Book Cube: 2026, page 4
Book Cube publications from 2026, page 4: bilingual notes, sources, and original Telegram posts.
- · #AI4SDLC
State of AI4SDLC on HighLoad++: materials for the report (Category AI4SDLC)
State of AI4SDLC on HighLoad++: materials for the report (Rubric AI4SDLC ) I spoke today. Saint HighLoad++ 2026 State of AI4SDLC: How AI is Changing Development Processes in Large Companies The main idea I had was that AI4SDLC is now about redesigning the engineering system: platform, processes, metrics, security, economics and the very role of an engineer. 1️ In the first part, I showed that the implementation of AI has already happened, but trust and team effect do not keep up with it.
- · #AI4SDLC
Dora ROI: how to calculate the effect of AI-assisted development without magic (Category AI4SDLC)
Dora ROI: how to calculate the effect of AI-assisted development without magic (Rubric AI4SDLC ) Deal with the new report DORA and Google CloudThe ROI of AI-assisted Software Development\. He's good at continuing the last Dora. 2025 "State of AI-assisted Software Development" handlerThe main thing was that AI works as an amplifier of an engineering system, and the effect depends on seven possibilities. (I'm sorry. handler separately). The main point of the previous report was that strong teams from AI get more speed, weak teams get more chaos.
- · #AI4SDLC
[2/2] The New SDLC: harness, factory model and agentic engineering economy (Category AI4SDLC)
\[2/2\] The New SDLC: harness, factory model and agentic engineering economy (Rubric AI4SDLC ) I continue to analyze the May whitepaper from Google, where first We discussed the shift from syntax to intent, and then the spectrum of vibe coding → agentic engineering and context engineering. Now let's talk about what surrounds the model and turns it into a working agent.
- · #AI4SDLC
[1/2] The New SDLC: From Vibe Coding to Agentic Engineering (Category AI4SDLC)
\[1/2\] The New SDLC: From Vibe Coding to Agentic Engineering (Rubric AI4SDLC ) I read May. whitepaper Addy Osmani, Shubham Saboo and Sokratis Kartakis from Google’s AI Agent Training Series. The document is useful not in fashionable terms, but in collecting stable mental models: the tools change every week, and the frame must survive this change. It doesn’t fit in one post, so there will be two:)
- · #Leadership
Archetypes and strategization (Category Leadership)
Archetypes and strategization (Rubric Leadership ) Yesterday, for the third time, I was trained by Andrei Shishakov on archetypes and personal strategizing. And it's not that I don't understand the first time, but rather that I like the way Andrei pitches his material and twists it year after year. For me, such training is often a way to go even deeper into self-reflection and get some insights: for the first time I studied with Andrei in the school. 2022 after all known events, the second time 2024 A year and a third time now.
- · #AI4SDLC
Comparing AI Coding Agents: A Task-Stratified Analysis of Pull Request Acceptance (Category AI4SDLC)
Comparing AI Coding Agents: A Task-Stratified Analysis of Pull Request Acceptance (Rubric AI4SDLC ) I read this one. paperIt shows why ranking agents without considering the type of task are not very useful. This article shows that the correct answer to the usual question "what coding agent is the best?" is it depends:) For their analysis, they took a PR kit from GitHub called AIDev, which I'm talking about. told Recently.
- · #Books
Hard & Soft: How the Russian information technology market was created (Books column)
Hard & Soft: How the Russian information technology market was created (Rubric Books ) I read this book by Boris Shcherbakov, which tells the story of the birth of IT in Russia, but it is rather an ironic autobiographical book about how the management practices of large companies looked in practice in the 90s. This is the story of a man who worked at Hewlett-Packard, Verysell, Party, Oracle and Dell, was at the origins of the Russian IT market and adapted Western management approaches to the local context.
- · #AI
WebMCP: How every website talks to agents — Tara Agyemang, Google Chrome(AI column)
WebMCP: How every website talks to agents — Tara Agyemang, Google Chrome(Rubric AI ) Interesting. speech Tara Agyemang from Google Chrome about WebMCP. Not the emergence of a new browser API, but a more practical idea: normal browser flow management for agents can be arranged not through “look at the screen and guess where to click”, but through explicit tools that the site itself gives to the agent.
- · #AI4SDLC
Your Attention Is the Bottleneck, Not Your Agents — Zack Proser, WorkOS (Category AI4SDLC)
Your Attention Is the Bottleneck, Not Your Agents — Zack Proser, WorkOS (Rubric AI4SDLC ) I saw it the other day. report Zack Proser from WorkOS, which talks about AI burnout from working with agents. I think you know what it feels like to work with Claude or Codex, when it seems like more has been done than before, but by the middle of the day, you’ll be out. I really started feeling that when I was running agents all day.
- · #AI
Stop Making Models Bigger, Make Them Behave — Kobie Crawford, Snorkel (AI column)
Stop Making Models Bigger, Make Them Behave — Kobie Crawford, Snorkel (Rubric AI ) I saw it. 20minute report Kobie Crawford from Snorkel AI conference AI Engineer London (10 April 2026). This is a rare report that goes against the usual "new model bigger and smarter." The main idea is simple and inconvenient for the race of parameters: the capabilities of the model are limited not so much by the architecture and number of parameters, but by the quality of data and tasks on which it was trained.
- · #AI4SDLC
AIDev: Studying AI Coding Agents on GitHub (Category AI4SDLC)
AIDev: Studying AI Coding Agents on GitHub (Rubric AI4SDLC) Read it. short from 9 February 2026 Learn how to learn coding agents from real PR, not demos. It presents a public dataset on which agentic software engineering can be studied not by demos, but by the tracks of real pull requests. Authors are Hao Li, Haoxiang Zhang and Ahmed E. Hassan of Queen's University. This is a good sign of trust: Hassan has long been involved in empirical software engineering and mining software repositories.
- · #AI4SDLC
AI Dev Podcast 3 How to Make a Code Review Tool / T-Bank Experience (Category AI4SDLC)
AI Dev Podcast 3 How to Make a Code Review Tool / T-Bank Experience (Rubric AI4SDLC ) Listened. The latest release of AI Dev Podcast, where my colleagues from T-Bank - Nadezhda Egoshina and Georgy Mkrtchyan - tell hosts Andrei Dmitriev (co-founder jug.ru and the author of the channel @dmitrievandrey8) Andrey Burakov (channel author @another\ sa)It’s like the AI Code Review. I listened with special interest: the topic is close, and the guys talk about the real way, not much embellishing.
- · #Writing
How I write longreads on topics that are important to me (Category Writing)
How I write longreads on topics that are important to me (Rubric Writing ) I decided to share a recipe for how I now write important and difficult articles on technical topics. From the last article,IDP is Dead? No, the GUI monopoly is dying.andAgent-first IDP: how bigtechs and clouds prepare for agents».
- · #PlatformEngineering
Agent-first IDP: how bigtechs and clouds prepare for agents (Category PlatformEngineering)
Agent-first IDP: how bigtechs and clouds prepare for agents (Rubric PlatformEngineering ) Wrote. continuation to an essay on internal platforms. Last time in the textIDP is Dead? No, the GUI monopoly is dying.I’ve explored why the internal platform is losing the monopoly of its GUI and turning into a platform for agents – a layer of capabilities that not only people but also agents turn to. It was about why. The new text is about how. And in the new article, I specifically did not go into futurology.
- · #AI4SDLC
SDLC with AI look through metrics - Anna Gromova @ AI Dev Conf (Category AI4SDLC)
SDLC with AI look through metrics - Anna Gromova @ AI Dev Conf (Rubric AI4SDLC ) Watched May. speech Anna Gromova from T-Bank from the AI Dev Conf conference. Anya and I often discuss AI performance metrics at work, so it’s especially helpful to me that there’s a record that can now be referenced: it’s a relevant and in-depth analysis of how we approach this topic at home. The main idea there is very practical: AI in development can not be properly evaluated if you do not understand the process of code delivery.
- · #AI4SDLC
What if the network was the sandbox? — Remy Guercio, Tailscale (Category AI4SDLC)
What if the network was the sandbox? — Remy Guercio, Tailscale (Rubric AI4SDLC) I saw it. report Remy Guercio from Tailscale from AI Engineer Confa, 1 June 2026It’s about how to run AI agents into an organization if they need keys, accesses, MCP tools, and sometimes production data. Remy Guercio himself is engaged in strategic projects at Tailscale, which many people know as the “convenient VPN on WireGuard”. But the company has long positioned itself more broadly as a platform to connect with zero trust. (zero trust) It's based on identity.
- · #AI4SDLC
State of AI4SDLC on AI Dev Conf: how AI bridges development bottlenecks (Category AI4SDLC)
State of AI4SDLC on AI Dev Conf: how AI bridges development bottlenecks (Rubric AI4SDLC ) Appeared. notebook AI Dev Conf 2026. The point is simple. AI assistants made coding faster, but it didn’t speed up the entire development. The bottleneck did not disappear, but moved up the process: to the quality of the task statement, to the verification of the result and to the operating model of the engineering organization.
- · #AI4SDLC
Turbo ML Conf 18 July: Come to the State of AI4SDLC (Category AI4SDLC)
Turbo ML Conf 18 July: Come to the State of AI4SDLC (Rubric AI4SDLC ) 18 July will be held in Moscow Turbo ML Conf T-Bank conference for those who push the boundaries of ML and turn ideas into working products. I’m giving a talk on “State of AI4SDLC: How AI Removes Development Bottlenecks” and why it’s worth going there all day. First the report. AI in development has already become everyday: coding, revision, planning, debugging.
- · #AI4SDLC
Building OpenCode with Dax Raad: Conversation with Open Source Coding Agent Creator (Category AI4SDLC)
Building OpenCode with Dax Raad: Conversation with Open Source Coding Agent Creator (Rubric AI4SDLC ) I saw it. fresh-out The Pragmatic Engineer, where Gergely Orosz talks to Dax Raad, the creator of OpenCode. It’s not just about the history of the product. This is a rare case where the author of one of the fastest growing AI coding tools consistently refuses to sell magic and speaks about AI with noticeable skepticism. OpenCode is an open-source coding agent that started out as a terminal-based tool and grew to a GUI.
- · #AI4SDLC
BDD, ADR, PRD, WTF: Capturing Decisions for Humans and AI Alike — Michal Cichra, Safe Intelligence (Category AI4SDLC)
BDD, ADR, PRD, WTF: Capturing Decisions for Humans and AI Alike — Michal Cichra, Safe Intelligence (Rubric AI4SDLC ) Dealing with report Michal Cichra of Safe Intelligenc, with whom he spoke at AI Engineer Europe 2026 London 8-10 April 2026. I liked the report because the author looked at how agents break the implicit agreements of the teams and what to do about it.
- · #PlatformEngineering
IDP is Dead? No, the GUI monopoly is dying. (Category PlatformEngineering)
IDP is Dead? No, the GUI monopoly is dying. (Rubric PlatformEngineering ) I wrote it here. essay How AI is not just changing the interface of platform products, but the very nature of platform products... and what platform teams can do about it. I've been thinking a lot about this lately, and I've been looking at cases as well. Gitlab, Github, Atlassian
- · #AI4SDLC
SWE-rebench: Lessons from Evaluating Coding Agents — Ibragim Badertdinov, Nebius (Category AI4SDLC)
SWE-rebench: Lessons from Evaluating Coding Agents — Ibragim Badertdinov, Nebius (Rubric AI4SDLC ) I watched a short one. report Ibragim Badertdinov Nebius SWE-rebenchHow to honestly understand that the coding agent really knows how to solve software engineering problems, and not just beautifully passes the familiar benchmark. The main problem with old and static benchmarks is that they quickly cease to be fresh. Tasks are publicly available, get into training data, prompts, scaffolding, retry cycles and non-obvious fitting appear around them.
- · #AI
Thinking Machines: Why AI Models Should Learn to Interact in Real Time (AI column)
Thinking Machines: Why AI Models Should Learn to Interact in Real Time (Rubric AI ) Dealing with article Thinking Machines (Mira Muratti is a former CTO OpenAI company.) It is about interaction models, in which the authors say that the very way of communicating with AI becomes a limitation. Now the usual pattern of step-by-step interaction looks like this: the person asked the request, the model thought, the model answered. Even with a voice, camera, or agent harness, the logic is often the same inside.
- · #AI4SDLC
[2/2] GitLab Act 2How much it matches GitHub, Atlassian and Agent IDE (Category AI4SDLC)
\[2/2\] GitLab Act 2How much it matches GitHub, Atlassian and Agent IDE (Rubric AI4SDLC ) In past I took down the GitLab Act. 2 as an attempt to rebuild the DevSecOps platform for agent development: scale Git for machine use, implement the orchestration of the entire life cycle, collect a context graph, embed governance and implement a hybrid model of human/agent/autonomous work. And in this post, I wanted to compare it to what other platforms are doing and how much the course is aligned.
- · #AI4SDLC
[1/2] GitLab Act 2Why GitLab is Rebuilding Development for Agents (Category AI4SDLC)
\[1/2\] GitLab Act 2Why GitLab is Rebuilding Development for Agents (Rubric AI4SDLC ) Dealing with GitLab Act 2. Formally, this CEO letter is about the new stage of the company: focus, restructuring, AI and DevSecOps. But more importantly, GitLab does not describe a set of AI features, but a change in the production model of development. In short, GitLab no longer wants to be a DevSecOps platform to which AI has been pinned. He wants to become a control plane for SDLC, where not only people work, but also agents.
- · #AI4SDLC
Can LLMs generate Enterprise Quality Code? Why the pass rate is not enough (Category AI4SDLC)
Can LLMs generate Enterprise Quality Code? Why the pass rate is not enough (Rubric AI4SDLC ) An LLM can pass the tests and still write code that the enterprise team will take a long time to clean up. This is the report of Prasenjit Sarkar of Sonar:Can LLMs generate Enterprise Quality Code?" The main idea is simple, but unpleasant: pass rate Not enough anymore.
- · #Documentary
The Story of C++: The World's Most Consequential Programming Language The Official Story (Category Documentary)
The Story of C++: The World's Most Consequential Programming Language The Official Story (Rubric Documentary) Yesterday I left. film The official history of the C++ language, which most people don't see directly, but uses its results every day. C++ is not just another programming language. This is the invisible infrastructure of the modern world: games, operating systems, finance, hardware, graphics, simulations, high-load services. A language that is loved, scolded, and tried to be replaced, and that still carries a huge layer of digital civilization.
- · #AI4SDLC
Panel discussion “AI yesterday, today, tomorrow” with AI Dev Conf (Category AI4SDLC)
Panel discussion “AI yesterday, today, tomorrow” with AI Dev Conf (Rubric AI4SDLC ) Appeared. panelI have been working on AI Dev Conf. The composition of the participants was representative: Alexey Totmakov from VK Tech, Rafael Tonakanyan from Sber, Andrei Kuleshov from Yandex SourceCraft and me in the role of moderator. I think it was a good conversation about the fact that AI in development has already become basic hygiene: models help write code, tests, documentation, disassemble legacy, review and move faster on routine tasks.
- · #Management
Upper Management Meeting: When AI Looks Like Magic to Management (Category Management)
Upper Management Meeting: When AI Looks Like Magic to Management (Rubric Management ) I looked.Upper Management Meeting" It’s very funny and a bit sad because the caricature hits a familiar place: AI is really very easy to tell beautifully. Especially if you don’t understand how it works. This is the managerial danger. The problem is not that managers are interested in AI. This is just fine: technology is already affecting products, processes, development, support, analytics and cost of work.
- · #Management
Organizational evolution: From products to user needs Eugene Sergueev (Category Management)
Organizational evolution: From products to user needs Eugene Sergueev (Rubric Management ) I saw it. report Evgeny Sergeev talks about the evolution of the engineering organization at Flo Health, and I had a very strong sense of déjà vu. Summer. 6 Back in the day, I took a pretty similar path: turning an app into a super-app, Conway's Law, Conway's reverse maneuver, Team Topologies, trying to understand where team boundaries help a product, and where architecture and user experience begin to break. (There's one. My talk with Techlead Conf).
- · #Books
The Developer's Playbook for LLM Security (Books column)
The Developer's Playbook for LLM Security (Rubric Books ) I read this one. book by Steve Wilson, who was the director of the project.OWASP Top 10 for LLM Applications" He asked her how much the book was. 2024 The year on the safety of large language models is still relevant 2026-m. The answer is: as a basic engineering framework, yes, very much. As a complete overview of the current agenda - no longer, because the region has noticeably moved ahead in two years.
- · #Software
How a Group of Developers Took Back Control from Enterprise Java Spring: The Documentary (Category Software)
How a Group of Developers Took Back Control from Enterprise Java Spring: The Documentary (Rubric Software ) I watched it this weekend. spring-documentary. I love technology movies, which show not only what happened, but how much pain technology was trying to alleviate. Spring’s story is especially good: it’s not just a story about the popular Java framework, but a story about how developers got tired of the heavy enterprise model and began to regain control of the code.
- · #AI4SDLC
Sber released AI-Disrupt PDLC – a beautiful PDF and strange 140-page DOC (Category AI4SDLC)
Sber released AI-Disrupt PDLC – a beautiful PDF and strange 140-page DOC (Rubric AI4SDLC ) I read it the other day. whitepaper Collected about AI-Disrupt PDLC. This is the concept that AI changes the entire product development cycle: intent, context, specs, agents, harness, evals, governance, risk ladder, evidence bundle, tiny teams and so on.
- · #AI4SDLC
State of AI4SDLC (Category AI4SDLC)
State of AI4SDLC (Rubric AI4SDLC ) Appeared. record I gave a talk at DevOps Conf in April about “State of AI4SDLC” or what’s really going on with AI in development and why the story isn’t just about code generation.
- · #Books
Root Cause: Stories and Lessons from Two Decades of Backend Engineering Bugs (Books column)
Root Cause: Stories and Lessons from Two Decades of Backend Engineering Bugs (Rubric Books ) I read a few chapters. books Hussein Nasser, a popular blogger (500k subscribers channel)What does it say about software engineering? I took it out of mere interest to read not just about architecture, databases, and distributed systems, but rather how it all breaks down in the marketplace. In fact, the author promises these stories in the description of his book.
- · #SciFi
Atomic Heart. Background to “Enterprises” 3826How Soviet Technoutopia Was Built Before Failure (Category SciFi)
Atomic Heart. Background to “Enterprises” 3826How Soviet Technoutopia Was Built Before Failure (Rubric SciFi ) Continuing his immersion in the world of Atomic Heart, he read a special edition of Atomic Heart. Background to “Enterprises” 3826Harald Horf, which differs in color inserts and thicker and whiter paper from the usual edition. About the game itself.
- · #AI
Why AI Is Making Infrastructure a Management Theme (AI column)
Why AI Is Making Infrastructure a Management Theme (Rubric AI ) Went out a few days ago. RBC Why businesses choose hybrid infrastructure. In this interview, RBC from 19 In May, Yandex Cloud CEO Grigory Atrepiev said that the corporate software market in Russia 2025 year-round 808 The two main factors of market change are information security and artificial intelligence.
- · #Brain
No Drama - Color Lama (Brain column)
No Drama - Color Lama (Rubric Brain ) I have long noticed that my head works better if I combine thinking with physical action. That's why I often spend one-on-ones walking around our office. Then we had pencils and coloring sheets added to every negotiator in our office — I liked the idea and bought myself a coloring book. I use them now in some online meetings or just when I have to think deeply about something and my hands need to be occupied. The result is an art therapy that helps me focus.
- · #Architecture
AI Dev Podcast 2Alexander Polomodov, Sergey Baranov / Architecture in the Age of AI (Category Architecture)
AI Dev Podcast 2Alexander Polomodov, Sergey Baranov / Architecture in the Age of AI (Rubric Architecture ) About a month ago, we recorded. podcast together with Sergei Baranov n (founder of ArchDays) Andrey Dmitriev n (co-founder of JUG.RU) in preparation for Today's AI Dev Conf Conf. It just so happened that I went on vacation and forgot to share this episode with you, and it was interesting.) It was about how LLM and multi-agent systems are changing software architecture and engineering processes.
- · #Engineering
Spec-driven development (SDD)Why AI brought specifications back into development (Category Engineering)
Spec-driven development (SDD)Why AI brought specifications back into development (Rubric Engineering ) The hot topic of spec-driven development seems to me a reincarnation of old bearded approaches. V-Model or RUPIn the area of the two thousandth were used to describe engineering processes. I was wondering why this happened and why this post came out:) The old spec-driven approach was to first describe requirements, then design, then implementation, then verification.
- · #Conference
Tickets for AI Dev Conf (Category Conference)
Tickets for AI Dev Conf (Rubric Conference ) I'm on the conference program committee. AI Dev ConfAt the same time, I’m giving a keynote talk on the State of AI4SDLC. The conference itself will be held tomorrow, and I have a couple of tickets that I am ready to give to subscribers of the channel. In order to qualify for a ticket, you need to recommend interesting whitepapers about the implementation of AI in the development, metrics and results in the comments.
- · #AI
Gemini 3.5Google Bets on Agent Workflow (AI column)
Gemini 3.5Google Bets on Agent Workflow (Rubric AI ) Watched yesterday's. Gemini announcement 3.5 Google I/O. Google has officially published it. 19 May 2026 The big part here, I think, is not the next benchmark race, but how the company formulates the next layer of models. In short, Google introduced the Gemini family. 3.5 starting with 3.5 Flash. A model. 3.5 Pro, according to Google, is already in use internally and should go rollout next month.
- · #Books
Empire of AI (Books column)
Empire of AI (Rubric Books ) I finished reading a book yesterday.Empire of AI" from Karen Hao, which he bought from Foyles during a trip to London. In general, it was difficult for me to break away from reading this book and while flying to Moscow, I read it by a third - you seem to already know the main events around OpenAI, but in a coherent story they begin to look quite different. This book is not about ChatGPT as a product or how transformers are built.
- · #AI
State of AI4SDLC: how AI is shifting development bottlenecks
State of AI4SDLC: how AI is shifting development bottlenecks Already in 21 I will be speaking at the conference on Thursday. AI Dev Conf with this report, where I will continue to talk about State of AI4SDLC . It is now clear to everyone that AI in development has ceased to be just an experiment, since it is widely used for all major scenarios: coding, review, planning, debugging and working with documentation.
- · #Robotics
The history of robotics: not one line, but several engineering shifts (Category Robotics)
The history of robotics: not one line, but several engineering shifts (Rubric Robotics ) Continue. story I want to talk about why robotics is interesting now, I want to take a step back. The current wave of AI, humanoid robots, simulators and fundamental models looks new, but it stands on the shoulders of giants who have made history before. The story didn’t look like one straight line—it’s more useful to look at it as a series of steps that gradually evolved together: mechanical action, programmability, world perception, and autonomous behavior.
- · #Robotics
Why Robotics Is Now the Right Time (Category Robotics)
Why Robotics Is Now the Right Time (Rubric Robotics ) I decided to start a new column about robotics, which is now of great interest from different sides. But I wonder why it's so dense. The answer is that robotics suddenly found itself at the point where several long lines of engineering came together. 1ая Industrial base Robots have long ceased to be fiction.
- · #Agents
The Multi-Agent Architecture That Actually Ships — Luke Alvoeiro, Factory (Agents column)
The Multi-Agent Architecture That Actually Ships — Luke Alvoeiro, Factory (Rubric Agents ) I saw something interesting. video Luke Alvoeiro from Factory (theirs Droid The first place in terminal bench v1) A system where multiple AI agents conduct long engineering tasks: plan, write code, check, repair and do not lose context for hours or days. The idea is that if before the question was “can a model write code?”, now the question is, can the system autonomously bring the task to a working state and prove that the result is correct?
- · #AI
Everything I Learned Training Frontier Small Models - Maxime Labonne, Liquid AI (AI column)
Everything I Learned Training Frontier Small Models - Maxime Labonne, Liquid AI (Rubric AI ) I watched this short one. report Maxime Labonne, Head of Post-Training, Liquid AI, where Maxim talks about how Liquid AI designs and trains edge models from architecture to reinforcement learning. The main point of the report is that small models cannot be perceived as smaller versions of large models.
- · #Architecture
[2/2] Clojure: The Documentary (Category Architecture)
\[2/2\] Clojure: The Documentary (Rubric Architecture ) Continue. analysis I will tell you about the remaining interesting moments that I remember.
- · #Architecture
[1/2] Clojure: The Documentary (Category Architecture)
\[1/2\] Clojure: The Documentary (Rubric Architecture ) I saw a new one. CultRepo documentary about Clojure And it's the story of how Rich Hickey's view of complexity became the language, the community, the Datomic database and the production stack for companies like Nubank. (Which I've been talking about. told). Not only does Rich appear in the film, but the other key people who built the language, wrote books about it, implemented it in companies and used it in large systems.