical mental models that help you think like a senior developer.
The blog post presents a theory that Reddit's recent API changes are driven by the interests of OpenAI, given the significant stake and influence that OpenAI stakeholders have in Reddit. The author suggests that Reddit's data is extremely valuable as training data for AI models, and by restricting access to this data, OpenAI can create a competitive advantage or moat. The post also speculates that Reddit's future actions might include buying out more third-party clients or extending the deadline for API changes.
Overcoming burnout is tough. It decreases productivity and affects your mental health negatively. However, there are ways to come back from burnout and put measures into place to make sure you don’t burn out again. For example, taking regular breaks, having regular self-check-ins, and following a well thought out plan are measures that can help you tackle burnout as a developer.
The Vercel AI SDK is an open-source library designed to help developers build conversational, streaming, and chat user interfaces in JavaScript and TypeScript. The SDK supports React/Next.js, Svelte/SvelteKit, with support for Nuxt/Vue coming soon. It features first-class support for LangChain and OpenAI, Anthropic, and HuggingFace, Edge Runtime compatibility, and callbacks for saving completed streaming responses to a database (in the same request).
OpenObserve is a cloud-native observability platform built specifically for logs, metrics, traces and analytics designed to work at petabyte scale. It’s easy to get up and running in less than 2 minutes.
GitHub surveyed 500 developers in the US to get a sense of how they use AI coding tools. Gergely Orosz of The Pragmatic Engineer examines the results and adds context on how the survey was conducted. 92% of developers in the survey said they use AI coding tools at work. The top 3 areas where AI coding tools helped the most were in learning, productivity, and focus.
2023.06.16·
newsletter.pragmaticengineer.com·3 min read·TLDR Web Dev
Apple’s Vision Pro isn’t an iPhone moment, or at least, not exactly. Rather, the Vision Pro actually looks a bit more like the original Macintosh, which was over $7,000 (adjusted for inflation) when it launched in 1984 and most people didn’t know why they needed one. Apple has decided that the capabilities of the Vision Pro are the minimum viable product - that it just isn’t worth making or selling a device without a screen so good you can’t see the pixels, pass-through where you can’t see any lag, perfect eye-tracking, and perfect hand-tracking. However, this still doesn’t guarantee success. This analysis breaks down Apple’s approach to the Vision Pro, how it compares to Meta’s approach to VR, and how long it might take for this battle to play out.
Reddit is facing backlash from its community due to changes aimed at increasing profitability, including a "premium access" pricing for API users, which threatens third-party applications. In response, thousands of moderators have blacked out over 8,000 subreddits in solidarity with developers.
AI Chrome Extensions can help you to tap into the extraordinary powers of AI to streamline your coding journey. This article features ten extensions that can help you leverage AI in your day to day work.
Since alerting subscribers in the United States that it would begin to curb password sharing on May 23, 2023, Netflix has had the four single largest days of U.S. user acquisition in the four and a half years that Antenna has been measuring the streaming service. Based on the most current data available, Netflix saw nearly 100,000 daily sign-ups on both May 26 and May 27.
The European Parliament has passed a draft law for regulating artificial intelligence (AI). The law restricts the use of AI in certain high-risk areas, such as in the legal system. It would also require companies to disclose more information about the data they use to create AI systems.
Synthesia announced a $90M Series C funding round led by Accel. The funds will boost their mission to make video creation easy for everyone, driving their AI research and enhancing their collaborative video editing platform with better avatar technology, real-time collaboration, and enhanced sharing functionality.
AMD announced its latest AI-dedicated GPU, the MI300X, set to ship later this year. The chip poses a robust challenge to Nvidia's dominance in the AI chip market. This launch aligns with AMD's strategic focus on AI, a sector forecasted to grow from $30 billion to over $150 billion by 2027. Designed for large language models, the MI300X, with its superior memory capacity and support for ROCm software, could drive down costs of AI applications and tap into Nvidia's market share.
Multivariate time series anomaly detection (MTAD) is important for many real-world applications like smart surveillance and risk management. In this study, the authors create a new system called Coupled Attention-based Neural Network (CAN) to better detect these anomalies. It uses a special network that can adapt to changing data, making it more accurate in different scenarios.
This research presents SoViT, a vision transformer that has been optimized for size and shape using advanced scaling methods, resulting in performance competitive with much larger models, but with less computation. Through tests across various tasks like image classification, captioning, visual question answering, and zero-shot transfer, SoViT demonstrates its effectiveness, indicating a new direction for optimizing, rather than simply scaling up vision models.
In this research, the authors created a better way to detect objects in combined color-depth (RGB-D) images using a new training method that requires less supervision and separates different kinds of data effectively.
OmniEvent is an open-source toolkit designed for event extraction, encompassing both event detection and argument extraction processes. The toolkit offers a comprehensive approach for various paradigms, providing fair and unified evaluations on commonly-used English and Chinese datasets. Its modular design ensures a high degree of extensibility.
Closed models are strong because they are actually code models disguised as language models. A lot of the open language models fall short because they don't gain reasoning skills from massive code pretraining. Here, we have Starcoder models instruction tuned on Evol instructions and achieving an astonishing 57 on the HumanEval Benchmarks. Models and code are available.
Audiocraft is a PyTorch library for deep learning research on audio generation. At the moment, it contains the code for MusicGen, a state-of-the-art controllable text-to-music model.
Senators Josh Hawley and Richard Blumenthal want to clarify that the internet's bedrock liability law does not apply to generative AI, per a new bill introduced Wednesday that was shared exclusively with Axios.
Researchers have warned that AI models that are trained on AI-generated content can collapse, meaning they forget the true underlying data distribution and produce erroneous responses. This can lead to problems such as discrimination and a distorted perception of reality.
Microsoft's CEO, Satya Nadella, has led the company to an impressive resurgence in AI tech by integrating OpenAI's language model into Bing via a chatbot named Sydney and launching Copilot, an AI tool that automates aspects of coding. Despite initial hiccups, Microsoft is becoming a heavyweight in the AI arena. The company sees its $10B investment in OpenAI as a key strategic move. Nadella views AI as an empowerment tool for humans and emphasizes the importance of aligning AI with core human values while continuing to explore its potential and manage unintended consequences.
Google will have to postpone starting its artificial intelligence chatbot Bard in the European Union after its main data regulator in the bloc raised privacy concerns.
Doctors are using ChatGPT to improve their bedside manner. ChatGPT can help doctors communicate more compassionately with patients by providing them with words to break bad news, express concerns about a patient's suffering, and more clearly explain medical recommendations.
Delio, a South Korean digital asset manager and lending platform, has temporarily halted customer withdrawals to protect user assets amid heightened market volatility and confusion following Haru Invest's recent suspension of crypto deposits and withdrawals. Delio has committed to regularly update its users about the situation and measures taken to safeguard their assets. Founded in 2018, Delio reportedly holds an estimated $1 billion in Bitcoin, $200 million in Ether and about $8.1 billion in altcoins.
Prometheum, a company that has been around since 2017, recently made headlines for getting approval for a Special Purpose Broker-Dealer license and for testifying in front of Congress earlier this week. Despite raising $50M through a Reg A+ filing, having ex-SEC and FINRA staff on their team, and gaining approval from SEC and FINRA, the firm has yet to launch a product. Their activities, including a fundraising event associated with an entity with a checkered past and hiring ex-SEC and FINRA personnel without a clear product, have sparked suspicion. Their sudden approvals and their listing of tokens in 2022, despite claiming no operational ability in 2023, have added to the suspicion.
EigenLayer has launched restaking on Ethereum's mainnet. This Stage 1 launch allows Ethereum stakers to securely manage their liquid staking tokens. The next phases will introduce Operators for validation tasks and multiple AVSs. Initial restaking limits have been set for safety, supporting three types of staked Ether and a 7-day withdrawal delay. The protocol has undergone rigorous security analyses and is offering a bug bounty to developers that find issues. EigenLayer aims for decentralization with three governance multisigs initially responsible for decision-making. As the protocol matures, the goal is to foster a decentralized, sustainable network.
The Graph, a decentralized protocol that indexes and queries blockchain data, has commenced the last phase of its transition from Ethereum to L2 solution Arbitrum. The move by Graph developers Edge & Node aims to lower barriers to entry for developers by reducing gas costs and speeding up transactions. With this shift, The Graph will now record and settle its own transactions on Arbitrum as opposed to Ethereum, potentially attracting more indexers and delegators due to the reduced costs.
The author of this thread, a full-time FAANG software developer, shares how he stays up to speed on the latest in crypto through various newsletters and social media accounts. He highlights Linn's comprehensive newsletter, Viktor's level-headed market updates, @thedailydegenhq's stats-focused daily updates, PageOne's unique narrative insights, and Chung's varied daily notes.
This blog post discusses the rise of 'digiphysicals', goods which exist physically but are linked to a digital (on-chain) record using NFC technology, in the crypto space. They allow real-world objects to interact with blockchain technology, granting wallet-like abilities and unlocking new consumer experiences. Digiphysicals are poised to revolutionize consumer goods by bringing on-chain interactions to physical environments, giving physical products their own on-chain history, creating new distribution channels for brands, and adding utility for consumers in both digital and physical realms. The post covers the evolution of these products, potential applications, and future predictions.
The Federal Reserve has decided to pause the interest rate hikes that had been ongoing for 15 months, keeping the current federal funds rate between 5% and 5.25%. Most policymakers believe further tightening is required, with projections suggesting two additional quarter-point rate hikes or one half-point increase by the year's end. Fed Chair Jerome Powell signaled that rates might rise "somewhat further" in 2023 to control inflation, although he refrained from specifying when the next hike could be. Despite this, inflation remains high and above the 2% target set by the Fed.
Floating Point Group (FPG), a crypto-focused institutional trading desk, was hit by a cyber attack, leading to a loss of approximately $15-20 million in crypto. The firm was successfully targeted despite having implemented cybersecurity measures, including obtaining a SOC 2 certification. Post-attack, FPG locked all third-party accounts, secured wallets, and temporarily halted trading, deposits, and withdrawals. FPG is now collaborating with the FBI, the Department of Homeland Security, regulators, and Chainalysis to investigate the breach and work towards asset recovery.
Someone borrowed $14,500 against two Rolexes using a novel blend of on-chain tech with Arcade.xyz and 4K Protocol. This is an interesting solution to create greater liquidity and less friction in markets with highly valuable physical collectibles or real world assets.
Apple rejected the latest version of non-custodial Lightning-enabled bitcoin wallet Zeus for violating its cryptocurrency transmission rules, following a similar conflict with Damus, another bitcoin-related app.
Amid significant regulatory pressures and crypto market downturn, crypto-based businesses are grappling with the decision to pivot, double down, or shut down. This entrepreneur gives some advice for other startup founders to follow during the bear market.
The new GPT API features function calling, making it easier to connect with external tools and APIs. It can output JSON instead of just strings when making function calls. Function calling basically gives developers lower-level access to the large language model, making developing on top of the GPT API easier. This article goes through an example of using the API properly and the implications of this new feature.
Netflix migrated their mobile apps to GraphQL in 2022. This was done with zero downtime and involved a total overhaul from the client to the API layer. Doing this safely for 100s of millions of customers without disruption was exceptionally challenging, especially considering the many dimensions of change involved. This blog post shares the broadly-applicable techniques (beyond GraphQL) Netflix used to perform this migration. The three strategies discussed are AB Testing, Replay Testing, and Sticky Canaries. Netflix first created a new GraphQL service over the previous legacy monolith API endpoints. It then deprecated both the service and legacy API in favor of GraphQL services owned by the domain teams.
Using HTTP to query for a search works, but it's not the most optimal UX if your query takes a long time. Users have to wait for the entire search to complete, which can take up to 30 seconds for sparse queries that are not requested often. A better UX would be to stream responses as you query, which can be accomplished with Server-sent events (SSE). This article explains how SSE works, how it differs from traditional web sockets, and how you can integrate it in Express.js.
The obvious benefit to working quickly is that you’ll finish more stuff per unit time. But there’s more to it than that. If you work quickly, the cost of doing something new will seem lower in your mind, so you’ll be inclined to do more. Google famously prioritized speed as a feature. They realized that if search is fast, you’re more likely to search. The reason is that it encourages you to try stuff, get feedback, and try again. Working fast ends up implicitly lowering the activation energy to do new tasks, which means you get more done and try more things.
This retrospective post covers things one developer learned while working on large TypeScript and React codebases. It recommends best practices to replicate and pitfalls to avoid.
Reactivity is an important feature of front end frameworks. There are a number of ways to provide this functionality. This post summarizes the three ways frameworks today provide reactivity: value-based (think states in React), observables-based (through subscriptions), and signal-based (which is close to observables-based, but without the need to subscribe/unsubscribe). It discusses the pros and cons of each reactivity model and a hot take on why signals are the best model out of the three.
Zap is a zig framework to create REST APIs quickly. It can be considered to be a microframework for web applications, similar to Flask for Python. Zap comes with multiple features, such as routes, endpoints, cookie support, middleware, and more.
Rewind-UI is a React component library for building modern web applications using Tailwind CSS. It offers a range of accessible, highly customizable components that can be easily integrated into any React project. Rewind-UI comes with a set of default styles that can be easily customized using Tailwind CSS classes. Furthermore, each component can be parameterized using a set of props that can be used to change the default styles of the component.
Most people dream of building great projects, but many never start. Of the few that start, even fewer finish. This guide breaks down how to plan successfully and overcome roadblocks to become one of the few that finishes the projects they start. Finishing requires effort and courage.