Lip-Bu Tan will become the CEO of Intel on March 18. Former CEO Pat Gelsinger was booted from the position on December 2 after several quarters of losses, rounds of layoffs, and canceled or spun-off side projects. Tan was previously on Intel's board from 2022 to 2024 and has been a board member for several other technology and chip manufacturing companies. Intel reported a loss of $18.76 billion in 2024, its first annual loss since 1986.
2025.03.13·arstechnica.com·3 min read·TLDR Google installed 100 million lithium-ion cells across its data centers to provide backup power to its servers. The lithium-ion cells hold twice the power and take up half as much space as the lead-acid batteries they replaced, allowing the company to cut the number of cells required by three-quarters - more space for servers. Batteries are driving change in a myriad parts of the economy. Panasonic has shipped over 10 billion cells from its Nevada plant since it opened in 2015.
2025.03.13·techcrunch.com·2 min read·TLDR Canadian startup General Fusion has created plasma inside its prototype reactor, a major milestone in proving its technology. The company uses technology that dates back to the 1970s - details are available in the article. The prototype is now forming plasmas daily - the next step is to compress the plasmas with a lithium liner to create fusion and heating from compression. The company hopes it will achieve scientific break-even equivalent by 2026.
2025.03.13·newatlas.com·3 min read·TLDR This post discusses how to make the Cursor experience as good as possible. It covers what to put inside the .cursorrules file, how to get the best output, and more. One of the biggest factors in a developer's experience with coding with AI is going to be what they are coding and which tech stack they are using. One of the best things about AI is that it can help developers tackle problems that would normally be outside of their reach.
2025.03.13·nickcraux.com·4 min read·TLDR The DuckDB team and MotherDuck have released a local UI for DuckDB as part of the ui extension. The DuckDB UI uses interactive notebooks to define SQL scripts and show the results of queries. It can show databases, table summaries, notebooks, column insights, and more.
2025.03.13·duckdb.org·5 min read·TLDR It was clear when Apple announced Apple Intelligence that some of its features were more ambitious than others. Apple hinted strongly that the various features wouldn't all ship at the same time - the more trivial features would ship first. Its more ambitious features have been delayed until sometime 'in the coming year'. Apple appears to have pitched a story that wasn't true and set a course based on that.
2025.03.13·daringfireball.net·23 min read·TLDR The excitement around AI's potential contributions to mathematics doesn't always come with a nuanced understanding of what mathematics truly is and isn't. This post attempts to articulate these differences and hopefully build a bridge for better collaboration. Most mathematicians are not interested in AI creating new mathematics, but in processing and organizing existing knowledge. Even if AI was able to produce a proof of a major conjecture, it would fail to deliver the deeper understanding that mathematicians truly seek.
2025.03.13·links.tldrnewsletter.com·25 min read·TLDR The latest Gemma model is aimed at developers who need to create AI to run in various environments.
2025.03.13·arstechnica.com·3 min read·TLDR FilePizza enables peer-to-peer file transfers inside the browser.
2025.03.13·github.com·GitHub repository·TLDR Manus uses Browser Use to execute various tasks like clicking through site menus and filling out forms.
2025.03.13·techcrunch.com·3 min read·TLDR Inference is a far more diverse workload compared to training - performance is predominantly determined by memory capacity, memory bandwidth, and compute - which of these is prioritized is heavily dependent on a model's architecture, parameter count, hosting location, and target audience.
2025.03.13·theregister.com·5 min read·TLDR You should only train a model on outputs, not on an analysis of its outputs or on the technique used to analyze its outputs.
2025.03.13·links.tldrnewsletter.com·30 min read·TLDR The real value of generative AI will come from agents that can do things for users directly.
2025.03.13·weightythoughts.com·16 min read·TLDR Sponsored
Ai2 is a nonprofit research institute dedicated to developing truly open AI solutions to solve the world's biggest problems. To continue innovating, they needed to scale compute performance and prepare a data center with significant room for growth.
2025.03.13·cirrascale.com·Sponsor·TLDR AI Game Art Forge introduces AI-generated templates to streamline game asset creation, enhancing speed, scalability, and creative control for developers. It allows customization while maintaining consistency and supports both indie developers and larger teams. By combining AI efficiency with human creativity, it ensures high-quality, adaptable workflows for game development.
2025.03.13·genies.com·4 min read·TLDR AI The AI scientist from Sakana Labs has its first reviewed and accepted publication. It generated the idea, conducted experiments, and wrote a novel paper that was accepted to an ICLR workshop (with the full consent of the conference organizers).
2025.03.13·sakana.ai·8 min read·TLDR AI Google has released the weights and a technical report for the Gemma 3 model. This model, which comes in four sizes, is approximately as performant as Gemini 1.5 Pro. It also has a strong understanding of 140+ languages. It seems to be close to the state-of-the-art dense model.
2025.03.13·blog.google·7 min read·TLDR AI Google has released the best dense models available. They score 9th in the arena, have phenomenal multilingual capabilities, and support more than 140 languages. Depending on the upcoming open launches, these may be the backbone of the community. They certainly are the best single GPU models and likely will be for some time.
2025.03.13·storage.googleapis.com·12 min read·TLDR AI TaylorSeer introduces a method to predict future timestep features in diffusion models using Taylor series expansion, significantly reducing errors in feature caching.
2025.03.13·arxiv.org·16 min read·TLDR AI CATANet enhances image super-resolution by aggregating long-range content-similar tokens.
2025.03.13·arxiv.org·12 min read·TLDR AI This project explores how language models perform stepwise implicit reasoning but struggle with generalization, especially when dealing with expressions containing variables as subtrahends.
2025.03.13·github.com·GitHub repository·TLDR AI Motion Anything improves conditional motion generation by introducing Attention-based Mask Modeling for better spatial and temporal control.
2025.03.13·steve-zeyu-zhang.github.io·2 min read·TLDR AI VideoPainter introduces a dual-stream architecture for video inpainting, significantly reducing learning complexity while improving background preservation and object generation. It also provides VPData and VPBench, the largest video inpainting dataset and benchmark to date.
2025.03.13·yxbian23.github.io·1 min read·TLDR AI PromptPex is a developer tool that treats prompts as functions and automatically generates test inputs, enabling systematic unit testing of AI model prompts.
2025.03.13·github.com·GitHub repository·TLDR AI A nice unification and simplification of diffusion models for continuous data. They use a new moment matching framework to create a few step generative model.
2025.03.13·github.com·GitHub repository·TLDR AI The Hugging Face team has announced the next update for its open replication of the DeepSeek reasoning model. It found that small models can outperform the larger ones when tuned specifically for competitive coding.
2025.03.13·huggingface.co·8 min read·TLDR AI Sponsored
Oxford Languages hosted Yasmin Greenaway, founder of Dermie.AI [18], for a discussion of AI in healthcare - from data privacy to the latest
2025.03.13·languages.oup.com·Sponsor·TLDR AI 2025.03.13·languages.oup.com·TLDR AI Wan is a generative image model that has recently been open sourced.
Another LoRA for the Wan video model.
By breaking the reasoning training process into two steps, this work shows that small models are still able to generalize across tasks and even modalities.
2025.03.13·forjadeforest.github.io·6 min read·TLDR AI MGX, an Abu Dhabi-based AI and tech investor, has made a landmark $2B investment in Binance - the largest single institutional investment ever into a crypto company and the largest ever paid in stablecoin. This investment, MGX's first in blockchain, aims to support blockchain innovation and solidify Binance's role in the global digital finance ecosystem.
Russia's central bank has proposed new regulations allowing high net-worth individuals (>$11.5M assets) and qualified institutions to trade crypto during a three-year testing period. Aiming to increase market transparency while maintaining a ban on crypto as a general payment method, this marks another step in Russia's evolving stance on digital assets following its legalization of crypto for international payments and mining amid ongoing sanctions.
Giza Protocol is building infrastructure to enable trustless, context-aware, and permissionless agents in Web3 powered by a three-layer infrastructure—Semantic Abstraction for AI-blockchain interaction, Decentralized Execution with EigenLayer's cryptoeconomic security, and Agent Authorization for non-custodial control. Giza aims to democratize algorithmic finance and enhance capital efficiency across protocols.
Bullet is a high-speed trading network built on Solana designed for traders who need ultra-low latency (2ms) and high performance. It outperforms both centralized and decentralized exchanges. Evolving from Zeta, it introduces a dedicated execution layer with real-time transactions, zero-knowledge proofs for transparency, and a network token for gas and node operations.
Acquiring new users in crypto is increasingly challenging due to market saturation with over 8,700 L1s and 5,200 L2s, fragmented global audiences, trust issues stemming from premature technology, and 415 Bitcoin “death” headlines, as evidenced by declining Google search trends for “crypto.” Developers are overcoming these barriers through personalized onboarding, coordinated marketing across stakeholders, consistent messaging, and storytelling to engage users, helping founders cut through attention fragmentation.
2025.03.13·links.tldrnewsletter.com·9 min read·TLDR Crypto WildcatFi addresses undercollateralized lending risks in DeFi by emphasizing transparency, requiring borrower financial disclosures, and implementing safeguards against manipulation while exploring collateral options to mitigate defaults. The protocol aims to enable on-chain debt financing for crypto businesses, arguing that improved data visibility and tailored terms can align risk-reward dynamics despite inherent credit challenges.
A trader used high leverage to build a massive ETH long position on Hyperliquid, then strategically withdrew collateral to trigger liquidation, offloading risk onto the HLP vault while hedging elsewhere for profit. The result: a $1.8M gain for the trader and a $4M loss for HLP, exposing vulnerabilities in liquidity mechanics rather than an outright exploit.
Amid declining liquidity and valuation pressures in Web3, revenue is re-emerging as a critical focus, pushing projects to adopt transactional fee models over speculative token launches. Industry leaders like Uniswap sustain consistent fees, while infrastructure providers like Polymarket and MetaMask generate significant revenue from transaction volumes, with some, like Ronin and Kaito, experimenting with token buybacks to tie value to cash flow.
2025.03.13·links.tldrnewsletter.com·8 min read·TLDR Crypto Ledger Donjon's security research revealed a potential supply chain attack on Trezor Safe 3 where an attacker could tamper with its microcontroller to compromise user funds, despite the device using a Secure Element. Trezor responded by addressing the vulnerabilities, highlighting the importance of cross-industry cooperation and continuous security improvements in the crypto space.
A user on Uniswap v3 lost $216k to a sandwich attack when swapping $221k USDC for USDT, demonstrating significant MEV risks in DeFi transactions.
US inflation came in lower than expected, with the Consumer Price Index (CPI) rising 2.8% YoY vs the estimated 2.9%, and Core CPI at 3.1%, slightly below the expected 3.2%.
The House voted 292-131 to overturn the Biden administration's DeFi broker rule, which would have imposed KYC on non-custodial software providers.
2025.03.13·links.tldrnewsletter.com·6 min read·TLDR Crypto European regulators are investigating OKX's Web3 platform for allegedly laundering $100M of the $1.5B stolen in the February Bybit hack (attributed to North Korea's Lazarus Group), raising questions about its compliance with the EU's MiCA framework.
2025.03.13·links.tldrnewsletter.com·3 min read·TLDR Crypto Figma's connected projects enable collaboration between separate teams, like agencies and clients, using their own paid plans. Admins manage access, security, and libraries, while each team covers its seat costs. Projects can be connected, shared, or disconnected, with plan limitations affecting features and the number of connected projects allowed. Figma plans to launch the new capability in the first half of 2025.
Demonstrated with Aloy from Horizon Forbidden West, Sony is testing AI-powered video game characters capable of real-time conversations. Using technology from Meta, OpenAI, and Sony's Mockingbird, the AI replicates speech and expressions. The project remains a prototype, with no plans for integration into games yet, amid concerns over industry job impacts.
Korean Air just refreshed its logo for the first time in 40 years following its $1.6 billion merger with Asiana Airlines. The new look, created by design firm Lippincott, features a sleeker Taeguk symbol, a custom typeface, and a simplified blue color scheme to give it a more premium feel. With this rebrand, Korean Air is setting the stage for global expansion while keeping its signature blue livery intact
2025.03.13·links.tldrnewsletter.com·3 min read·TLDR Design Designers can use AI to overcome imposter syndrome by generating meaningful sample content, organizing positive feedback, and understanding business needs. These approaches help build confidence, improve iteration, and enhance communication of design impact.