← Ch 4 · Contents · ↑ Contents
Chapters: Ch 1 · Ch 2 · Ch 3 · Ch 4 · Ch 5
From self-1.md (Oct 2024 – Feb 2026)
Competitive Programming Patterns — Advanced LeetCode
A collection of advanced algorithmic patterns for competitive programming — segment trees, tries, Dijkstra variants, union-find — going beyond the basic 15-pattern framework into territory required for hard LeetCode and ICPC-level problems. [→ algorithms-data-structures]
System Design: Instagram Scale — 2.5 Billion Users Deep dive into how Instagram architected their systems to support 2.5 billion users; sharding, caching, CDNs. x.com — https://x.com/systemdesign42/status/1800491019663970354
A detailed thread by Neo Kim (@systemdesign42) covering Instagram’s architecture at 2.5 billion users: how they scaled horizontally with sharding, reduced latency with CDNs, managed media storage on S3, and designed their feed ranking and notification systems. [→ data-engineering; algorithms-data-structures]
Rearchitecting Core Services at X (Twitter) Engineering overview of X’s infrastructure rearchitecture: monolith-to-microservices, cost reduction. x.com — https://x.com/cambridgemike/status/1835774409786986572
Mike (@cambridgemike)’s thread on X’s infrastructure overhaul after Elon Musk’s acquisition: decomposing the monolith, cutting infrastructure costs by 60%+, and migrating core services to a leaner microservice architecture. [→ data-engineering]
B-Trees — Why They Are Provably Optimal Edward Kmett thread on the mathematical optimality of B-trees for many practical applications. x.com — https://x.com/kmett/status/1811904210021286014
Edward Kmett’s thread explaining the external memory model of computation and why B-trees achieve the optimal I/O complexity for sorting, searching, and range queries — connecting algorithm theory to practical database and file system design. [→ algorithms-data-structures]
Breaking Out of a Fold — Short-Circuiting Functional Iteration Blog post on techniques for early termination inside fold/reduce operations in functional programming. ethankent.dev — https://ethankent.dev/posts/breaking_fold/
Expands on the core insight that fold is inherently total (must process all elements) and surveys techniques for encoding early termination: returning Either/Option in the accumulator, using foldRight laziness, or switching to find/takeWhile. [→ algorithms-data-structures; scala-functional-programming]
Graph Neural Networks for Exact Scheduling (2022 Paper) Research using GNNs to accelerate exact solving of scheduling problems in industrial settings. arxiv.org — file:///Users/vishwas/Downloads/2022Exactsolvingschedulingproblemsacceleratedbygraphneuralnetworks_RG.pdf
A 2022 research paper proposing the use of Graph Neural Networks to accelerate branch-and-bound solvers for NP-hard scheduling problems, achieving significant speedups by learning good branching heuristics from problem structure. [→ algorithms-data-structures; machine-learning-ai; data-engineering]
AI for Production Scheduling — Systematic Literature Review Survey of AI techniques (PSO, neural nets, RL) for production scheduling; real industrial results. mdpi.com — https://www.mdpi.com/2079-9292/12/23/4732
A MDPI Electronics systematic literature review covering AI-based production scheduling methods: particle swarm optimisation, neural networks, and reinforcement learning. Finds that AI solutions consistently reduce costs, improve energy efficiency, and outperform classical schedulers in industrial trials. [→ data-engineering; machine-learning-ai]
Topological Data Analysis — Book Recommendation Frank Nielsen’s recommendation for an introductory TDA textbook covering persistent homology and simplicial complexes. x.com — https://x.com/FrnkNlsn/status/1814972379187093563
Frank Nielsen (a leading information geometry researcher) recommends a starting textbook for Topological Data Analysis — the study of shape and connectivity in high-dimensional data using persistent homology, Betti numbers, and Mapper algorithms. [→ mathematics-science; machine-learning-ai]
Divergence as Adjoint of Gradient — PDE Primer Gabriel Peyré’s explanation of divergence/gradient adjoint relationship via integration by parts; Laplacian workhorses. x.com — https://x.com/gabrielpeyre/status/1816699972789628992
A tweet by Gabriel Peyré (researcher in optimal transport and image processing) explaining that the divergence operator is the negative adjoint of the gradient via integration by parts — a fundamental identity underlying PDEs, variational methods, and many imaging algorithms. [→ mathematics-science]
Warnsdorff’s Rule — Knight’s Tour Heuristic Warnsdorff’s 1823 heuristic for constructing knight’s tours: always move to the square with fewest onward moves. github.com — https://raw.githubusercontent.com/douglassquirrel/warnsdorff/refs/heads/master/5_Squirrel96.pdf
The original 1823 heuristic and Squirrel’s 1996 formalisation: at every step, move the knight to the square with the minimum number of valid onward moves. This greedy local rule produces a valid tour in O(n²) time on standard boards without backtracking. [→ algorithms-data-structures]
Universal Algebra Pilot Project — Terry Tao Terry Tao’s blog post on exploring new human-machine collaboration methods in universal algebra. terrytao.wordpress.com — https://terrytao.wordpress.com/2024/09/25/a-pilot-project-in-universal-algebra-to-explore-new-ways-to-collaborate-and-use-machine-assistance/
Terry Tao describes a pilot project using large language models and proof assistants to explore conjectures in universal algebra — specifically around equational theories and free algebras — demonstrating a new mode of human-machine mathematical collaboration. [→ mathematics-science; machine-learning-ai]
Grad Students Find Inevitable Patterns in Number Sets — Quanta New proof marking first progress in decades on arithmetic Ramsey theory: order emerging from disorder in large sets. quantamagazine.org — https://www.quantamagazine.org/grad-students-find-inevitable-patterns-in-big-sets-of-numbers-20240805/
Graduate students proved a new result in arithmetic Ramsey theory: any sufficiently large set of integers must contain certain structured patterns (arithmetic progressions or other configurations) — the first significant progress in decades on this class of problems. [→ mathematics-science]
Complexity Theory’s 50-Year Journey to Knowledge Limits — Quanta History and current state of meta-complexity theory: how hard is it to prove hardness? quantamagazine.org — https://www.quantamagazine.org/complexity-theorys-50-year-journey-to-the-limits-of-knowledge-20230817/
A Quanta survey of meta-complexity theory — the study of how hard it is to prove that computational problems are hard. Covers circuit complexity lower bounds, natural proofs barriers, and why proving P≠NP remains so elusive after 50 years. [→ mathematics-science; algorithms-data-structures]
Elliptic Curve Murmurations Discovered With AI — Quanta AI-discovered statistical patterns in elliptic curve families; mathematicians working to explain the phenomenon. quantamagazine.org — https://www.quantamagazine.org/elliptic-curve-murmurations-found-with-ai-take-flight-20240305/
ML models trained on large families of elliptic curves discovered unexpected oscillating statistical patterns (dubbed “murmurations”) correlated with the rank of the curve. Number theorists are now working to find a theoretical explanation for this AI-discovered phenomenon. [→ mathematics-science; machine-learning-ai]
How Base 3 Computing Beats Binary — Quanta Ternary (base-3) computing’s theoretical advantages and potential applications in cryptography and computing. quantamagazine.org — https://www.quantamagazine.org/how-base-3-computing-beats-binary-20240809/
A Quanta article on the theoretical advantages of balanced ternary (base-3) computing: it can represent integers more efficiently than binary and has interesting properties for cryptography and certain arithmetic algorithms — though practical hardware remains binary. [→ mathematics-science; algorithms-data-structures]
Geometric Langlands Conjecture — Monumental Proof — Quanta 30-year effort culminating in proof of the geometric Langlands conjecture; connecting geometry, algebra, and number theory. quantamagazine.org — https://www.quantamagazine.org/monumental-proof-settles-geometric-langlands-conjecture-20240719/
A 30-year effort by a team of mathematicians culminating in the proof of the geometric Langlands conjecture — a vast generalisation connecting algebraic geometry (D-modules on moduli spaces) and representation theory (automorphic forms), forming the geometric analogue of the Langlands program. [→ mathematics-science]
Scientists Find a Fast Way to Describe Quantum Systems — Quanta New algorithm enabling efficient classical description of quantum system states, bypassing exponential blowup. quantamagazine.org — https://www.quantamagazine.org/...
A Quanta article on a new algorithm allowing efficient classical description of certain quantum system states — avoiding the exponential state-space blowup that makes quantum simulation hard. Relevant to quantum simulation and quantum advantage questions. [→ mathematics-science; machine-learning-ai]
OpenClaw Workflows — AI Agent Orchestration (Feb 2026) Framework for orchestrating multi-step AI agent workflows with dependency management and parallelism. github.com — https://github.com/...
OpenClaw is a framework for building multi-step AI agent workflows with dependency graphs, parallel execution, and state management — relevant to building production agentic systems on top of Claude or similar models. [→ machine-learning-ai; infrastructure-devops]
Intelligent AI Delegation — Arxiv (Feb 2026) Research on automatic delegation strategies in multi-agent AI systems: when and how to hand off tasks. arxiv.org — https://arxiv.org/abs/...
An arxiv paper on adaptive delegation in multi-agent AI systems: how to automatically decide when a task should be delegated to a subagent, which agent to choose, and how to handle authority transfer — formalising the orchestration problem. [→ machine-learning-ai]
Algorithms & Interview Prep
Algorithms by Jeff Erickson Jeff Erickson’s (UIUC) comprehensive open textbook on algorithm design: sorting, searching, graph algorithms, dynamic programming, NP-completeness, and randomized algorithms. One of the most mathematically rigorous free algorithm textbooks available online. https://jeffe.cs.illinois.edu/teaching/algorithms/
GitHub: ossu/computer-science Open Source Society University’s structured path to a free, self-taught computer science education: curated MIT, Stanford, and other university course materials organized into a coherent CS degree curriculum covering math, algorithms, systems, and theory. https://github.com/ossu/computer-science
Arpit Adlakha: Senior Software Interview Roadmap (@arpit20adlakha) A detailed Twitter thread outlining the roadmap for clearing senior software engineer interviews at Uber L5A/L5B and Google L5/L6: algorithm patterns, system design topics, behavioral interview frameworks, and resource recommendations with approximate time investment. https://x.com/arpit20adlakha/status/1805084468870521100
Ashish Pratap Singh: 15 LeetCode Patterns (@ashishps_1) Twitter thread listing fifteen algorithmic patterns that cover the majority of LeetCode problems: Prefix Sum, Two Pointers, Sliding Window, Fast & Slow Pointers, In-place LinkedList Reversal, Monotonic Stack, Top K Elements, Overlapping Intervals, Modified Binary Search, and more. A structural framework for approaching unknown problems. https://x.com/ashishps_1/status/1814884401249198569
CS & Algorithms Channels
Computerphile (YouTube) https://www.youtube.com/channel/UC9-y-6csu5WGm29I7JiwpnA
University of Nottingham’s companion channel to Numberphile, featuring short computer science videos with academic researchers. Covers algorithms, cryptography, programming language theory, complexity, and computer history. [→ algorithms-ds; cs-theory]
PapersWeLove (YouTube) https://www.youtube.com/channel/UCoj4eQh_dZR37lL78ymC6XA
Community meetup recordings of researchers and practitioners presenting influential computer science research papers. Papers We Love is one of the most intellectually stimulating CS communities; the talks cover distributed systems, type theory, algorithms, and foundational CS. [→ algorithms-ds; cs-theory]
Steven Skiena (YouTube) https://www.youtube.com/channel/UCh27N8m4Jp_rkwx26Q4iclQ
Author of The Algorithm Design Manual — lectures from Stony Brook covering algorithm analysis, graph algorithms, dynamic programming, and computational geometry. His practical, problem-solving-oriented approach complements more theory-heavy textbooks. [→ algorithms-ds]
Jordan has no life (YouTube) https://www.youtube.com/channel/UCbvDQKzAJ5GwCjTrv4FWkxg
System design interview prep and distributed systems concepts explained by a working engineer. Covers consistent hashing, CAP theorem, database internals, and scalable architecture patterns. [→ algorithms-ds; infrastructure]
Pavel Mavrin (YouTube) https://www.youtube.com/channel/UC9WDkCyUlNTncq1Y-mFSgLQ
Competitive programming and algorithms by a high-rated Codeforces competitor. Covers advanced graph algorithms, segment trees, and mathematical problem-solving techniques used in competitive programming. [→ algorithms-ds; competitive-programming]
SF Systems Club (YouTube) https://www.youtube.com/channel/UCOYdObvVg2lqJix4mnE9MvA
San Francisco systems programming club — talks on operating systems internals, databases, compilers, and low-level systems topics. A Bay Area community for those interested in the engineering below the application layer. [→ algorithms-ds; systems]