Search the archive

IPHONE 15 DESIGN SHOWN OFF IN LEAKED IMAGES SHARED BY UNKNOWNZ21

YEARS

SCIENTISTS GROWING SKIN THAT CAN BE SLIPPED ON LIKE A GLOVE

SCIENTISTS GROW ELECTRODES IN BRAIN, THANKS TO A SIMPLE VISCOUS GEL

XC

xc is a task runner that simplifies the process of creating executable documentation. It allows users to define tasks using human-readable Markdown, meaning that even those without the xc tool installed can access useful commands via the README.md file. xc also provides editor tools for VSCode and Vim.

SONNER

Sonner is a toast component for React. It is styled by default and includes an animation for swiping to dismiss messages. Sonner allows users to easily customize their toast messages.

TECH LAYOFFS ARE FEEDING A NEW STARTUP SURGE

The tech industry has seen a significant decrease in the number of employees in the past two years, with at least 260,000 workers laid off. Many of these individuals have decided to become entrepreneurs, taking advantage of their severance packages to pursue their own business ventures. Investors view startups as an opportunity as they are agile, cost-efficient, and it can be a sign of success if customers are willing to pay for a new product during a recession. Early indications show that this trend of laid-off workers becoming entrepreneurs is on the rise.

GOOGLE ASKS SOME EMPLOYEES TO SHARE DESKS AMID OFFICE DOWNSIZING

ON-CHAIN

A NEW KIND OF BUG SPELLS TROUBLE FOR IOS AND MACOS SECURITY

INTRODUCING AMBIENT 0.1

HOW I BROKE INTO A BANK ACCOUNT WITH AN AI-GENERATED VOICE

SETTING

Setting is a library that simplifies the process of creating beautiful preference panels.

INSTAGRAM FOUNDERS LAUNCH PERSONALIZED NEWS APP

TOME RAISES $43M SERIES B FOR AI POWERED STORY TELLING

A mix between powerpoint, figma, and docs - Tome is a creative tool powered by generative technologies. It has been growing quickly and looks like a potentially interesting combination of several ideas that have been in the community for the past several months. This brings their total money raised to just over $75m.

APIS ARE THE FUTURE OF CHATBOTS

LATENT VIDEO DIFFUSION MODELS

One of the main bottlenecks of generative video work is the inability of current methods to scale to the massive number of frames or maintain temporal consistency in generation. This work tries to solve both problems at the same time by working in a projective latent space. They obtain impressive FID scores on standard benchmarks and have fun videos on their project page.

DRAMATICALLY IMPROVES QUALITY

By building on the fascinating “score based” interpretation of diffusion models, this work investigates why diffusion models fail at various tasks that require composition. They determine that the sampler, and not the underlying model, must be the source of error. They then propose a new energy-based parameterization and Metropolis-corrected sampler. They show strong improvement on a variety of tough tasks. Also, as a smaller win, they have a nice background section on all these topics!

TASKS

Modular deep learning has emerged as a promising solution for developing models that specialize in multiple tasks without incurring negative interference and that generalize systematically to non-identically distributed tasks. This article provides a survey of modular architectures, explores various additional purposes of modularity, and reports various concrete applications where modularity has been successfully deployed.

TUNING

The article discusses the use of parameter-efficient adaptation methods such as prompt tuning (PT), in-context learning (ICL), and instruction prompt tuning (IPT) for fine-tuning large language models. The paper empirically studies when and how in-context examples improve prompt tuning, and observes that IPT does not always outperform PT, PT is unstable, and prompts learned for a specific source task via PT exhibit positive transfer when paired with in-context examples of a different target task.

PAPERS ON THE UX OF AI PROGRAMMING ASSISTANTS

It could be argued that in addition to the powerful model, co-pilot’s success is because of the innovative user experience. In any case, this page provides a list of papers that study the effectiveness of different choices on users of programming assistant tools.

The article presents a method called Vid2Avatar, which learns human avatars from monocular in-the-wild videos without the need for ground truth supervision, external segmentation modules, or priors extracted from large datasets. The method solves the task of surface reconstruction and scene decomposition by jointly modeling the human and background in the scene using two separate neural fields.

MOCHIDIFFUSION

This app brings stable diffusion to Macs. This app uses Apple's Core ML Stable Diffusion implementation to achieve maximum performance and speed on Apple Silicon based Macs while reducing memory requirements. It also runs on Intel based Macs too.

HOW NVIDIA DOMINATED AI - AND PLANS TO KEEP IT THAT WAY

HOW I BROKE INTO A BANK ACCOUNT WITH AN AI-GENERATED VOICE

MIND-CONTROLLED AI ROBODOGS

SHOULD AI CHATBOTS HAVE ROBOT RIGHTS?

AI-CREATED IMAGES LOSE U.S. COPYRIGHTS

The U.S. Copyright Office has ruled that images in a graphic novel created by AI should not have been granted copyright protections, one of the first decisions by a U.S. Court or Agency on the topic of AI art and copyrights.

REACHES 'PROOF-OF CONCEPT' STAGE

TESLA ANNOUNCES NEW ENGINEERING HEADQUARTERS IN CALIFORNIA

GOOGLE CLAIMS BREAKTHROUGH IN QUANTUM COMPUTER ERROR CORRECTION

Google has achieved a breakthrough in correcting errors in quantum computers. Currently, quantum computers struggle to produce results due to their qubits only holding their quantum states for a short time. Information is lost before the machine can complete its calculations. Correcting these errors is the hardest technical challenge the industry faces. Google's researchers recorded a 4% reduction in error rates when scaling up their technique. The research puts Google on the path to its first practical quantum computer. There is a chance that the error-correction technique may not work when applied to larger quantum systems.

FOR MARCH

IASQL

IaSQL provides a 2-way connection between a cloud account and a PostgreSQL database. This connection allows users to treat infrastructure as data in PostgreSQL. A quick demo is available in the repository.

GENERATIVE AI IS COMING FOR THE LAWYERS

THE RADICAL PROMISE OF NUCLEAR FUSION

TO THE PUBLIC WITH NEW FEATURES

LANDSCAPE

The MAD Landscape is an annual report that looks at the technologies and products in the data and AI ecosystem.

HIGHLIGHT.IO

highlight.io is a full-stack monitoring platform that features session replay, error monitoring, and logging.

SCIENTISTS SAY THEY GENE HACKED MICE TO DOUBLE REMAINING LIFESPAN

HUMANS WHO ARE NOT CONCENTRATING ARE NOT GENERAL INTELLIGENCES

INTRODUCTION TO DATA-CENTRIC AI

This is the first-ever course on Data-Centric AI, an emerging science that studies techniques to improve datasets.