Forget AI Agents. You Need an AI Exoskeleton.

9:17

Forget AI Agents. You Need an AI Exoskeleton.

Tiago Forte 24.03.2026 11 657 просмотров 549 лайков

Machine-readable: Markdown · JSON API · Site index

Смотреть на YouTube

Поделиться Telegram VK Бот

Транскрипт Скачать .md

Анализ с AI

Описание видео

🔗 Join the Founding Cohort: https://bit.ly/4c0WyAk I've spent 1,200 days testing one question: what does building a Second Brain actually mean now that AI exists? In this video, I share four things I believe that most AI creators disagree with — including why autonomous agents are the wrong bet, and why your notes are more valuable now, not less. 📩 Read my blog post “Introducing The AI Second Brain”: https://fortelabs.com/blog/introducing-the-ai-second-brain/ Resources: Harvard Business School / Procter & Gamble AI study: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5188231&__cf_chl_tk=c1XEOteeyR6IGrMpAwryu8mTAw8Nl11nwGMxzGPXDz0-1774264146-1.0.1.1-2Az9F222lS0nrPxn5ruKtm4vD6Kr8hxqdm_broQodxo CHAPTERS: 0:00 Why I stopped teaching for 1,200 days 0:29 The question I've been trying to answer 0:54 AI Agents 1:13 The cognitive exoskeleton 1:50 What happened on February 5 3:03 Contrarian Take #1 3:53 Contrarian Take #2 5:29 Contrarian Take #3 6:25 Contrarian Take #4 7:43 The AI Second Brain

Оглавление (10 сегментов)

Why I stopped teaching for 1,200 days

You may remember what happened in early 2023. After 6 years and 6,000 students, I stopped teaching Building a Second Brain. It wasn't a clean, strategic pivot. Something about that moment felt existential to me, not just professionally, but personally. — It forced me to return to the bedrock of who I am and what I stand for. —

The question I've been trying to answer

— It's now been 1,200 days, and I've spent that time trying to answer one question. What does it actually mean to extend your mind with technology, now that AI can perform most knowledge work? — And what I kept finding, there are two very different visions of what AI is supposed to do for us.

AI Agents

The first one is everywhere right now. It's the vision of AI as a replacement. Autonomous agents doing your work for you, taking over tasks, running in the background while you step aside. — AI as a substitute for human effort and judgment.

The cognitive exoskeleton

The second vision is one I first wrote about way back in 2015, before large language models, before ChatGPT, before any of this. I described it as a cognitive exoskeleton, both protecting us from the ravages of forgetfulness and amplifying our blows as we take on creative challenges. An exoskeleton doesn't replace the person wearing it. It makes them more powerful, more capable, more themselves. That's the vision I've been building toward for a decade now, and I think we're finally there.

What happened on February 5

On February 5th of this year, something happened in the AI world that I think changed the equation. OpenAI and Anthropic, the two leading AI companies, both released a new generation of models. For the first time, AI could read your files, use tools, take real actions on your computer. These aren't just chat interfaces. — They're what we call agent harnesses. What I always noticed is that as long as AI was mostly about chatting, — you couldn't build a real system around it. Each conversation started from zero. No memory, no continuity, and therefore no compounding. The tools and the environment that you give AI to work with matter just as much as its raw intelligence. Just as with humans, by the way. In 2021, I wrote in my book Building a Second Brain, "I began to see my second brain not just as a note-taking tool, but as a loyal confidant and thought partner. " At the time, that was just a metaphor. Now, it's something you can actually build.

Contrarian Take #1

My first contrarian take is that agents are the wrong bet. There's an enormous amount of time and energy going into autonomous AI agents right now. The idea that you set something in motion, step back, and it does the work for you. I've spent a lot of time testing this, and here's my honest assessment. The potential of agents is dramatically overblown. They're way harder to build than everyone claims. They're far less reliable than advertised, and they take enormous amounts of time and effort to maintain. More importantly, they're just the wrong metaphor. The right metaphor is an exoskeleton, not a replacement for your effort, but an amplifier of it. You're still in the suit, you're just stronger.

Contrarian Take #2

stronger. My second contrarian take is that it's not about doing more. It's about experiencing more. There's a recent study I keep coming back to. Researchers at Harvard Business School ran this carefully designed experiment with 776 professionals at Procter & Gamble, in which they pitted individuals working with AI against two-person teams working without it. A single person with AI reliably matched the performance of a two-person team. They worked faster. They produced higher-quality work. And they better integrated outside perspective. But here's the part that was really surprising, and I think contradicts many people's assumptions. Those working with AI reported significantly higher levels of positive emotions. For example, excitement, energy, and enthusiasm, and lower anxiety and frustration versus those working without AI. They didn't just perform better according to external metrics. — They had more fun. That's what amplification looks like. Not just offloading work, but doing better work and enjoying it more. Oh, and by the way, the model used in that study was GPT-4o, released in May 2024. That is now nine generations ago. Ancient technology by AI standards. The models are far more powerful and fun to use today.

Contrarian Take #3

My contrarian take number three is that your notes aren't obsolete. They're your most valuable asset. The question I hear most often from our community these days, "If AI can generate any content on demand, what's the point of taking notes? " And here's my answer. Personal knowledge management, PKM, is the best preparation for succeeding with AI I can imagine. If AI is like compound interest for your knowledge, then you've already made a deposit if you take notes. You're not starting from zero. Everything we've learned and built and discovered together over the last decade was a prelude to this. The foundation still matters more than ever. You still need a reliable way to capture and feed new ideas into your system, if that is, you want it to produce an answer that's different from what anyone else in the world would receive. My

Contrarian Take #4

contrarian take number four, you have more time than you think. Everything with AI these days is presented with this breathless urgency. Adopt the latest thing right now or join the permanent underclass. I differ from almost every AI content creator on this point. I'm confident we have plenty of time. An estimated 84% of the world population still hasn't adopted generative AI. If you pay anything for any AI product, you're roughly one in 133 people worldwide who do so. If you are watching this video, you are so incredibly early to this. The history of our relationship with information has evolved from papyrus to hard drives over centuries. It's not going to be completely upended in the next 36 months. The biggest risk right now, I believe, isn't falling behind. It's succumbing to fear. The moment you give in to fear, learning stops. You lose your judgment, and you start making decisions that are rooted in short-term survival. You have time, but use it wisely. This is an invitation between people who

The AI Second Brain

share a worldview, not a pitch to strangers. This April, I'm opening the very first cohort of the AI Second Brain. Three weeks live, hands-on, from April 15th to May 1st. I will be there personally for every session, every Q& A, every piece of feedback. What we'll be building isn't a short-term productivity hack. It's the fulfillment of everything we've been working toward together. A real AI-powered second brain, customized to how your mind actually works. You've been part of this movement long enough to know the difference between a shiny new tool and a genuine shift. This is the shift. The founding cohort opens March 26th. If you've been waiting for the right moment to go deep on AI, this is it. I'll compress literally thousands of hours of learning by me and my team into just three weeks, and bring you to the frontier of what AI is capable of now in your own work and life. The link is in the description. I hope you'll join us.

Другие видео автора — Tiago Forte

Ctrl+V

Экстракт Знаний в Telegram

Экстракты и дистилляты из лучших YouTube-каналов — сразу после публикации.

Подписаться

Лучшие методички за неделю — каждый понедельник