Odysseus landed successfully on the Moon on 22 February. Built by Intuitive Machines, Odysseus is the first private lunar lander and the first US lunar lander since 1972. It is currently around 300 kilometers from the lunar south pole, an area that may contain ice. The lander experienced a malfunction that required a software patch hours before landing. It will collect data for up to seven days until night falls at the landing site.
BlueSky (the company) is using BlueSky (the microblogging application) to prove the viability of the Authenticated Transfer Protocol (atproto), a federated protocol for large-scale distributed social applications. atproto can't be controlled by any one organization due to its federated nature - various parts of the system that communicate with each other can have multiple people running them. This article takes a close look at how atproto works, explaining its structure, how the system is supposed to work, potential issues, and more.
Enhance your team’s AI security expertise with Snyk’s Buyer’s Guide for Generative AI Code Security [9]🛡️. Learn how to choose the right tools and educate your team effectively 🔥. Secure AI-generated code without disrupting development workflows 🎯.
Functional UI Kit is a design system that focuses on accessibility, development experience, and unified designer-developer experience. It uses Figma variables and CSS variables that share the same names, usage, and inheritance structure. Each Figma variable has a direct counterpart in CSS so there's no confusion. Functional UI Kit is packed with best practices, unified prop names, and consistent design and code.
React Strict DOM aims to improve and standardize the development of styled React components for web and native. It integrates React DOM and StyleX to improve the speed and efficiency of React development without compromising performance, reliability, or quality. The tool is used at Meta to ship features faster to more platforms with fewer engineers.
This post summarizes a developer's learnings from three years of running a SaaS app. In three years, their product had grown from a toy to something people were paying for. The developer learned to stop being afraid of success and calmed down on their software releases and content marketing. They experimented with onboarding, which massively paid off, and started to understand their own product philosophy. The developer learned that, once their product became sustainable, they only had to focus on quality.
Apple introduced the Dynamic Island with its iPhone 14 Pro models in 2022. The feature makes the space surrounding the front camera and Face ID sensors more useful compared to the notch in previous models. Apple explored a variety of ideas for the feature before arriving at its current design. This article looks at some of the ideas Apple considered and presents mock-ups of what they might have looked like.
For the next 12 hours, try Dollar Flight Club for just $1. [14] Fly to Hawaii from $109. Greece from $247. Or Mexico from $137. The $1 deal is only here for 12 hours. Join 1 million members [14] for just $1 today.
The JSON Lines text format, also called newline-delimited JSON, is a format for storing structured data that may be processed one record at a time that works well with Unix-style text processing tools and shell pipelines.
Meta's new test generator recommends fully-formed software improvements that are verified to be both correct and an improvement to current code coverage.
Testing LLMs is not simple. Probabilistic output makes failures hard to identify, while running the models repeatedly tends to become very expensive quickly. In this blog post [5], QA Wolf engineer John Gluck covers 6 things the team learned about building automated black-box regression tests for genAI applications [5], including:
Google has acknowledged an issue with its AI model Gemini. The model injected inappropriate diversity into historical images, reflecting problems with bias in training data. The flaw sparked debate around diversity, equity, and inclusion in tech. Google implied that it will make improvements in the future but stopped short of a full apology for failing to properly contextualize historical figures in image generation.
OpenAI is working on a web search to compete more directly with Google. It is unclear if the product will be standalone or part of ChatGPT. Competition in the search space is filling up quickly with the addition of Copilot on Bing, newcomers like Perplexity, and Google's Gemini Copilot. A YouTube Short featuring Microsoft CEO Satya Nadella talking about competing with Google is available in the article.
REINFORCE is a simple, standard, and easily understood RL method. It is hard to train stably when used in simulators. PPO is much more performant and stable in general. Gemini uses REINFORCE and GPT-4 is believed to use PPO.
AlphaFold is used to predict the state of a protein after folding. By adding flow matching, which is invertible, you can dramatically improve modeling power on the entire landscape of proteins.
Researchers have developed a new way to make LLMs more efficient and easier to use by employing a method that focuses on 'expert-level sparsification', which reduces model size without losing performance. This is particularly useful for Mixture-of-Experts LLMs, which are powerful but usually too big to handle easily.
How do you transition from experimental models to highly scalable applications? apply() is a free community conference [12], where you can learn from seasoned engineers and technical leaders. Don’t miss out on the chance to hone your AI/ML skills in these FREE practical
GeneOH Diffusion is a new technique that improves how models understand and interact with objects using hands. This method focuses on making these interactions more natural by correcting errors in hand movements and relations with objects.
A model based on CodeLlama and DeepSeek Coder was able to get 85%+ on the HumanEval benchmark for programming by training on a synthetic multi-turn dataset and using human feedback.
Anthropic's research scientists have been working on a method of understanding deep neural networks that uses Circuits. These Circuits aim to identify subparts of models that get used for certain tasks. The research team has released a monthly update on the experiments they attempted and the results.
INSTRUCTIR is a new benchmark aimed at making search engines smarter in understanding users' intentions. Unlike current methods, which mostly focus on the query itself, INSTRUCTIR evaluates how well search engines can follow user instructions and adapt to various and changing search needs.
Join this Kolena webinar [19] to explore unique challenges in computer vision - including precise object detection and real-time decision-making - through the lens of autonomous vehicle systems. Join
Sam Altman's request for $7 trillion aims to support the rapidly escalating costs of advancing generative AI models like GPT, suggesting an exponential growth in resource needs for future iterations. This ambition underscores a pivotal moment in AI development, balancing between rapid technological progress and the broader implications of such swift advancement on safety and societal readiness.
Carson Group, a $30 billion investment advisory platform, is set to offer four spot Bitcoin ETFs to its clients, greenlighting options from BlackRock, Fidelity, Bitwise, and Franklin Templeton. It based its selections on AUM, trading volume, and cost-effectiveness. The move puts Carson Group at the front of the pack amongst RIAs approving spot Bitcoin ETFs, marking a contrast to firms like Vanguard, which blocked clients from buying these SEC-approved ETFs.
Citibank, in partnership with WisdomTree and Wellington Capital, has successfully simulated the tokenization of a private equity fund using blockchain technology. The process proves that tokenization is possible in a controlled environment, can remain legally compliant, and is compatible with existing banking systems. The move could potentially introduce efficiencies into the $6 trillion private equity industry. The success of this test will inform Citibank's decisions about offering related services.
Parcel is set to launch its token PRCL in April, to further decentralize its ecosystem. Token holders will participate in governance and have improved data and trading utility. Initially, 7-8% of the total one billion tokens will be in circulation. PRCL will give access to high-grade real estate data and future incentive programs. Eligibility for the initial community airdrop can be verified on Parcl's website.
Uniswap put out a proposal to modify the governance of its DEX, including adjusting the fee structure to benefit users who delegate and stake their UNI tokens. The proposal will be put to a vote on March 1, followed by an on-chain vote on March 8. This was a huge shift for Uniswap and for governance tokens in general, as very few tokens actually allow holders to participate in the revenues of the protocol. Many are speculating that this indicates positive regulatory progress in the background and could lead to more protocols choosing to include token holders in fee-sharing mechanisms.
For the next 12 hours, try Dollar Flight Club for just $1. [9] Fly to Hawaii from $109. Greece from $247. Or Mexico from $137. All on the best airlines. The $1 deal is only here for 12 hours. Join 1 million
This thread explains how to create an income flywheel using perpetual DEXs. It is possible to go both long and short on two different DEXs, leveraging the funding rates by emulating the strategy that the newly launched stablecoin protocol Ethena is implementing. The key is to pick exchanges that are profitable, offer good tokenomics, and provide good delta through arbitrage. Currently, Vertex and HMX are the author’s top picks owing to their stable income, continuous tech development, and beneficial tokenomics.
This post provides a detailed analysis of the various trends, strategies, and factors that influenced the success of different cryptocurrencies during the 2021 bull run. Ideas like investment timing, understanding market rotations, identifying emerging narratives, recognizing high-yield scenarios, and main character phenomena all played important roles last cycle. By studying these patterns, traders can better prepare for future cryptocurrency bull runs.
2024.02.26·
cryptonarratives.substack.com·18 min read·TLDR Crypto
This is a comprehensive overview of projects and potential use cases at the intersection of AI and cryptocurrency. It demonstrates the broad applicability and potential of these technologies across various domains. Potential projects range from enhancing DeFi experiences to revolutionizing game development, optimizing content creation, improving data privacy, and expanding blockchain capabilities through AI-driven analysis and secure computational networks.
The combination of Zero Knowledge Proofs with machine learning, known as ZkML, has the potential to become one of the next big narratives. ZKPs allow those building machine learning models to conduct inferences without exposing the data used but still ensuring computational correctness. They also help tackle privacy concerns like membership inference attacks and model inversion attacks. ZkML infrastructure is already being developed by several notable startups in the space, which could give rise to a hype cycle for this tech.
Avalanche experienced a six-hour outage on Friday due to a bug triggered by high demand during an inscription mint. The bug was fixed and the network has been operational without delay since.
This thread from a notable crypto VC provides a perspective on where he thinks crypto venture is in this cycle. He notes that deals are getting more and more competitive, especially on the infrastructure side, and that some FOMO is even setting in.