iiinova — where your digital self keeps evolving

Lennart Nacke, PhD @lennartnacke.com · 28/08/2026

AI development is exponential. Our committees slowly review policies. AIs rapidly ingest entire libraries. And now, AI companies are mass-buying, scanning, and destroying old books. All to gain an edge. The Internet is full of slop.

A chaotic assortment of vintage leather-bound books stacked and scattered, highlighting concerns over their destruction by AI companies.

110

Lennart Nacke, PhD @lennartnacke.com · 27/08/2026

Richard Feynman once wrote on his blackboard: "What I cannot create, I do not understand." He didn't mean abstract theory. He meant building physical, testable systems from scratch. → 1965: Won the Nobel Prize in Physics for reconstructing quantum electrodynamics using visual diagram

A blackboard displays a handwritten quote about understanding and creation, alongside other text and markings.

101

Lennart Nacke, PhD @lennartnacke.com · 29/07/2026

Most people want the keynote. That sexy recognition. They don't see the work you've put in to get it. They see the citations, the title, the chair position. Many assume public expertise just ships with completing a PhD. Far from true. What they don't see:

An iceberg graphic illustrates the hidden efforts behind expertise, highlighting extensive work, research, and overcoming self-doubt.

110

Lennart Nacke, PhD @lennartnacke.com · 09/07/2026

Telling AI to be brief does not work to reduce tokens. I have spent the last two weeks obsessively optimizing my token count. So, I want to share it with you. With Claude Fable 5 coming out again (and being renewed once more) I'd just gotten sloppy with my prompts. It's just too damn smart.

Five methods to optimize Fable tokens, focusing on reducing noise and code length while maintaining meaning and efficiency.

250

Lennart Nacke, PhD @lennartnacke.com · 04/06/2026

Most experts already know they need to publish better quality. The blocker isn't time or ideas. It's the judgment that arrives the second you open a new document. Relocate that judgment. Let the rough version live first. Then edit. Spartaaaa!

120

Lennart Nacke, PhD @lennartnacke.com · 29/05/2026

200+ papers. 41,000 citations. Invisible. Lesser experts were winning clients I deserved. Staying current does not build your authority. Publishing does. Even badly. Even once a week. Consuming feels like staying keen. You read three posts from consultants in your space.

Bold text on an orange background emphasizes the importance of publishing for experts to avoid invisibility and gain recognition.

100

Lennart Nacke, PhD @lennartnacke.com · 27/05/2026

❌ Thought leader. We need to stop using that word. Not because visibility doesn't matter. It absolutely does. But because thought leadership rewires what AI experts write about. For the collaborator who needs proof first. For the hiring manager drowning in AI noise.

The phrase "thought leader" is crossed out, emphasizing "expert" in green text on a dark blue background.

110

Lennart Nacke, PhD @lennartnacke.com · 25/05/2026

Oxford just found AI agrees with who pushes hardest. Pressure alone gets you there. Not reasoning. AI folds 73% of the time.

Research on AI alignment highlights issues of sycophantic consensus, emphasizing the need for visible disagreement and principled revisions.

120

Lennart Nacke, PhD @lennartnacke.com · 23/05/2026

Weaker experts are taking your clients. You have 7 fixable reasons why. Most clients are roughly 70% through their decision before they talk to anyone in sales. They read and watch. Then they pick someone. If that someone isn't you, one of these bottlenecks is why:

A flowchart outlines seven authority bottlenecks hindering experts from becoming the obvious choice for clients, with improvement strategies.

100

Lennart Nacke, PhD @lennartnacke.com · 16/05/2026

Most academic founders kill their own ideas by pitching like it's a research paper. You wait for the committee. You polish the slides. You beg for permission. Here is how ideas actually get approved today: 1. Build a messy prototype. 2. Give it to a real user.

Two historic meetings contrasting the approval of ideas, featuring notable leaders in formal settings, showcasing different eras of leadership.

100

Lennart Nacke, PhD @lennartnacke.com · 16/05/2026

POV: AI-native expertise in 2026. Every few weeks there’s a new prompt pack, a BlueSky tactic that takes the edge off your judgement, or the inevitable “Experts are obsolete” take. But let's be real for a second: 1. Serious buyers will always search for visible expertise

A man sits distressed at a laptop surrounded by quotes on future AI trends and digital content, highlighting expert challenges in 2026.

100

Lennart Nacke, PhD @lennartnacke.com · 15/05/2026

Precision is how experts protect themselves. But it's not how they get remembered. When ChatGPT launched, I went silent. Not because I had nothing to say. I had more to say than almost anyone in my field. I'd spent 15 years studying human-computer interaction. AI was literally my jam.

100

Lennart Nacke, PhD @lennartnacke.com · 13/05/2026

We just won Best Paper at DIS 2026. Here's what VR companies are still getting wrong. But then: → Consent settings buried in headset-hostile menus → Data sharing framed as required for immersion → Headset discomfort exploited to rush consent → 2005 web design in a body-tracking medium

The document presents research on user experiences and privacy concerns in VR, highlighting deceptive design practices affecting consent and comfort.

100

Lennart Nacke, PhD @lennartnacke.com · 13/05/2026

Deep expertise has been declared dead since 2000. Some notable "deaths" include: • Google • MOOCs hit • LinkedIn Learning • Podcasts over professors • Gurus without credentials • AI writing tools everywhere • ChatGPT > all knowledge work • AI agents > all tech consultants

Graph illustrates the declining market value for expert knowledge from 2000 to 2026, marking key milestones in expertise replacement trends.

210

Lennart Nacke, PhD @lennartnacke.com · 10/05/2026

Expert founders: Remember that stretch where no one gave your ideas the attention they deserved? Your advisors half-assed it, investors ghosted, and you had to figure shit out solo. That wasn't a setback. It forged your edge.

A humorous question from Jolanta asking about the realization that a PhD supervisor might not understand the student's work.

100

Lennart Nacke, PhD @lennartnacke.com · 10/05/2026

People dismiss science as some elite puzzle too tangled for normal folks. I call Bullshit on this. It's just testing what works. Adopt this mindset: Question assumptions. Run small experiments in your daily grind. Track cause and effect.

210

Lennart Nacke, PhD @lennartnacke.com · 09/05/2026

You know that researcher who downloads 50 papers a week, tabs exploding like fireworks, but never actually reads or applies a damn thing? Yeah, that's procrastination. Sorry. One path: Hoard knowledge like a digital squirrel. Feel productive. Achieve zilch. The other:

A man gives a thumbs-up in a cluttered office, overwhelmed by piles of papers and two computer screens.

320

Lennart Nacke, PhD @lennartnacke.com · 09/05/2026

I ran a positioning exercise in my Authority Lab today. I asked them to name their enemy, because your enemy shapes your story. Without one, your positioning has nothing to push against. Silence.

Snow White interacts with forest animals in a cottage, expressing hope for love. An ominous witch appears in the final scene.

120

Lennart Nacke, PhD @lennartnacke.com · 09/05/2026

I never let AI call all the shots for me. When I'm chewing on a tough problem, I hit it with requests for options. Multiple angles. Different frameworks. Fresh takes. Complex shit rarely has one right answer. Hand over your judgment, and you're training yourself to stop thinking.

A man gestures refusal, stating he uses AI instead of a brain being offered by a hand, highlighting a preference for technology over human thought.

200

Lennart Nacke, PhD @lennartnacke.com · 04/05/2026

Most researchers use AI to edit their writing right away. But they shouldn't. Not because AI is bad. Because you've never heard what your own thinking sounds like unedited. And now you're letting something else clean it up before you have. You're outsourcing a skill you never owned.

The graphic contrasts personal writing with AI-generated text, emphasizing the importance of maintaining one's unique voice and ownership.

100

Lennart Nacke, PhD @lennartnacke.com · 30/04/2026

Most PhDs I know can explain quantum mechanics but can't write a thread that gets 10 retweets. Expertise in your field ≠ expertise in getting people to actually listen to you. Stay humble about what you don't know. Especially the stuff that feels beneath you.

100

Lennart Nacke, PhD @lennartnacke.com · 22/04/2026

CFOs finding out that $50K a month in API calls somehow costs more than the $45K salary they were hoping to cut.

010

Lennart Nacke, PhD @lennartnacke.com · 05/04/2026

I use AI 8+ hours a day. And sometimes it scares me. Not the productivity part. The way it becomes an obedient mirror. A new MIT study supports what I'd been noticing: AI sycophancy causes delusional spiralling. Even in highly reasoning users. We're talking perfectly rational brains.

A person wearing white sunglasses holds a research paper, illuminated by vibrant colored lights in the background.

100

Lennart Nacke, PhD @lennartnacke.com · 03/04/2026

I launched a community for expert founders yesterday. The first thing I asked every member wasn't "What's your niche?" or "Who's your audience?" It's not like we're doing bro marketing here. It was quite simple: Can you explain what you do in one sentence and without jargon? Most couldn't.

Participants engage in a live workshop focused on positioning for experts transitioning to entrepreneurship.

320

Lennart Nacke, PhD @lennartnacke.com · 22/03/2026

People see the keynote. The book deal & big following They don’t see the decade that made it all possible. Nobody sees the 10,000 hours of domain work you logged before you posted a single thing. Nobody sees the 3 years you spent writing to an audience of 12 people. Mostly your mom.

An iceberg graphic contrasts visible success with unseen efforts, illustrating the hidden work behind perceived expertise.

110

Lennart Nacke, PhD @lennartnacke.com · 18/03/2026

I've coached 30+ experts with the same flaw. They had depth. Not clarity. One had 20 years of experience. Amazing at the work. Strong client results. But when someone asked what she actually did, she froze. The positioning got vague. The message got fuzzy. And the value got lost.

The IDEA Framework outlines steps for experts to position knowledge, automate content creation, and acquire opportunities effectively.

100

Lennart Nacke, PhD @lennartnacke.com · 18/03/2026

000

Lennart Nacke, PhD @lennartnacke.com · 13/03/2026

AI thought leaders: "I've spent 19 hours testing 236 prompts across 6 models to find the optimal workflow for maximum productivity. Here are my 21 findings (thread)." Domain expert with 15 years experience after 1 hour with Gemini 3.1 Pro and Antigravity:

A man in a black suit and sunglasses performs a lively dance pose (Gangnam style) in a stable, surrounded by horses and a bright background.

020

Lennart Nacke, PhD @lennartnacke.com · 01/03/2026

Your team says they're frustrated with the project, but you don't really know what that looks like day to day.

Concentric emotion wheel categorizing core feelings (Happy, Sad, Angry, Fearful, Surprised, Disgusted) with layered, specific related emotions.

120

Lennart Nacke, PhD @lennartnacke.com · 01/03/2026

AI "experts" in 2026: 6 AM: Post that AI will take your job (now) 9 AM: Ask ChatGPT to write the next post 12 PM: Launch a masterclass on prompting 3 PM: Argue hard in the comments 6 PM: Zero shipped products Actual experts using AI: Got the work done with Claude Code. Went home.

A man, Tony Stark, in a tailored suit stands confidently with outstretched arms amid a dramatic explosion and rugged landscape.

020

Lennart Nacke, PhD @lennartnacke.com · 22/02/2026

Your best skill is slowly killing your expert career. And you can't even see it happening. You thought the competition was out of your league. Meanwhile, someone with 3 papers and a strong LinkedIn presence got the keynote instead. This is the Dunning-Kruger damage nobody talks about.

Graph illustrating the Dunning-Kruger effect, showing confidence levels versus experience, with a humorous tagline about awareness.

100

Lennart Nacke, PhD @lennartnacke.com · 20/02/2026

Germany wants real names on the internet. The same country that gave us the strictest privacy laws in the world. You can't have both maximum privacy and maximum accountability. Choose.

000

Lennart Nacke, PhD @lennartnacke.com · 18/02/2026

Your PhD won’t teach you how careers are built. Not in seminars. Not in advisor meetings. But every researcher who lasts figures it out. Usually after it costs them years. Graduate school has a hidden curriculum. Rules. Incentives. Power. Nobody puts it on the syllabus.

A diagram illustrates grad school teachings, contrasting visible syllabus elements with hidden skills crucial for career success.

110

Lennart Nacke, PhD @lennartnacke.com · 18/02/2026

Your research design will fail. Not because you’re careless. Because real data is messy. During my PhD failure 2 almost ruined a publication. And your method only looks rigorous on paper. The fix is boring:

Infographic outlining pitfalls in research design and strategies for early testing of assumptions to ensure reliable results.

111

Lennart Nacke, PhD @lennartnacke.com · 16/02/2026

How to be more productive (without another app): • Split your work into two buckets: 1. Efficiency (speed) 2. Effectiveness (return on effort) • Chase high-leverage opportunities (rewards compound over time) • Create systems so discipline isn't required for small decisions #productivity #phd

Flowchart outlining productivity strategies, distinguishing between efficiency and effectiveness with tips on discipline, tools, and high-leverage opportunities.

130

Lennart Nacke, PhD @lennartnacke.com · 09/02/2026

I've given you tons of free protocols yesterday. 13,734 researchers read that article. Most still won't publish faster. Here's the thing: saving a framework is not using a framework. You read my newsletters. You read the posts. You nodded. You hit the bookmark icon.

Left shelf cluttered with 39 saved posts crossed out; right organized six-week bookshelf labeled Week 1 Intro to Week 6 Submit with a green check.

100

Lennart Nacke, PhD @lennartnacke.com · 09/02/2026

James Patterson publishes 12+ books a year. Most authors just publish one. His secret? He stopped writing first drafts. Here's what PIs can learn from him:

Middle-aged man seated on a wooden floor surrounded by multiple paperback books arranged around him, wearing a navy sweater and looking up.

101

Lennart Nacke, PhD @lennartnacke.com · 07/02/2026

Most PhD students think their journey looks like this: Year 1: Review Literature Year 2: Collect Data Year 3: Analyze Year 4: Write Year 5: Defend Clean. Linear. Predictable. But that's not how research actually works. It's a mess. Here's what the real PhD timeline looks like...

Top: straight goal-to-achieve line. Bottom: tangled PhD journey with loops labeled setting a goal, starting, learning, doubting, practicing, feeling lost, struggling, getting it, achieving it.

100

Lennart Nacke, PhD @lennartnacke.com · 04/02/2026

Four professors apply for the same grant. One writes nonstop. One networks hard. One resubmits old work. One co-leads. Who wins?

Four professors vying for a grant: A stressed overload with paper stacks, a well-networked handshake at a reception, a tired archivist sorting files, and a virtual team meeting.

000

Lennart Nacke, PhD @lennartnacke.com · 03/02/2026

Nobody taught me how to write. I came up through engineering. Then a PhD where I struggled with my supervisor. The assumption was simple: Collect data and the paper writes itself. It doesn't.

Hands rest beside a keyboard, glasses, earbuds and ink bottle; red paper reads "NOBODY TAUGHT ME HOW TO WRITE" with arrow pointing right.

220

Lennart Nacke, PhD @lennartnacke.com · 01/02/2026

"So what?" Two words that kill grant proposals. If you can't answer them, you've confused your research problem with your research gap. Let me explain:

Solid circle labeled "Research Problems" with an arrow pointing to a dashed circle labeled "Research Gaps," on a dark background.

100

Lennart Nacke, PhD @lennartnacke.com · 01/02/2026

I'm skeptical of most AI advice on social media. Here's why: Salt and sugar look the same to the untrained eye. Someone can look smart on the surface. Polished threads. Confident takes.

Two hands sprinkling fine white grains with text warning to be careful who you trust, since salt and sugar look the same.

110

Lennart Nacke, PhD @lennartnacke.com · 30/01/2026

"I'm stressed" is academic for "I have no idea what's wrong." My PhD student said this during one of our weekly meetings. Third week in a row. So, I asked her to get specific. What exactly was stressing her? "Everything. The data. The writing. The timeline."

A colorful emotion wheel showing core feelings (sad, mad, scared, joyful, powerful, peaceful) with concentric rings of related, specific emotion words.

100

Lennart Nacke, PhD @lennartnacke.com · 29/01/2026

The research enterprise is sprinting toward a cliff. We're publishing faster than we can think. Reviewing faster than we can read. Citing faster than we can understand. Something's got to give.

100

Lennart Nacke, PhD @lennartnacke.com · 29/01/2026

Think of your lab as McDonald's. Not the food. The franchise system. Ray Kroc built a system so clear that anyone could deliver consistent results. Your lab needs the same approach:

Flowchart comparing McDonald's franchise system to a lab: boxes show training, documented process, ordinary worker/student producing consistent burger or publication-quality draft.

100

Lennart Nacke, PhD @lennartnacke.com · 28/01/2026

Prism looks like a gift to researchers. But in information economics, when the product is free, you're not the customer. You're the inventory.

Document editor showing a research paper with highlighted edits, a large "OpenAI's Prism" overlay, and a revision suggestions panel at bottom.

110

Lennart Nacke, PhD @lennartnacke.com · 22/01/2026

Most PIs spend 600 hours rewriting student drafts by Year 5. But there's a pattern that cuts this to 150 hours. The Dependency Multiplication Matrix:

Left column shows PI repeatedly spending 2 hrs per draft rewriting students' work, cumulative hours rising to 600 by Year 5; right column shows PI teaching students to self-improve, cutting time to 150 hrs.

130

Lennart Nacke, PhD @lennartnacke.com · 16/01/2026

5 steps to developing a point of view: 1. Map the terrain 2. Build bridges to existing work 3. Shift your lens 4. Weigh the evidence 5. Commit to your stance Do this, and panels will notice.

Colorful five-step infographic showing numbered panels with icons (lightbulb, gears, magnifying glass, checklist, target) guiding development of a point of view.

020

Lennart Nacke, PhD @lennartnacke.com · 07/01/2026

Three months of reading. Zero clarity. Nobody teaches PhD reading. It's a skill. Here's why reading in your PhD breaks people:

Tweet-style exchange: Rachel account notes "reading for pleasure implies reading for pain," Neil Renic, Twitter's favourite deeply sarcastic academic, replies "they're calling it a PhD," poking fun at PhD reading struggles.

110

Lennart Nacke, PhD @lennartnacke.com · 07/01/2026

Only 70% of PhD students ever finish. I've supervised dozens to completion. After 20+ years, here's the 5-step system that actually works:

Tweet text comparing finishing a PhD to rescuing a group project where your partner was your less-experienced self four years earlier.

110