Search the archive

REDIS ADOPTS DUAL SOURCE-AVAILABLE LICENSING

All future versions of Redis will be released with source-available licenses starting with Redis 7.4. Redis will no longer be distributed under the three-clause Berkeley Software Distribution and will be instead dual-licensed under the Redis Source Available License and Server Side Public License. The new licenses enable Redis to sustainably provide permissive use of its source code, allowing it to continue to be freely available to developers, customers, and partners through Redis Community Edition as Redis moves on to its next phase of development as a real-time data platform with a unified set of clients, tools, and core product offerings.

DOCUSIGN'S 7,000 EMPLOYEES AND THE IRON LAW OF B2B SAAS SALES

Docusign's over 7,000 employees are mostly in sales. Around 24% of its workforce is in engineering, product development, and customer operations. The company's sales/engineering split isn't that unusual. Even if a company is well positioned in a space, its future and profits are always uncertain - putting resources into new products or features just in case makes sense as the cost of being overtaken is very high.

BRINGING AI TO THE MASSES WITH ADAM D'ANGELO OF QUORA

In this interview, CEO of Quora Adam D'Angelo talks about how building infrastructure for creators can democratize AI. The interview covers how social networks are a compliment to AI, multi-model and multi-modal AI, bringing AI to the masses, AI infrastructure for creators, and much more. The video of the interview is available.

NVIDIA ANNOUNCES “MOONSHOT” TO CREATE EMBODIED HUMAN-LEVEL AI IN ROBOT FORM

Nvidia's Project GR00T aims to create a general-purpose foundation model for humanoid robots to enable robots to learn skills and solve various tasks on the fly.

APPLE FACING IMMINENT U.S. ANTITRUST LAWSUIT

The US Justice Department is preparing to sue Apple for violating antitrust law.

PGWIRE

pgwire implements PostgreSQL Wire Protocol and provides APIs for building Postgres-compatible access layers for data services.

SO YOU THINK YOU WANT TO WRITE A DETERMINISTIC HYPERVISOR?

Determinism, which is when machines always pass through the same sequence of states to produce the same output when given a particular input, produces consistent and knowable results.

STABILITY AI IS FALLING APART

Several key developers at Stability AI, creators of the popular text-to-image generator Stable Diffusion, have resigned.

ELON MUSK'S SPACEX WANTS ITS STARSHIP BACK IN THE AIR FASTER THAN EVER

SpaceX aims to launch the next Starship as soon as early May.

Skeptical? Try calling it yourself >> Bland.ai

(P.S. TLDR readers can sign up here and access something even crazier... [5])

STABLE DIFFUSION MAKER LEAVES STABILITY AI

Stability AI's research scientist Robin Rombach, crucial to developing the Stable Diffusion model, is leaving the company, marking a significant departure amidst a year of technical team changes.

INTRODUCING COPILOT4D: A FOUNDATION MODEL FOR SELF-DRIVING

Waabi's Copilot4D is a pioneering foundation model that leverages LiDAR data to understand and predict the 3D dynamics of the environment over time, advancing the capabilities of autonomous machines.

NLX RAISES $15M SERIES A

NLX, an enterprise conversational AI platform, has raised additional funding from Cercano, Comcast, and others. The platform is used to build chat, voice, video, and conversational systems.

DATA AUGMENTATION WITH DIFFUSION MODELS

DreamDA offers a new approach to data augmentation, utilizing diffusion models to synthesize diverse, high-quality images that closely match the original data distribution.

VISION-LANGUAGE MODELS WITH INTERACTIVE REASONING

Chain-of-Spot (CoS) introduces an Interactive Reasoning technique that significantly enhances how Large Vision-Language Models (LVLMs) process and understand images. By focusing on key regions of interest within images in response to specific questions or instructions, CoS enables LVLMs to access detailed visual information without compromising image resolution.

ENHANCING VIRTUAL TRY-ON WITH PRE-TRAINED DIFFUSION MODELS

StableVITON is a novel approach to image-based virtual try-on. This method focuses on maintaining clothing details while leveraging the generative power of pre-trained diffusion models. StableVITON learns semantic correspondences between clothes and the human body in a pre-trained model's latent space.

Phospo in 60 seconds [13] or try Phospo Pro

TRITON PUZZLES

Triton is a library for writing CUDA kernels in a Python-like way. It is gaining popularity. This repository has a set of puzzles in increasing difficulty that encourage learning the tool.

DIFFUSION-BASED VIDEO TRANSLATION

FRESCO is a novel approach that combines intra-frame and inter-frame correspondences to significantly improve the spatial-temporal consistency in video translation tasks.

ENHANCED IMAGE EDITING WITH CONSISTENCY MODELS

This project enhances the capabilities of diffusion models for tasks like image editing and restoration by introducing Generalized Consistency Trajectory Models (GCTMs). These models streamline the process, making it possible to modify images with remarkable precision and efficiency by translating between any two distributions with just one step.

NEW BREAKTHROUGH BRINGS MATRIX MULTIPLICATION CLOSER TO IDEAL

Tsinghua University and UC Berkeley researchers have achieved a significant breakthrough in matrix multiplication, presenting an innovative technique that has already spurred further enhancements. This advancement in a core computational operation could lead to substantial time, power, and cost savings across various applications. This represents the most considerable progress in reducing the computational complexity of matrix multiplication since the previous milestone in 2010.

STYLIZED IMAGE BINNING ALGORITHM

This is a tutorial on creating a pixel-art-like image processing tool with interactive web elements like sliders for customization using a binning algorithm in JavaScript. The binning technique uses parameters such as bin size and gap to convert images into stylized, pixelated artwork by averaging pixel brightness within bins. The implementation involves manipulating pixel data on HTML canvas elements and optimizing looping structures for efficiency.

USING LLMS TO GENERATE FUZZ GENERATORS

LLMs like Claude can generate effective fuzzers for parsing code, automating a process that traditionally requires significant human effort. While LLMs are typically not precise enough for static analysis, they appear well-suited for creating fuzzers due to the stochastic nature of fuzzing. A hybrid approach that combines LLM-driven static analysis and targeted fuzzing could be promising for identifying and exploiting vulnerabilities in code.

model reliability and performance. Register ↗️

JARS AI LAUNCHES PLATFORM FOR INTERACTIVE AI SHOWS (PRODUCT)

JARS allows anyone to create episodes of their favorite shows with friends.

OPENAI COULD RELEASE GPT-5 IN A FEW MONTHS

OpenAI could release GPT-5 in the summer.

BEIJING COURT RULES AI-GENERATED CONTENT COVERED BY COPYRIGHT, ESCHEWS US STAND

Beijing Internet Court ruled that an AI-created image was copyright-protected artwork due to the human creator's intellectual input, awarding damages in a landmark case with significant implications for the AI industry and copyright disputes worldwide.

Start building on Hedera

ETHEREUM FOUNDATION REMOVES WARRANT CANARY AFTER ‘VOLUNTARY ENQUIRY FROM A STATE AUTHORITY’

The Ethereum Foundation has reportedly received a confidential inquiry from a state authority, resulting in the removal of its warrant canary from the organization's website. The Foundation, a non-profit that supports Ethereum without directly controlling it, has mistakenly removed the warrant canary once before in 2019. Ethereum showed little reaction in price to the news. Experts suggest such inquiries are fairly common in the crypto industry.

COINBASE FILES TO LIST DOGE, LTC, AND BCH FUTURES TRADING

Coinbase has applied to the CFTC to list Litecoin, Bitcoin Cash, and Dogecoin futures trading. Should the CFTC give the green light, these digital asset products could be available for trading as early as April 1. The move marks a significant step towards further legitimacy for these cryptocurrencies.

ROBINHOOD ROLLS OUT CRYPTO WALLET TO ANDROID USERS GLOBALLY

Robinhood has officially released its crypto wallet app for Android users worldwide. The move follows the company's successful debut of the wallet to iOS users last year. This comes as Robinhood saw a 10% YoY increase in its crypto business and an 89% uptick in its notional crypto trading volume.

STARKNET, AN ETHEREUM LAYER 2, PLANS 'PARALLEL EXECUTION' TO MIMIC SOLANA'S SPEED FEATURE

Starknet, an Ethereum L2, is planning to add a "parallelization" feature as part of a Q2 upgrade. The addition aims to increase transaction processing efficiency, resulting in better throughput and transaction finality. This comes as a part of Starknet's 2024 roadmap, which includes steps to curtail fees. Notably, Starknet is the largest network in the ZK rollups category of L2 networks, which are designed for quicker settlements based on zero-knowledge cryptography.

THE UNDERVALUED LIDO OF BITCOIN

Activity in the Bitcoin ecosystem has picked up since the announcements and launches of the spot Bitcoin ETFs. Projects such as Babylon Chain are innovating, creating a way to generate additional yield for Bitcoin holdings through staking and validating Proof of Stake chains. Meanwhile, new and innovative projects like Ankr are entering the market with offerings like Liquid Staked Bitcoin.

BULLISH BRAINDUMP ON BASE

This thread explores the growth and potential of the Ethereum L2 Base. It asserts that Base differentiates itself as a 'builder chain'. While Base was relatively slow to gain traction, the current uptick in speculative activity, especially in memecoins, interest in the ecosystem has been increasing significantly. The affordable gas fees and increasing user base present ample opportunities for developers to build applications.

ON THE ETH MISINFORMATION

Since the SEC allowed the ETH Futures ETFs to trade on its regulated security exchanges, it implicitly agreed ETH is a non-security and not under SEC jurisdiction. This move, made after Ethereum transitioned to a Proof of Stake model in 2022, suggests any delay or denial towards an ETH ETF would be contradictory to the agency’s past actions, as the SEC has indicated Ethereum is beyond its jurisdiction.

FED CHAIR POWELL: NO ’SECRET LAB’ BREWING A DIGITAL DOLLAR SURPRISE

Federal Reserve Chair Jerome Powell has dismissed rumors that the bank is secretly developing a CBDC, stating that no decision has been made on whether the US should proceed with such a project. However, Powell confirmed the bank is closely monitoring developments in the digital asset space.

TON FOUNDATION LAUNCHES TONCOIN COMMUNITY REWARDS INITIATIVE WORTH OVER $100 MILLION

The TON Foundation launched The Open League, a $115+ million Toncoin rewards initiative for ecosystem growth featuring competitions, airdrops, and more starting in April.

MEMECOIN MANIA HITS SOLANA

Solana's rise in popularity, driven by memecoins, surging metrics, and excitement around upcoming airdrops and ecosystem growth, is causing network congestion.

NON-FINANCIAL APPS ON BASE

This thread lists some of Jesse Pollak’s (the creator of Base) favorite non-financial apps on the chain, which include projects such as Blackbird and Parallel.

ANOTHER CRYPTOPUNKS NFT JUST SOLD FOR $16 MILLION IN ETHEREUM

An alien CryptoPunks NFT sold for $16 million, marking the second-largest sale in the collection’s history.

APPLE IPHONE 16: EVERYTHING WE KNOW, FROM DESIGN TO AI FEATURES

The iPhone 15 lineup brought significant upgrades, with the Pro model featuring a titanium body and a customizable Action Button, while the standard version included features like Dynamic Island. Expectations are high for the iPhone 16, rumored to have larger screen sizes and potential AI enhancements. The release is anticipated for September, with prices likely to remain similar to the iPhone 15's.

ADOBE SUBSTANCE 3D’S AI FEATURES CAN TURN TEXT INTO BACKGROUNDS AND TEXTURES

Adobe has introduced two new integrations for its Substance 3D design software suite that allow 3D artists to create assets quickly for their projects using text descriptions. Using the Text to Texture feature in Substance 3D Sampler, designers can create photorealistic or stylized textures for 3D models. Another feature is Generative Backgrounds for Substance 3D Stager, which lets designers use text prompts to create background images for objects they compose into 3D scenes. Text to Texture and Generative Background are free during beta and available in Substance 3D Sampler 4.4 and Stager 3.0 beta versions.

SPACE FORCE DOESN'T WANT ANYONE TO SEE ITS SECRET LOGO MOCKUPS

The U.S. Space Force logo has been criticized for its obvious Trekkie influence. If you've wondered what other logo designs were considered, you will see them sometime soon. As a tardy response to Reason's Freedom of Information Act (FOIA) request, the Air Force redacted all images of alternate logo designs that didn't make the cut.

WILL AI DEVICES KILL UX?

AI will require UI/UX to be simple enough to support the navigational UI of devices like the AI pin and Rabbit R1. The coming years will probably be about very focused, value-oriented apps that can be integrated into broader ecosystems. In the future, marketplaces will be driven by autonomous agents that find the best solution for unique use cases. Focus and simplicity will be essential.

CREATING A UX STRATEGY

Approximately 50% of engineering time is devoted to reworking. Building a solid UX strategy as soon as possible will save you a lot of trouble and time. Having a baseline set of goals and guidelines will keep you and your team from getting confused while doing the best work possible.