LearnByDoing With Steven

·

Channel · 4 Shows

Education

Education

Updated Weekly
Education

Education

Updated Weekly
Education

Education

Updated Weekly
Education

Education

Updated Weekly

11/23/2025

Stanford Admissions: Requirements, Essays, and Applications

These sources provide a comprehensive guide to Stanford University's admissions requirements for both undergraduate and graduate applicants. The information covers the holistic review process, emphasizing that essays, personal experiences, and recommendation letters are critical, especially given the highly selective acceptance rates. For undergraduates, the materials detail requirements such as SAT/ACT scores (with specific superscoring and submission policies), the number of required teacher recommendations, and a breakdown of the numerous supplemental essay prompts. For graduate admissions, the texts explain the decentralized application process managed by individual schools, the requirement for a highly specific Statement of Purpose demonstrating fit with faculty research, varying GRE and GMAT policies by program (including the GSB's mandatory testing but no minimum score), and English proficiency requirements for international students, along with visa and funding considerations.
11/23/2025

UC Berkeley Admissions and Application Guide

These documents provide a comprehensive guide to the highly competitive undergraduate and graduate admissions processes for the University of California system, with a significant focus on UC Berkeley. The texts meticulously outline academic requirements for first-year and transfer applicants, stressing the importance of exceeding minimums, such as maintaining an average GPA near 3.9 and maximizing course rigor through A-G requirements. A key strategic element highlighted across sources is that UC Berkeley is "test-free," meaning SAT/ACT scores are not considered for admission, which shifts the academic assessment burden entirely to grades and coursework rigor. Furthermore, the documents detail the application logistics, including the strict October 1 to December 1 filing window, the necessity of demonstrating English language proficiency for international students, and the critical role of the Personal Insight Questions (PIQs) in providing qualitative context for holistic review.
09/16/2025

From_POC_to_PROD__Hard_Lessons_in_Generative_AI

From_POC_to_PROD__Hard_Lessons_in_Generative_AI
Feb 19

Comparative Analysis of Large Model Inference Optimization Frameworks

This report provides a comparative analysis of specialized large language model (LLM) inference frameworks designed to overcome hardware limitations and high computational costs. It distinguishes between high-throughput server solutions like vLLM, which uses PagedAttention to eliminate memory fragmentation, and SGLang, which optimizes complex, multi-turn interactions through RadixAttention and structured generation. For local deployment, the text evaluates Ollama and LM Studio, highlighting how they leverage llama.cpp and the GGUF format to run models on consumer-grade hardware. The study further explores critical performance-enhancing technologies such as quantization, speculative decoding, and continuous batching. Ultimately, the sources serve as a guide for selecting the right infrastructure based on specific needs, ranging from cloud-scale API services to private local assistants.
May 18

The One-Person Company (OPC) Paradigm: AI-Driven Industrial Re-Architecture

The One-Person Company (OPC) Paradigm: AI-Driven Industrial Re-Architecture The definition of entrepreneurship is undergoing a fundamental structural transition. Driven by advanced generative AI frameworks and automated operational pipelines, the "One-Person Company" (OPC) is no longer a simple legal designation, but the core building block of the modern digital economy. An objective analysis of the 2026 China OPC landscape reveals critical structural trends: 1️⃣ Scale and Velocity: By mid-2025, China's active one-person limited liability companies exceeded 16 million, representing over 25% of all national business entities. This supply-side explosion is highly concentrated in digital-native sectors like autonomous agent development and specialized digital media. 2️⃣ The Institutional Sandbox: Over 20 major municipalities have established targeted support frameworks, shifting from basic rent-free physical spaces to deep infrastructure provisioning, including computing power vouchers, desk registration protocols, and dedicated micro-seed funding. 3️⃣ Severe Revenue Polarization: Despite ultra-low startup barriers, commercialization remains highly competitive. The revenue profile is sharply pyramidal, with roughly half of exploration-phase founders earning under 7,000 RMB monthly, while a tiny elite of domain-specific, asset-reusable builders achieve multi-million RMB annual run rates. 4️⃣ Organizational Inertia: Large technology corporations are struggling to adapt their enterprise service pipelines to cater to this highly decentralized micro-client market, leaving a critical gap in lightweight API access and distribution support. As the industry enters its secondary phase of development, the priority for solo builders must shift from superficial tool experimentation to the validation of concrete commercial orders, converting individual labor into reusable, long-term digital assets. Source Report: https://www.opcquan.com All my links: https://linktr.ee/learnbydoingwithsteven #OnePersonCompany #ArtificialIntelligence #StartupEcosystem #DigitalEconomy #BusinessInsights #Automation #TechPolicy #AIAgents #Entrepreneurs #learnbydoingwithsteven
11/11/2025

Reflections on AI's Future from Former FAIR Director Tian Yuandong

The text provides a transcript from an interview with Tian Yuandong, a former Research Director at FAIR (Meta AI Research), following his layoff during a major restructuring at Meta. The discussion begins by confirming the Meta AI department layoffs and then transitions into a broader conversation about the future of AI research and industry trends. Yuandong shares his thoughts on the potential of Large Language Models (LLMs), the advantages of Reinforcement Learning (RL) as an active learning method, and his belief that the Scaling Law represents a pessimistic trajectory for AI development due to its reliance on exponentially increasing data and resources. He also reflects on his time at FAIR, expressing regret for not having focused more on engineering work, and concludes by discussing the shifting landscape for AI talent and the importance of balancing frontier research with practical applications.
Jul 20

1 Agents You Can t Answer For w_learnbydoingwithsteven.github.io_linktr.ee_learnbydoingwithsteven

1 Agents You Can t Answer For w_learnbydoingwithsteven.github.io_linktr.ee_learnbydoingwithsteven
Jan 10

LLM-D: Optimizing Distributed AI Inference with Intelligent Routing

The provided text introduces LLM-D, an open-source project designed to optimize AI inference by treating requests like planes managed by an air traffic controller. By utilizing a distributed architecture on Kubernetes, this system significantly reduces latency and operational costs for complex tasks like RAG and agentic workflows. It functions through an inference gateway that intelligently routes prompts based on current system load and the likelihood of cached data. A key innovation is the disaggregation of the prefill and decode phases, allowing them to scale independently while sharing resources. Ultimately, this approach achieves massive improvements in response times and throughput, making it ideal for high-demand, mission-critical AI environments.
12/29/2025

The Global Wind Power Landscape: Market Shifts and Technical Horizons

The 2024 global wind energy market reached record milestones, with annual installations hitting 117 GW and total capacity surpassing 1,136 GW. This expansion is characterized by a significant shift in leadership, as Chinese manufacturers now occupy the top four spots globally, though their activities remain largely centered on domestic and emerging markets. While the industry faces economic hurdles such as high interest rates, trade tariffs, and supply chain constraints, technical innovation continues through the development of larger turbines and floating offshore platforms. Advanced digital tools and artificial intelligence are also becoming essential for optimizing maintenance and enhancing energy output in harsh environments. Furthermore, as infrastructure ages, the sector is increasingly focused on circular economy solutions for recycling complex composite turbine blades. Ultimately, the reports highlight a future defined by a bifurcated market where national energy security and cost-efficiency drive the transition toward wind as a primary power source.
09/18/2025

The_Modern_AI_Tech_Stack__An_Engineer_s_Guide

The_Modern_AI_Tech_Stack__An_Engineer_s_Guide

Learn By Doing With Steven 数能生智