Search the archive

A MANAGEMENT OVERHAUL

LANDING

THIS STARTUP WANTS TO CREATE A PENICILLIN FOR VIRUSES

PERSEUS

Perseus is a frontend web development framework. It supports generating page state at build-time, request-time, incrementally, or at any other time. Perseus aims to support every major rendering strategy to provide developers with the ability to efficiently create super-fast apps.

CRONER

Croner is Cron for JavaScript and TypeScript. It lets users trigger functions, evaluate cron expressions, and obtain a list of upcoming run times without any dependencies. Croner operates in-memory, requiring no configuration files or databases. It supports Node.js, Deno, and Bun. Croner features built-in overrun protection and error handling. Users can pause, resume, or stop execution after a task is scheduled.

METAVERSE

Programmers capable of building out virtual reality games, apps, and technology earned total compensation packages from $600,000 to up to $1 million from Meta. The company is paying much more than other gaming companies. It is also trying to reign in costs while investing in the metaverse. Meta's high salaries are part of a larger trend where the company is willing to outspend competitors to lure in the best talent. It will likely be years before the company's metaverse ambitions come to fruition.

INSIDE GOOGLE'S PLAN TO KILL THE COOKIE

PL/RUST

PL/Rust is a loadable procedural language that enables writing PostgreSQL functions in the Rust programming language.

WE’RE ONE STEP CLOSER TO READING AN OCTOPUS’S MIND

IS THE EMPLOYER-EMPLOYEE SOCIAL CONTACT IN TECH ... NOW BROKEN?

MINUTES

TABBY

Tabby is a self-hosted AI coding assistant that can be used as an alternative to GitHub Copilot.

CANCER AND HEART DISEASE VACCINES ‘READY BY END OF THE DECADE’

GOOGLE CEO SUNDAR PICHAI SAYS SEARCH TO INCLUDE CHAT AI

Google plans to add conversational artificial-intelligence features to its flagship search engine, Chief Executive Officer Sundar Pichai said, as it deals with pressure from chatbots such as ChatGPT and wider business issues. Advances in AI would supercharge Google’s ability to answer an array of search queries, Mr. Pichai said in an interview with The Wall Street Journal. He dismissed the notion that chatbots posed a threat to Google’s search business, which accounts for more than half of revenue at parent Alphabet Inc.

GPT-4 IS A REASONING ENGINE

SEGMENT - DETECT - IN PAINT

A RECIPE FOR TRAINING LARGE LANGUAGE MODELS

ML PAPERS OF THE WEEK

Another great list of papers to refer to regularly. There is lots of hype circulating online at the moment and so it is good to get back to the source sometimes and understand the trends. There are many lists like this one, but it seems high quality and thoughtfully curated.

FULL PROOF GENERATION WITH PROOF REPAIR

One common problem with generative language models is their propensity to hallucinate. Folks have pointed out that sometimes code these models generate is subtly wrong, but looks plausible. The act of verifying is sometimes more important than generating in this case. Verification is hard though, especially for code. This paper studies proof generation which can be checked by a formal system. There are lots of great insights in this work, my favorite is that when given error messages, the system can prove 1.5% more proofs.

AUTOMATICALLY GENERATE LANGCHAIN TOOLS OR CHATGPT PLUGINS

[/%E2%80%8B%E2%80%8Bhttps://github.com/hey-pal/toolkit-ai?utm_source=tldrai]

LLAMA.CPP PROJECT REVOLUTIONIZES META'S LANGUAGE MODEL ACCESS

Georgi Gerganov's llama.cpp project optimizes LLaMA performance, boasting 100x faster load times and halved memory usage, enhancing the user experience on edge and personal computer devices.

INST-INPAINT: INSTRUCTING TO REMOVE OBJECTS WITH DIFFUSION MODELS

Introducing Inst-Inpaint, a novel inpainting framework that seamlessly removes objects from images based on natural language input, eliminating the need for time-consuming binary masks. Using the GQA-Inpaint dataset, the framework outperforms GAN and diffusion-based baselines, demonstrating quantitative and qualitative improvements.

PRETRAINING GAUSSIAN PROCESSES

IN AI RACE, MICROSOFT AND GOOGLE CHOOSE SPEED OVER CAUTION

STABILITY AI IS ON SHAKY GROUND

PYTHON OPTIMAL TRANSPORT UPDATE

Optimal Transport theory is a field with a rich history that has fully embraced computation. One of the standard frameworks has received an upgrade which improves speed, adds examples, and increases general reliability. If you want to compare two distributions, this is likely a good place to start.

SAVE MONEY BY REDUCING YOUR GPT-4 PROMPT

[/%E2%80%8B%E2%80%8Bhttps://www.promptreducer.com/?utm_source=tldrai]

AI SHELL

AI Shell is a CLI tool that converts natural language to shell commands.

TABBY

Tabby is an open-source alternative to GitHub Copilot.

JAPANESE POLITICAL FACTION SIGNALS POTENTIAL THAW IN CRYPTO WINTER

SUSHISWAP HACKED, HEAD CHEF SAYS 'REVOKE ALL CHAINS'

REVENUE FOR CHAINLINK

AEVO LIVE ON MAINNET

USING LAYERZERO TO TRANSFER NFTS BETWEEN CHAINS

ENDGAME PERILS OF RESTAKING

HOW BANKS ARE EYEING DEPOSIT TOKENS FOR B2B PAYMENTS

FARMING

ARMSTRONG

MILLION

HOW SHAREHOLDERS ARE SELLING OPENSEA STOCK DESPITE RESTRICTIONS

MAKERDAO’S ENDGAME IS MAKING SOME DELEGATES UNEASY

CHROMA’S $18M SEED FUNDING

Another big funding announcement for an open source tool for AI. This time the fast growing embeddings database company raises cash. They’re hoping that folks will use their tools to add state and memory to different LLM applications. The goal is for this to reduce hallucinations and increase factuality. They’re growing quickly, it’ll be interesting to see how all the different LLMOps tools come together.

FIXIE RAISES $17M SEED FUNDING

HOW DOES TRAINING A MODEL WITH RLHF ACTUALLY WORK?

Training a model on an unsupervised dataset is so last year. Using human feedback data is all the rage and while it sounds easy in theory - it turns out to be tricky in practice. The HuggingFace folks have written another great blog showcasing StackLLaMA which answers StackExchange-like questions.

VISION TRANSFORMERS WITH MIXED-RESOLUTION TOKENIZATION

The paper proposes a new approach to Vision Transformers by introducing a mixed-resolution sequence of tokens to replace the standard uniform grid, where each token represents a patch of arbitrary size. This approach uses the Quadtree algorithm and a novel saliency scorer to construct a patch mosaic that processes important image regions with higher resolution while routing low-saliency areas in low resolution, resulting in significant accuracy gains on image classification while controlling for the computational budget.

VideoCrafter is an open-source video generation and editing toolbox for crafting video content, which contains three models: Generic Text-to-video Generation, Personalized Text-to-Video Generation with LoRA, and Video Generation with More Condition Controls

TM2D: BIMODALITY DRIVEN 3D DANCE GENERATION VIA MUSIC-TEXT INTEGRATION

The paper proposes a new task that utilizes both music and text for 3D dance generation, which can guide humans to perform richer movements. To tackle the challenge of the existing datasets containing paired motion with a single modality, the authors utilize a 3D human motion VQ-VAE to project the motions of the two datasets into a latent space, and propose a cross-modal transformer architecture to integrate text instructions. The approach generates realistic and coherent dance motion conditioned on both music and text, while keeping comparable performance conditioned on two single modalities.

E2B

E2b is an open-source IDE powered by AI agents that do the work for you.

SAMSUNG WORKERS MADE A MAJOR ERROR BY USING CHATGPT