Dr. Tali Režun

Web3 & AI Scientist, Vice Dean at @COTRUGLI & Co-founder; @BlockLabsLux | @4thtechProject | @immu3_io | @PollinationX_io | @RecordedPodcast

[Dr. Tali Režun @talirezun]

New Article from the From Lab to Life collection: Are We Becoming Obsolete, or Finally Free?

"After 30 years in tech and research, I've reached a profound conclusion about our AI future.

'The future belongs to those who can effectively collaborate with artificial intelligence, not those who fear or ignore it.'

But here's what most people miss: this isn't about becoming obsolete. It's about becoming more human.

In my latest article from the 'From Lab to Life' collection, I break down complex AI developments into language everyone can understand. Because the future of AI shouldn't be decided by tech experts alone - it affects all of us.

The transformation ahead offers an unexpected gift: the possibility of returning to what makes us most human. Meaningful conversations, empathy, community support - what I call the 'human touch' - this remains irreplaceable.

AI may process information faster than any human, but it cannot offer the warmth of genuine human connection or the deep satisfaction that comes from meaningful relationships.

I believe we're heading toward a world where humans become curators of experience, facilitators of connection, and guardians of meaning in an increasingly automated world.

This is part of my #FromLabToLife series - simplifying complex technology so everyone can understand what's coming and how to prepare.

Are we becoming obsolete or finally free? Read my full article: https://t.co/KwXcudUcUY

#AI #FromLabToLife #FutureOfHumanity #Technology #HumanPotential"

[Dr. Tali Režun @talirezun]

https://t.co/HgVStSS6Mr

[Dr. Tali Režun @talirezun]

🧠 The Curator — latest update v3.0.1-beta.18 is live
#TheCurator turns your documents and conversations into an interconnected knowledge wiki — compiled knowledge that compounds with every source, not retrieval-on-demand. The files stay plain Markdown, readable in Obsidian as a living graph.

This release came almost entirely from open-source community feedback. The headline:
🔧 The Health section is now an AI-powered maintenance system.
As a knowledge base grows past a few thousand pages, it accumulates broken links, orphan pages (nothing links to them), and duplicate concepts. Fixing those one by one is impossible at scale. Now the AI does the per-item judgement and you approve the fixes in batches — with a preview every time:
• Fix broken links — the AI re-points links to the right page, or cleans up the ones pointing nowhere (on one real wiki: 1,000+ → 0).
• Rescue orphans — it finds the best existing page to connect each isolated note to.
• Merge duplicates — clears near-identical pages in one reviewed pass.
(Plain-English version: a "clean up my wiki" button that's smart enough to ask before it touches anything — and everything's revertable.)

🛠 Plus reliability fixes from the community:
• Large documents no longer fail to import (and silently lose their content).
• Edits stop overwriting existing prose.
• Works across both Google Gemini and Anthropic Claude.

This matters most for Shared Brains — collective wikis for teams and organisations — where maintenance has to scale.

Every change is Git-tracked and validated against live data on both AI providers before shipping.

Huge thanks to the contributors who reported these. 🙏

🔗 The Curator open-source repo link in the comments (mac installer available)👇

#KnowledgeManagement #AI #SecondBrain #OpenSource #Obsidian

[Dr. Tali Režun @talirezun]

𝗧𝗢𝗞𝗘𝗡𝗦...𝘁𝗵𝗲 𝗻𝗲𝘄 𝗰𝘂𝗿𝗿𝗲𝗻𝗰𝘆 𝗼𝗳 𝗶𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝗰𝗲.
And the exchange rate just quietly changed on all of us.

There's growing noise about frontier models becoming more expensive. From my direct observation: it's true. But the story behind it matters more than the headline.

For two-plus years, companies like OpenAI and Anthropic have been subsidising your access to intelligence. A $20/month subscription delivered dramatically more compute value than the equivalent API bill would cost. I've seen both sides firsthand — the gap was significant, and it was intentional. Early adoption needed a price entry point that made sense for individuals and businesses.
That honeymoon period is ending.

Subscriptions are tightening. Usage limits are appearing. Rate caps are becoming real friction. And the reason isn't corporate greed — it's physics.
→ Compute costs money
→ Energy is not abundant
→ Data centres are not infinite
→ Running a 200K token context window at scale is extraordinarily expensive

Something had to give.

𝗦𝗼 𝘄𝗵𝗮𝘁 𝗱𝗼𝗲𝘀 𝘁𝗵𝗶𝘀 𝗺𝗲𝗮𝗻 𝗳𝗼𝗿 𝗽𝗿𝗮𝗰𝘁𝗶𝘁𝗶𝗼𝗻𝗲𝗿𝘀?
It means mastering context engineering — giving an LLM exactly what it needs, not more, not less — is no longer just best practice. It's economic discipline.

Here's what 2 years of building with coding agents has refined in my own workflow:
→ Every project begins with structured markdown files: architecture, blueprint, UI/UX, security notes. Complete. Precise. Permanent.
→ When a coding task runs, I don't throw the entire project's docs at the agent. I reference only the documents relevant to that specific task.
→ The result: dramatically lower token consumption, faster responses, measurably better output quality.
The foundational documentation isn't overhead. It's your token budget strategy.

The practitioners who thrive in the expensive-tokens era will be those who treat context as a resource to be engineered, not a convenience to be dumped.

Less context is not the answer. Precisely targeted context is.

Tokens are the new currency of intelligence. Spend them like it.

What's your approach to managing context costs in production workflows?

#AI #ContextEngineering #FromLabToLife #LLM #AIStrategy #TokenEfficiency #FrontierAI

[Dr. Tali Režun @talirezun]

Full text of Magnifica Humanitas (English, official Vatican publication):
https://t.co/Q38rHpq3iN

Available in nine languages: Arabic, German, Spanish, French, Italian, Polish, Portuguese, Russian, and English.

Five chapters. Introduction and Conclusion. 200+ numbered paragraphs covering: the history of Catholic Social Doctrine, the foundations of human dignity, the governance and ethics of AI, truth and democracy in the digital age, the transformation of work, new forms of digital slavery — and a direct prohibition on autonomous lethal weapons systems.

The most comprehensive institutional statement on AI and human dignity published to date. If you read one governance document on AI this year, this is it.

[Dr. Tali Režun @talirezun]

Pope Leo XIV published Magnifica Humanitas on May 15 — an encyclical on AI and human dignity addressed to the entire human family. Not a policy paper. Not a corporate framework. 200+ paragraphs across five chapters.

I read it cover to cover. And I keep coming back to one question it asks:

"Does AI make human life on earth more human in every aspect? Does it make it more worthy of man?"

I don't think we answer that question enough in our field.

What it actually says:
This isn't a rejection of technology. It's a demand for honest accounting.

→ Data is a common good. Algorithms, platforms, and digital infrastructure are explicitly classified alongside traditional goods — private ownership is not absolute. A 135-year-old principle extended into the digital age.
→ Lethal autonomous weapons can't be delegated to machines. "It is not permissible to entrust lethal or irreversible decisions to artificial systems." The question of who bears responsibility when a machine kills remains largely unanswered.
→ The "just war" framework is challenged as outdated. Dialogue, diplomacy and forgiveness are named as humanity's better tools now.
→ The attention economy is named as a form of dependency — not a side effect, but a deliberate feature.
→ Data colonialism is condemned. Health and demographic data extracted from vulnerable populations as a new form of domination.
→ Invisible AI labor is given moral standing. Data labelers, content moderators, rare-earth miners — the human chain that makes "seamless" AI possible.

Which of these surprised you most?

Why this matters beyond religion:
→ This is the first time an institution with 1.4 billion members has issued a detailed governance framework for AI — covering labor, weapons, data ownership, child safety, environmental cost, and epistemic manipulation in a single document.
→ It doesn't just call for ethical AI. It tries to name the structures that produce unethical AI.

I find that distinction more honest than most of what I read in tech. But I'm curious whether you do too.

I've been documenting AI's real-world impact in my From Lab to Life series. If there's one thing practitioners keep discovering, it's this: intelligence is no longer the bottleneck. Power, access, and accountability are.

This encyclical says the same thing from a very different direction.

Full text link in the comments — five chapters, worth the time.

#AI #FromLabToLife #DigitalEthics #Leadership #Technology

[Dr. Tali Režun @talirezun]

📖 Article links — read, follow, and star:

🔵 GitHub (open-source, free):
https://t.co/ipLvVkif9W

✍️ Medium:
https://t.co/2SbUw7ZUFG

📬 Substack:
https://t.co/aKvCKq6Y3H

———

⭐ The framework is open source — if it brings value, smash that star on the GitHub repo → https://t.co/pu7J62CMaU
Every star helps other founders and researchers discover it.

👥 Follow me on Medium for more practitioner research, field notes, and framework updates.

📩 Subscribe on Substack to get every new article directly in your inbox — Article 2 is already in progress.

This is a living framework. Your feedback shapes what gets built next.

[Dr. Tali Režun @talirezun]

𝟵𝟱% 𝗼𝗳 𝗲𝗻𝘁𝗲𝗿𝗽𝗿𝗶𝘀𝗲 𝗔𝗜 𝗽𝗶𝗹𝗼𝘁𝘀 𝗽𝗿𝗼𝗱𝘂𝗰𝗲 𝗻𝗼 𝗺𝗲𝗮𝘀𝘂𝗿𝗮𝗯𝗹𝗲 𝗯𝘂𝘀𝗶𝗻𝗲𝘀𝘀 𝗶𝗺𝗽𝗮𝗰𝘁.

⚠️ Not a technology problem. An organisation problem.

That's the MIT NANDA finding from August 2025 — and it's the founding premise of everything I've been building for the past three years.

Today I'm publishing Article-1 of the #ØØT Research Series: the complete overview of the Organisation of Tomorrow (ØØT) framework.

#ØØT is an open-source GitHub framework for partner-run, AI-augmented organisations.

👉 Not a consulting methodology.
👉 Not a SaaS platform.
👉 Not a framework you need a vendor to implement.

A complete operational stack — free to adopt, built on open standards, zero vendor lock-in.

The article "𝗕𝘂𝗶𝗹𝗱𝗶𝗻𝗴 𝘁𝗵𝗲 𝗢𝗿𝗴𝗮𝗻𝗶𝘇𝗮𝘁𝗶𝗼𝗻 𝗼𝗳 𝗧𝗼𝗺𝗼𝗿𝗿𝗼𝘄" covers everything a founder needs to understand before building or transforming:

→ The 5 theses behind #ØØT — grounded in MIT, McKinsey, Microsoft, HBS, and METR research
→ The Klarna Test — a mandatory blocking gate that prevents the failure mode Klarna made famous
→ Three generations: what's operational today, what ships in 6–12 months, and what's research-stage
→ Cloud track vs. privacy track — full operational parity, full data sovereignty, your choice
→ The Collecting Brain — why your firm's knowledge graph is its most valuable compounding asset
→ SKILL.md files — how to make AI methodology portable across every tool, forever
→ Three install paths — including a 60–90 min agent-assisted option for non-technical founders
→ Why this works beyond tech: any industry, any founder

This is my contribution from 30 years of building companies at the frontier of technology.

#ØØT is the first open-source organisational framework to address all four structural gaps simultaneously:
resistance management · output-based compensation · agentic knowledge infrastructure · data sovereignty

No comparable framework occupies all four corners.

If you're building from scratch or upgrading what you have — this is where to start.

🔗 All article links (GitHub · Medium · Substack) are in the comments below 👇

#OrganisationOfTomorrow #OpenSource #AI #FutureOfWork #AgenticAI #Entrepreneurship #AIStrategy #KnowledgeManagement #DigitalTransformation

[Dr. Tali Režun @talirezun]

𝗧𝗵𝗲 𝗺𝗮𝗰𝗵𝗶𝗻𝗲 𝗶𝘀 𝗿𝗲𝗮𝗱𝘆. 𝗧𝗵𝗲 𝗾𝘂𝗲𝘀𝘁𝗶𝗼𝗻 𝗶𝘀 — 𝗮𝗿𝗲 𝘆𝗼𝘂?

Coding agents that write production code. Design agents that ship interfaces. Research agents that synthesise thousands of pages in minutes. Automation pipelines that run while you sleep.

All of it exists. Right now. Today.

And yet most business professionals are still standing on the outside — watching the revolution happen to other people.

That's the gap #ChasingJarvis was built to close.

This is the first post in a series I'm launching — a visual summary of the Chasing Jarvis course, delivered as a mid-century magazine cover collection. Each cover captures one core module. One idea. One shift in how you think about AI.

Because here's what I've learned running this course across @COTRUGLI MBA cohorts:

→ The tools are not the hard part
→ The hard part is knowing what to ask, how to structure your intent, and how to give an AI agent the context it needs to actually deliver
→ That skill — context engineering — is the new competitive advantage. And it has nothing to do with coding.

VS Code. Google AI Studio. Augment Code. Claude. NotebookLM. Nano Banana Pro.

I teach all of it. Not as a technology course — as a business transformation course.

Because the professionals who learn to orchestrate these tools today are not just more productive.

They are building the companies the rest of the market hasn't imagined yet.

The series continues. Follow along and leave your thoughts in the comments. 👇

#ChasingJarvis #AIAgents #AITools #VanguardMBA #COTRUGLI #ContextEngineering #AIEducation

[Dr. Tali Režun @talirezun]

𝗖𝗵𝗮𝘀𝗶𝗻𝗴 𝗝𝗮𝗿𝘃𝗶𝘀 𝗕𝗲𝗹𝗴𝗿𝗮𝗱𝗲. 𝗗𝗼𝗻𝗲. 𝗔𝗻𝗱 𝘄𝗵𝗮𝘁 𝘁𝘄𝗼 𝗱𝗮𝘆𝘀 𝗶𝘁 𝘄𝗮𝘀. 🚀

Two days. Two cohorts of Vanguard MBA students. And MVPs that genuinely surprised me.

Factory floor optimization tools. Bank regulation compliance dashboards. Workflow automation systems. All built by business professionals. All powered by AI agents. All deployed with a live URL by end of Day 2.

This is what 𝗖𝗵𝗮𝘀𝗶𝗻𝗴 𝗝𝗮𝗿𝘃𝗶𝘀 is designed to prove — that the gap between idea and shipped product is no longer technical. It’s contextual. Give an AI agent the right documentation, the right structure, the right intent — and it builds.

The Belgrade cohort didn’t just learn that. They demonstrated it.

A huge thank you to the incredible @COTRUGLI Business School team for making this event possible — the logistics, the energy, the care for every detail. You know who you are. This doesn’t happen without you.

And the venue — the Tolstoy Members Club in Belgrade. Originally the ambassadors’ club of ex-Yugoslavia — a place built for diplomacy, where envoys from around the world once gathered. The lecture hall where we ran both days was carefully renovated but deliberately preserved in its original style from over 50 years ago. Dark wood, high ceilings, the gravitas of history in every corner.

There is something poetic about teaching the next generation of AI builders in a room where the world’s diplomats once negotiated the future.

The Chasing Jarvis journey continues. Next stop: coming soon.

#ChasingJarvis #AIAgents #VanguardMBA #COTRUGLI #ContextEngineering #Belgrade #MVPBuild #AIEducation

[Dr. Tali Režun @talirezun]

Chasing Jarvis. Day 1. Belgrade. In full effect. 🚀

We are halfway through an incredible day with the Vanguard MBA cohort at @COTRUGLI Business School — and the energy in the room is exactly why I build this course.

This morning we went deep on LLM foundations. Not theory for the sake of theory — the kind of understanding that changes how you work tomorrow. How tokens work. Why context windows matter. What makes an AI agent fundamentally different from a chatbot.

Then the first group work session hit.

NotebookLM. Audio podcasts generated in minutes. Claude Dashboards built in Artifacts. Business professionals — many of whom had never touched a developer tool — producing real outputs inside the first two hours.

This afternoon: AI Tools Deep Dive. VS Code, GitHub, Google AI Studio, @augmentcoden, Claude Code. Two tracks running in parallel — beginner-friendly and production-ready — because the goal is never to leave anyone behind.

The groups are strong. The questions are sharp. And the builds are already happening.

A few things I keep seeing prove true, cohort after cohort:

→ The technical barrier is almost always psychological, not real
→ The moment someone shares a live link they built themselves, something shifts permanently
→ Non-technical founders are not at a disadvantage — they are one context package away from shipping

Day 2 tomorrow: AI agents, MCP, context engineering, and every group deploys an MVP with a live URL.

#ChasingJarvis #AIAgents #VanguardMBA #COTRUGLI #AIEducation #ContextEngineering #Belgrade

[Dr. Tali Režun @talirezun]

🚀 Day 2. @COTRUGLI Chasing Jarvis. Belgrade.

Today we built the foundation — LLMs, AI tools, first live builds.

Tomorrow we will go further. Day 2 is where the real shift happens: from using AI to orchestrating it.

Here's what's on the roadmap 👇

📁 01 — AI Agents, MCP & Context Engineering
The conceptual leap from passive assistant to autonomous agent. Autonomous loops, MCP protocol, and why context engineering beats prompt engineering every time.

📁 02 — MVP Research & .md Bundle (Group Work)
Research your app idea. Validate market fit. Build your 4-file context package: CONCEPT.md · ARCHITECTURE.md · FEATURES.md · UI_UX.md

📁 03 — Lunch Break
Rest, reflect, and prep your .md bundle for the afternoon.

📁 04 — Track A & B: MVP Approach
Track A: Google AI Studio · Track B: Augment / Claude Code
A live 4-step build demo. Watch an app emerge from a context package.

📁 05 — Building Your MVP (Group Work)
Feed your .md bundle to a coding agent. Build → Test → Iterate → Deploy to Firebase. Share your live URL.

📁 06 — Wrap-Up & Next Steps
MVP presentations, course reflections, LinkedIn tribe launch — and a look at what comes next in the Chasing Jarvis journey.

By the end of today, every group will have shipped something real.

That's the whole point.

#ChasingJarvis #AIAgents #VanguardMBA #COTRUGLI #ContextEngineering #MVPBuild #AIEducation

[Dr. Tali Režun @talirezun]

🚀 Chasing Jarvis is landing in Belgrade.

This week, @COTRUGLI Business School brings its AI Agents & Entrepreneurial Innovation program to the Vanguard MBA cohort — and Day 1 kicks off on Wednesday.

For those who haven't followed the journey: Chasing Jarvis is built on a single premise — we are learning to build and ship real digital products using AI coding agents.

Here's what Day 1 looks like 👇

📁 01 — Intro to LLMs & AI Tools
From the history of AI to token prediction, neural networks, and context windows. The foundation everything else builds on.

📁 02 — NotebookLM + Interactive Dashboards (Group Work)
Research with NotebookLM, generate an audio podcast, build a Claude Dashboard in Artifacts.

📁 03 — Lunch Break
Rest, network, and prepare for the afternoon.

📁 04 — AI Tools Deep Dive
VS Code, GitHub, Google AI Studio, @augmentcode, Claude Code, @lmstudio, Nano Banana Pro — the full toolkit.

📁 05 — Let's Code! (Group Work)
Track A: Google AI Studio · Track B: Augment or Claude Code / @antigravity
Build something. Share a live link.

📁 06 — Wrap-Up Day 1
Group demos, key learnings, Day 2 preview.

Day 2 goes deeper — AI agents, MCP, context engineering, and live MVP builds.

If you're building something with AI or teaching others to do the same — follow along. The roadmap is just getting started.

#ChasingJarvis #AIAgents #VanguardMBA #COTRUGLI #ContextEngineering #AIEducation

[Dr. Tali Režun @talirezun]

👉 #TheCurator open-source repo: https://t.co/HgVStSS6Mr

👉 #TheCurator MAC installer: https://t.co/1FKhjw9ju9

[Dr. Tali Režun @talirezun]

🧠 𝗦𝗶𝘅 𝗿𝗲𝗹𝗲𝗮𝘀𝗲𝘀 𝗶𝗻 𝘁𝗵𝗿𝗲𝗲 𝘄𝗲𝗲𝗸𝘀. 𝗛𝗲𝗿𝗲'𝘀 𝘄𝗵𝗮𝘁 𝗰𝗵𝗮𝗻𝗴𝗲𝗱 𝗶𝗻 𝗧𝗵𝗲 𝗖𝘂𝗿𝗮𝘁𝗼𝗿 𝟯.𝟬.𝟭. 𝗕𝗲𝘁𝗮 𝟭𝟯

For those new here: #TheCurator is the free, open-source app I built to turn the PDFs, articles, and notes you read into a living, interconnected knowledge wiki — fully searchable, chattable, growing smarter with every source. Local. No subscription. Your data stays on your machine.

Three weeks ago, community members started using it at real scale and hit edge cases I'd never seen on my own setup. Their bug reports drove a wave of fixes. Sharing the highlights because most of them affect everyone.

🔍 𝗖𝗵𝗮𝘁 𝗷𝘂𝘀𝘁 𝗴𝗼𝘁 𝗱𝗿𝗮𝗺𝗮𝘁𝗶𝗰𝗮𝗹𝗹𝘆 𝘀𝗺𝗮𝗿𝘁𝗲𝗿 𝗼𝗻 𝗹𝗮𝗿𝗴𝗲 𝘄𝗶𝗸𝗶𝘀
Question: "What articles have I ingested by Dr. Tali Rezun?" — the chat returned 6 articles. Turned out the chat was loading only a fraction of the wiki on mature domains. Critical bug. Fixed.
Now when you ask about a specific person or topic, #TheCurator detects the "pivot" — loads that entity's page plus every article connected to it — and figures out whether you want a complete list or a focused synthesis.

Same question. Same wiki. 33+ articles now. Roughly 5× more comprehensive answers.

🛡 𝗦𝗺𝗮𝗿𝘁𝗲𝗿 𝗮𝘂𝘁𝗼𝗺𝗮𝘁𝗶𝗰 𝗶𝗻𝗴𝗲𝘀𝘁 𝘀𝗮𝗳𝗲𝗴𝘂𝗮𝗿𝗱𝘀
When the AI reads a new source, #TheCurator now catches and repairs more of its own mistakes automatically:
→ Creates missing parent concept pages when the AI describes sub-ideas but forgets the umbrella
→ Wikilinks items inside hub pages so they actually connect in your knowledge graph
→ Auto-merges near-duplicate concepts ("expert-roundup" vs "experts-roundup") before they pile up

📊 𝗖𝗹𝗲𝗮𝗿𝗲𝗿 𝗶𝗻𝗴𝗲𝘀𝘁 𝗿𝗲𝗽𝗼𝗿𝘁𝘀
The old "Ingest finished with 12 warnings" message was scaring people. Most of those were actually successes — the safeguards reporting what they caught and fixed.

Now each entry is color-coded: ✓ Auto-fixed / ⚠ For review / ⚠ Attention. At a glance, you know exactly what (if anything) needs you.

🔒 𝗡𝗼 𝗺𝗼𝗿𝗲 𝗹𝗼𝘀𝘁 𝗳𝗶𝗹𝗲𝘀
→ Wiki pages can no longer be corrupted by clicking Update or Sync during an ingest
→ A process killed mid-write now leaves the old file intact — never zero-byte
→ Cryptic npm errors now include the exact next-step command
→ Auto-restart after updates is more reliable across machines

✅ 𝗠𝗼𝗿𝗲 𝘁𝗲𝘀𝘁𝘀, 𝗹𝗲𝘀𝘀 𝗴𝘂𝗲𝘀𝘀𝘄𝗼𝗿𝗸
Test coverage grew from 365 to 791 assertions across the critical paths. The chat has dedicated tests now. The ingest pipeline runs stress tests against live AI to verify quality contracts before every release.

Big thank you to the community members who put in the time to debug, diagnose, and write up issues in an actionable form. Three of these bugs would have stayed hidden for months without your sharp eye.

If you've been waiting to try the latest #Curator, this is a good moment.

If you're already using it — Settings → Check for Updates. If you are new, there is a link to the MAC installer in the comments 👇

PS: Smash a GitHub STAR to support the project 🦾

#TheCurator #SecondBrain #KnowledgeManagement #AITooling #OpenSource #Obsidian

[Dr. Tali Režun @talirezun]

Every time someone hits Claude's usage limits, I see a similar root cause.

Not too many messages. Not the wrong plan tier.

Too much context chaos.

And here's the thing — this is exactly what I've been teaching in my context engineering work for the past year. The limits problem isn't new. It's the same problem I keep seeing derail AI coding agent sessions, bloated CLAUDE.md files, and multi-agent builds that collapse at Phase 2.

The scoreboard isn't message count. It's context weight.

Every stale conversation thread, every pasted document, every tool connector running in the background that Claude has to carry into the next answer — it all draws from the same pool. And yes, https://t.co/8R2sVI009l, Claude Code, Claude Desktop, and CoWork all share a single usage budget.

→ The practitioners who never hit limits aren't the ones on Max plans.

→ They're the ones who treat context as infrastructure.

This is the core of what I teach in context engineering: before you build anything, build your documentation. Not as an afterthought — as the foundation.

A structured library of lean, up-to-date markdown files covering architecture, requirements, UI/UX decisions, security constraints, and project conventions is not just good engineering practice. It turns out it's also the most efficient way to feed an AI agent exactly what it needs — nothing more, nothing less.

Here's what this looks like in practice:

→ One markdown file per concern. Architecture is architecture. Security is security. Don't mix them into a single bloated instruction file that loads every time.

→ Files stay current. A context file that describes last month's architecture is worse than no context file — it's misinformation. Updating documentation is part of the build process, not optional cleanup.

→ The agent reads what it needs, when it needs it. A well-structured context library means you point Claude at the relevant file for the task at hand. You're not pasting. You're referencing.

→ /compact and /clear become safe. When your decisions live in files, clearing a conversation costs you nothing. The knowledge is preserved. The chat history is just noise.

For beginners: start with three files. A project brief (what you're building and why), a decisions log (what you've chosen and why), and a current status file (what's done, what's next). That alone will transform your sessions.

For advanced builders in Claude Code, CoWork, or Claude in Chrome: your CLAUDE.md should be a map, not a manual. Durable rules and file pointers only. Everything else lives in the files it points to.

Anthropic confirmed this direction in May 2026 — doubling Claude Code limits for Pro and Max users. More headroom helps. But clean context architecture is what makes that headroom last.

The limit is a signal. It's telling you your context isn't structured enough yet.

That's a solvable problem. And the solution is the same one I've been teaching all along.

[Dr. Tali Režun @talirezun]

Chasing Jarvis is coming to Belgrade. 🇷🇸

When I designed this course at @COTRUGLI, the core thesis was simple: experts who learn to work with AI don't just work faster — they produce a fundamentally different category of output.

There's a name for this: the Centaur model.

The human contributes spec, taste, judgement, and the context the model hasn't seen. The AI contributes implementation at speed. Neither is sufficient alone. Together, they form something that outperforms both.

This isn't intuition. The evidence is converging from multiple directions:
→ HBS Cybernetic Teammate (Dell'Acqua et al., 2025, n=776 at P&G): individuals with AI matched the output of two-person teams without AI. Teams with AI produced more top-tier solutions. AI broke down silos between R&D and commercial roles.
→ DORA Report 2025: AI-assisted teams show measurable lifts in delivery speed — but only when paired with mature engineering practices. Without those, AI accelerates the production of bugs.
→ Karpathy, Software 3.0 (Sequoia AI Ascent 2026): "Spec is human, implementation is AI, review is human." Agents are like intern entities — they need taste and judgement from the human in the loop.
→ HBS + MIT NANDA + DORA + Microsoft Frontier Firm 2025 all converge on the same finding: the highest-performing teams in 2025–2026 are neither AI-sceptical nor AI-maximalist. They redesigned their workflows around the Centaur pattern.

Chasing Jarvis is that redesign — applied to business professionals.

Not a prompt engineering workshop. Not AI literacy 101. A hands-on course where you build real products, using real agents, grounded in a real understanding of what the human brings that the model cannot.
Belgrade, see you there. 👇

[Dr. Tali Režun @talirezun]

🚀 𝗡𝗲𝘅𝘁 𝗪𝗲𝗱𝗻𝗲𝘀𝗱𝗮𝘆 & 𝗧𝗵𝘂𝗿𝘀𝗱𝗮𝘆 — 𝗖𝗵𝗮𝘀𝗶𝗻𝗴 𝗝𝗮𝗿𝘃𝗶𝘀 𝗴𝗼𝗲𝘀 𝗹𝗶𝘃𝗲 𝗶𝗻 𝗕𝗲𝗹𝗴𝗿𝗮𝗱𝗲.

On May 27–28 at Tolstoy, Belgrade, my colleagues from @COTRUGLI Business School and I are taking a room full of students on a two-day, hands-on AI and AI agent journey.

No theory for the sake of theory. We build.

📍 What the two days look like:

▸ Day 1 — Foundations & Tools
We start with LLMs — how they actually work, from token prediction to scaling laws to context windows. Then we move into the full AI tool landscape: NotebookLM, Claude Desktop, Google AI Studio, VS Code, GitHub, Augment Code, Claude Code, Nano Banana Pro. Students don't just hear about these tools — they use them live.

▸ Day 2 — AI Agents, MCP & Building Your MVP
The conceptual leap from LLM to autonomous AI agent. Model Context Protocol — what it is, why it matters, and how to connect agents to real-world tools. Context engineering as the highest-leverage skill in the AI toolkit. And then: students build and deploy their own MVPs.

🛠 Two tracks:
→ Track A (beginner-friendly) — Google AI Studio, Gemini API, rapid prototyping
→ Track B (advanced) — Claude Code, @antigravity, @augmentcode, @opencode production workflows

Every group leaves with a working app and a live link.

This is what I believe business education needs to look like in 2026. Not slides. Not case studies about AI. Actual building — by business professionals, in real time, using the same tools that are reshaping industries right now.

And if you're a business leader thinking about how to upskill your team on AI agents and practical AI implementation — this is the format that works.

#ChasingJarvis #AIAgents #ContextEngineering #AIEducation #COTRUGLI #VanguardMBA #AITools #ClaudeCode #MVP #Belgrade

[Dr. Tali Režun @talirezun]

👉 Medium: https://t.co/rnGxbtd0TG

👉 Substack: https://t.co/fW5Cvle7TQ

👉 GitHub: https://t.co/pspZyZYNWY

[Dr. Tali Režun @talirezun]

What if your team's collective intelligence never retired, quit, or forgot?

Every organisation leaks knowledge.

The senior partner who leaves takes 15 years of client patterns with her. The PhD student who graduates takes 500 papers with him. The engineer who moves teams takes the why behind every architectural decision — leaving only the what.

We've been building a fix.

#TheCurator v3.0 introduces Shared Brain — a collective, compounding knowledge graph that grows smarter every time any member of your team reads, researches, or learns.

Here's what that looks like in practice:
🔬 Research teams 4 researchers × 20 papers/week × 50 weeks = 4,000 papers/year. No one person reads them all. With Shared Brain, every ingested paper compounds into a collective wiki — contradictions flagged, concepts cross-linked, sources cited. Friday's meeting goes from 2 vague hours to 30 focused minutes.

🎓 University cohorts 20 students. One semester. Five papers each per week. By December: a 500-page collective wiki that no individual could have built alone — every concept cross-referenced, every contributor's insight attributed. The cohort's shared understanding, queryable forever.

🏢 Consulting firms "Why did we choose this approach for a similar client in 2022?" — answered in seconds. Institutional memory that survives partner departures. New hires productive in days, not months. The firm's real competitive moat is the patterns it has recognised over years. The brain remembers them.

💡 Engineers & product teams "Why did we pick PostgreSQL over MongoDB for the auth service?" — Claude reads the collective via MCP, returns the answer with a citation to a 2023 architectural decision record. That's not search. That's reasoning over your team's accumulated knowledge.

Read more in the latest article "The Shared Brain: When Second Brains Start Thinking Together" writen by @DrazenKapusta and me. @COTRUGLI Business School COlab · The Curator Research Series · May 2026

Full article links in the comments 👇