Search the archive
Full-text search across article titles and summaries. Refine with sector, year, content type, and reading-time filters.
Results
14,500 matchesHOW SLOW IS SELECT * ?
A technical dive into the not-so-obvious costs of using SELECT *
OPEN SOURCE IS FUELING THE FUTURE OF NUCLEAR PHYSICS
Nuclear physicists have formed an unlikely friendship with open source software.
3 WAYS TO SET HEADERS WITH ASTRO AND NETLIFY
ELON MUSK FOUNDS NEW AI COMPANY CALLED X.AI
CONSIDERS BING DEFAULT
YOUR ROBOTIC AVATAR IS ALMOST READY
SPACEX WILL TRY TO LAUNCH MOST POWERFUL ROCKET EVER MONDAY
WEB LLM
Web LLM is a project that runs large language models directly in the browser. It runs everything within the browser with no server support. Web LLM aims to provide AI assistants with privacy for everyone. A web demo is available.
WEBGPU FUNDAMENTALS
This site contains a resources for learning WebGPU. WebGPU is an API that allows developers to draw triangles, points, and lines to textures and run computations on the GPU. It is an extremely low-level API that requires a large amount of code and serious organization of data for many apps. The resources assume readers have previous knowledge of JavaScript.
HYPE GROWS OVER “AUTONOMOUS” AI AGENTS THAT LOOP GPT-4 OUTPUTS
FULL-TIME JOBS
BLUESKY IS MY FAVORITE TWITTER CLONE YET
CHATGPT IS GOING TO CHANGE EDUCATION, NOT DESTROY IT
Many teachers believe that chatbots can be used to make lessons more interactive, teach students media literacy, generate personalized lesson plans, and save teachers time on admin.
DBDEV
dbdev is a package manager for PostgreSQL trusted language extensions.
APPLE PLANS TO LAUNCH MORE THAN JUST ITS NEW HEADSET AT WWDC
In addition to a mixed reality headset, Apple's annual conference will also feature new Mac laptops, an update to the Apple Watch's software, and updated versions of iOS, iPadOS, and macOS.
EFFICIENCY TRADES OFF AGAINST RESILIENCY
INNOVATION IS OVERRATED
PRINCETON LAUNCHES CENTER FOR LANGUAGE AND INTELLIGENCE
Recent progress in large language models has been dominated by large industry players. However, many of the key ideas still come from academia. Princeton is hoping to continue contributions to language and intelligence with this new center. They’re hiring for research scientists, engineers, and postdoc scholars.
WORLD'S SIMPLEST TEXT-TO-IMAGE MODEL?
Laion just released the Paella image generation model. It uses a discrete, token based, latent space and a cross entropy denoising objective. One goal of this project is to make an image synthetics model that almost anyone can understand. Their sampling code is only 12 lines long!
GOOGLE RUSHING TO ADD AI TO SEARCH ENGINE
SCALING, EMERGENCE, AND REASONING IN LLMS
SO LONG HYPERPARAMETERS
What if we didn’t have to tune the pile of hyperparameters we do to make deep learning work? This is a pretty technical paper that presents Automatic Gradient Descent which is a new optimizer that takes into account network architectures and eliminates a set of previously required hyperparameters. They work through all the math thoroughly, and the resulting optimizer seems to be fairly simple. They don’t have it working yet for all network architectures, but that feels like only a matter of time.
DINOV2: LEARNING ROBUST VISUAL FEATURES WITHOUT SUPERVISION
Recent breakthroughs in natural language processing have paved the way for foundation models in computer vision, producing all-purpose visual features through self-supervised methods and large, diverse datasets. By utilizing an automatic pipeline for curated data and training a 1-billion-parameter ViT model, researchers have created smaller models that surpass existing all-purpose features in most benchmarks at image and pixel levels.
OPENAI CONSISTENCY MODELS CODE
We wrote recently about the new type of genitive model from OpenAI that is faster than diffusion and more stable than a GAN. They just released code and model checkpoints on a set of small experimental data sets. This model is a cool next step in generative image modeling, and now we can play around with it!
TRANSFORMER
Combining spectral and multi-headed attention layers in the novel SpectFormer architecture improves transformer performance for image recognition tasks. This approach achieves state-of-the-art accuracy on ImageNet-1K and consistently strong results in transfer learning, object detection, and instance segmentation tasks across various datasets.
COST TO TRAIN ANTHROPIC'S NEXT MODEL
Anthropic recently announced plans to spend a billion dollars on training a massive language model. They announced this plan before their strategic partnership with Google, which may offer cheaper compute via their TPU pods. In any case it seems like one of the largest expenses of this endeavor is actually AI researcher salaries, followed closely by compute costs.
91% OF ML MODELS DEGRADE OVER TIME
EUROPEAN PARLIAMENT PREPARES TOUGH AI MEASURES
The European parliament is preparing tough new measures over the use of artificial intelligence, including forcing chatbot makers to reveal if they use copyrighted material.
GPT5 IS NOT IN TRAINING YET
GENERATING CODE WITH TESTS
WOLVERINE
Wolverine uses GPT-4 to edit your scripts and tell you what went wrong.
U.S. HOUSE COMMITTEE PUBLISHES DRAFT STABLECOIN BILL
CRYPTOCURRENCY
Bhutan's sovereign investment arm, Druk Holding & Investments, quietly invested millions of dollars in cryptocurrencies like Bitcoin and Ether. This was revealed in the bankruptcy filings of crypto lenders BlockFi and Celsius, both of which Druk was an institutional customer. The fund has been accused of defaulting on a $30 million loan from BlockFi, though Druk’s CEO claims the matter has been settled. Druk's involvement in cryptocurrencies raises questions about the nation's relationship with the volatile crypto market, and it appears to be the first sovereign wealth fund to directly own cryptocurrencies.
SUI ANNOUNCED ITS TOKEN LAUNCH MECHANICS
The crypto ecosystem received long-awaited details on the Sui public launch over the weekend with the announcement of its upcoming public token sale. In conjunction with this, the SUI Token Community Access Program was announced, with plans to reward early Sui network supporters such as testers, community managers, and builders with retrospective awards and recognition sale opportunities. The general public can also participate by purchasing SUI tokens on exchanges like KuCoin and OKX. Instead of an airdrop, SUI tokens are being offered at a discounted price to early contributors and ecosystem participants. This is a different approach than some of the most recent high-profile token launches like ARB and BLUR - both of which conducted airdrops to distribute tokens.
MONEY-MARKET FUNDS
Another stablecoin company, Ondo Finance, has come on the scene. This stablecoin will be backed by money-market funds invested in US government securities. This mechanism, created by former Goldman Sachs employees, allows holders to receive a yield on their stablecoins from the yield generated via the money-market funds backing the coin. Investors will receive a digital coin called OMMF, which will invest only in low-risk money market funds. This aims to provide the price stability and on-chain utility of stablecoins while offering better investor protection and passing on yield to holders. The long-term goal of OMMF is to provide an alternative to bank settlement networks or other stablecoins, as well as to become a preferred collateral asset in crypto lending.
PYRAMIDS OF NFT VALUE
A COMPLETE GUIDE TO PRESALES
SEC’S GENSLER TAKES ON CRYPTO DEFI EXCHANGES WITH REFRESHED RULE PLAN
MOST ACTIVE INVESTORS
WHY CRYPTO TOKENS SHOULD NOT BE CONSIDERED SECURITIES
UNISWAP DAO APPROVES PROPOSAL TO DEPLOY V3 ON POLYGON ZKEVM
LIQUIDITY GAP
BITCOIN THIEF ONCE WORTH $3.4 BILLION GETS A YEAR IN PRISON
After pleading guilty late last year, James Zhong was recently sentenced to one year in prison for stealing 50,000 Bitcoin from Silk Road. Initially worth $600,000, the Bitcoin's value reached $3.35 billion before being seized by authorities.
WEB-LLM: LARGE-LANGUAGE MODELS AND CHAT LOCALLY IN WEB BROWSERS
Web-LLM brings language model chats directly onto web browsers. Everything runs inside the browser with no server support and accelerated with WebGPU. It brings a lot of fun opportunities to build AI assistants for everyone and enables privacy while enjoying GPU acceleration.
ARCHIVE AND MORE
Wayback is a tool that supports running as a command-line tool and docker container, purpose to snapshot web pages to time capsules. It is a self-hosted archiving service integrated with Internet Archive, archive.today, IPFS and beyond.
NEW BEDROCK SERVICE
Sponsored
INFRASTRUCTURE ENGINEERS JUST GOT A COPILOT
WHAT THE HECK IS HAPPENING AT UBER?
The recent Wall Street Journal profile of Uber CEO Dara Khosrowshahi, which featured him driving for Uber and discovering flaws in the product, has been criticized by analysts, investors and ex-employees. They are concerned about the timing of the CEO's revelation, the level of compensation he receives,S and the company's product management.
JUNIOR TO SENIOR: AN ACTION PLAN FOR ENGINEERING CAREER SUCCESS
Making the jump from junior to senior can seem daunting and unclear. This guide will help you to enhance your developer skills and take your career to the next level.