ETERNAL ALEX COEFFICIENT
The Eternal Alex Coefficient measures how much a former employee still shapes decisions years after leaving.
Full-text search across article titles and summaries. Refine with sector, year, content type, and reading-time filters.
The Eternal Alex Coefficient measures how much a former employee still shapes decisions years after leaving.
Sponsored
Welcome to Delve Shipmas—5 days of incredible product launches to bring your team the cheer they really need.
Sam Altman recently appeared on Big Technology Podcast to discuss OpenAI's strategy to win, where the product lineup is going in the coming year, how OpenAI's over $1 trillion in AI infrastructure commitments make sense, and the company's future plans for AI devices and AI cloud. The strategy to keep OpenAI ahead involves getting people to use its products, keeping them there with the best models, serving the use cases they want with reliable compute, then expanding into areas like enterprise and hardware. The full conversation, along with links to the audio, is available in the article.
Tesla won its fight over Elon Musk's record-breaking pay package approved by shareholders in 2018. The Delaware Supreme Court said the lower court's ruling would have deprived Musk of compensation for several years of work, so he should get the full pay package. Musk's 2018 stock options have grown in value to $139 billion. It was initially valued at more than $2.3 billion, already shattering previous compensation records.
The Trump administration launched a partnership in October to build at least $80 billion worth of new, large-scale nuclear reactors. The money will support the construction of a type of pressurized water reactor developed by Westinghouse that can generate about 1,110 megawatts of electric power. The partnership includes profit-sharing mechanisms that will give the US government some of the upside if the initiative succeeds.
China reportedly has a prototype EUV chipmaking machine. While it has not yet produced a working chip, it is apparently creating EUV light. The machine is expected to produce its first chip in 2028. The Dutch company ASML had previously been the sole producer of EUV machines.
Sponsored
Incubated by Insta360 [10] and third parties, A1 combines 8K, 360 capture with intuitive motion control. Weighing sub-249g, the drone offers an experience like no other, fully immersing pilots with their VR-style Vision Goggles. This is flight unlike anything else. Grab yours today from Antigravity's store, shipping globally at
Efficiency becoming a metric will create pressure and cause exploration to look wasteful. High-token sessions will start to look sloppy. Over time, AI will turn into another optimization target, and eventually into something that people virtue signal about rather than something that reflects real impact. Some of the most valuable work exists in the messy, high-token exploratory phase that resists easy measurement.
Design is about finding the right problem, the right intent, and the right vision. The features you design and build today should be considered a step toward that vision. Without the right context and goal, you could be iterating toward a direction that wasn't chosen intentionally. The industry is not very patient, but that doesn't mean you should devalue the why behind your designs in favor of output.
Lawyers for David Ellison sent a letter to Warner Bros. Discovery on the night of December 3 expressing serious concerns about the fairness and adequacy of the company's sale process. This surprised Warner Bros., as the company believed it had run a fair process. Ellison's inner circle tried to retract the statement, claiming some partners didn't even know the letter was going out, but it was too late. In response, Warner Bros. decided to sell to Netflix.
Software quality startup Graphite has been acquired by code generation startup Cursor. Graphite started out as Screenplay, which couldn't find product-market fit right away. The conversations around Graphite started right before the pandemic lockdowns. The team started working in person together as soon as it was advisable, allowing them to knit a culture early and build trust. The three founders took three distinct roles, and they all remain vital to the company.
Josh Woodward, a 16-year Google veteran, was tasked with growing Gemini while also keeping his existing job running Google Labs.
It's important to take the time to learn new technologies, the capabilities of those technologies, and where they can help your business, rather than just shoehorning in new technologies everywhere you can.
Almost everybody has stopped saying that large language models are stochastic parrots.
Scalable quantum computing is now plausibly imminent.
The ad tiers scored all net additional subscribers for both Netflix and Disney+ this year.
2025 was a strong and eventful year of progress in LLMs.
Sponsored
Many financial systems of record (like ERP) were built at a time when AI tools were the stuff of science fiction. Now, they need to work together.
OpenAI has introduced new personalization options in ChatGPT, letting users adjust enthusiasm, warmth, and emoji use directly. These controls, available in the Personalization menu, offer "More," "Less," or "Default" settings, expanding tone customization beyond the existing base style and tone feature.
Cursor has acquired Graphite, a company known for its performance-focused internal developer portal. This marks Cursor's third acquisition as it aims to build a comprehensive AI-powered dev platform.
Anthropic's Bloom is an open-source tool for generating automated behavioral evaluations of AI models. Bloom assesses specific behaviors like self-preferential bias and sabotage by creating scenarios and quantifying behavior occurrence across models. It efficiently differentiates between aligned and misaligned models and correlates strongly with human judgment, enabling scalable and reliable behavior evaluations.
LLM use is rising. People are increasingly using different LLMs, different products, and in different places. ChatGPT remains dominant and keeps acquiring new users, but Gemini's growth has been faster over the last few months. OpenAI's revenue seems to be on track, but consumer revenue is likely decreasing as a share. A substantial share of workplace AI use involves workers adopting tools on their own rather than waiting for employer-provided access.
What happens when agents run out of memory determines whether they continue productively or have to start from scratch. This post explores an evaluation framework that measures how much context different compression strategies preserve. Structured summarization retains more useful information than alternative methods without sacrificing compression efficiency.
Benchmarks are the most widely misunderstood part of the AI ecosystem. The narrative keeps implying a universal increase in intelligence, but the numbers can be misleading. To navigate this noise, look at the aggregate, look at the relative, and verify with your own tasks. The only benchmark that matters at the end of the day is your own workload.
This document contains a diary for an experiment aimed at teaching an LLM using GRPO to generate regex given a description. It details the performance, learnings, modifications, and key takeaways from each experiment. The initial training run was on December 17. It saw the model quickly learning how to generate valid regex tags, but the model was basically generating random regex strings.
Andrej Karpathy has outlined paradigm shifts of LLMs in 2025, including fast inference engines, model distillation trends, real-time agents, neural GPUs, and the rise of high-quality open models like DeepSeek-V2 and RWKV.
Sponsored
One in three employees uses AI tools without approval, risking data leaks and compliance violations. With this Enterprise AI Governance Kit [13] from You.com [13], you will get ready-to-use templates to help protect your organization, including security checklists, usage policies, and governance frameworks.
Qwen-Image-Layered is a model capable of decomposing an image into multiple RGBA layers. Each layer can be independently manipulated without affecting other content. They can be resized, repositioned, and recolored. The approach enables high-fidelity and consistent editing.
MiMo-V2-Flash is a powerful, efficient, and ultra-fast foundational language model that excels in reasoning, coding, and agentic scenarios. It serves as an excellent general-purpose assistant for everyday tasks. The model is available globally on Hugging Face, AI Studio, and Xiaomi's API platform. Benchmark results are available in the article.
jax-js is a machine learning framework for the browser. It brings JAX-style, high-performance CPU and GPU kernels to JavaScript, so users can run numerical applications on the web. The library is written from scratch and has no external dependencies. It can run anywhere a browser can run.
MCP servers give agents the tools they need to accomplish tasks. This post discusses how to multiplex MCP servers to simplify the connection to various tools within them. Multiplexing allows multiple MCP servers to be used over a gateway in a single interaction. It allows agents to access multiple MCP servers with different stacks, clouds, applications, and frameworks for specialized tasks.
tcgen05 is the set of PTX instructions that programs Tensor Cores on the latest NVIDIA Blackwell GPUs. This post contains a tutorial for Blackwell in plain CUDA C++ with PTX. It documents the author's process of learning tcgen05 and reaching 98% of CuBLAS speed. Readers can follow the tutorial using Modal or any other B200 cloud providers.
METR topics are public, making it easy to game METR horizon length measurements for a frontier lab. The horizon length under METR's assumptions might be adding little information beyond benchmark accuracy. There is a meme going around based on a team achieving a one to four hour range on the METR plot. This post explains why the plot has been interpreted incorrectly.
AI is incredibly good at some tasks while being really bad at others. This is the 'Jagged Frontier' of AI ability. The jaggedness is likely going to remain a big part of AIs going forward. However, the growing frontier will outpace jaggedness.
Masayoshi Son has already sold SoftBank's entire $5.8 billion stake in Nvidia, offloaded $4.8 billion of its T-Mobile US stake, and slashed staff to come up with the money.
A lot of research progress on agentic reinforcement learning made its way into Gemini 3 Flash, but it was too late for Pro.
World simulators are models trained to predict how the world evolves over time, frame-by-frame, using large amounts of video and interaction data.
Chinese GPU maker Moore Threads' Huagang architecture promises significant performance gains in gaming and AI.
In 2026, US companies will apply a factory mindset to energy, mining, construction, and manufacturing by pairing AI and autonomy with skilled labor to standardize complex work.
Sponsored
The transparency of traditional blockchains forces a difficult choice: utility or privacy. While transparency can build trust, it can also expose user activity and personal data. Midnight [4] empowers innovators to break free from this limitation, to a place where privacy is not just a feature, but foundational.
Representatives Max Miller and Steven Horsford proposed amending the IRS Code to exempt stablecoin transactions under $200 from capital gains taxes if issued by GENIUS Act-permitted issuers pegged to the dollar with stable trading ranges, while excluding brokers and dealers from the benefit. The draft allows taxpayers to defer income recognition on staking and mining rewards for up to five years rather than immediate taxation, addressing "phantom income" concerns, and extends securities lending treatment to digital asset lending arrangements.
Coinbase filed federal lawsuits against Connecticut, Michigan, and Illinois, arguing that prediction markets fall under the CFTC's exclusive federal jurisdiction rather than state gaming boards. The Illinois Gaming Board issued cease-and-desist letters to Kalshi, Robinhood, and Crypto.com in April over sports event contracts, claiming state law violations. Chief Legal Officer Paul Grewal argues Congress deliberately excluded only specific commodities like onions from CFTC jurisdiction, making all other subjects, including sports, federally regulated.
Quantum computing is transitioning from “remote theory” to a plausible engineering timeline, and Bitcoin's slow governance means the network needs to start post-quantum prep now, not when a “Q-day” headline hits. The core fears include keys getting cracked, panic, contentious forks, and institutional confidence breaking if the community has to decide under duress. Furthermore, there's a world where millions of BTC in old “pay-to-public-key” outputs (including Satoshi-era coins) that can't be proactively migrated, forcing a future decision between freezing them or letting a quantum-capable attacker scoop them.
JPMorgan launched its JPM Coin tokenized deposits on Coinbase's Base blockchain, driven by institutional demand for bank deposit products on public chains that can be interest-bearing, unlike stablecoins restricted by the GENIUS Act. The permissioned, whitelisted JPMD serves clients using Coinbase for crypto transactions as collateral and margin payments, offering different risk profiles than stablecoins or off-chain bank accounts with cutoff time issues. Tokenized deposits are the "cousin of stablecoins" as banks face interoperability challenges exporting products beyond their four walls versus stablecoins' broader distribution. The bank maintains sole control over the smart contract with proper key storage and role separation, positioning public blockchain infrastructure as the destination for innovation and customer activity.
Sponsored
No slippage. No Fees. Move USDC from any chain straight into USDH on Hyperliquid for free!. Only on Across Protocol.