iiinova — where your digital self keeps evolving

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

A stranger skims your 300 papers in one word: Teacher. WTF. You're holding office hours. You're rerunning the methods with a student. You're 15 years into the claim they hear in 5 seconds. Conduct the research they skipped over. Win the decade they booked as teaching.

000

Lennart Nacke, PhD @lennartnacke.com · 06/09/2026

Don't build your identity around having the right answer. Build it around the process of choosing one.

000

Lennart Nacke, PhD @lennartnacke.com · 06/09/2026

We chase certainty to avoid regretting our decisions. Turn to political promises. Read another paper. Ask AI. Eventually, you must choose based on what you know. It's uneasy. It's awkward. It's uncomfortable. Behind the painful distress lies satisfaction.

100

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

4. Write it in your own words. Reusable and good.

000

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

How to turn any published paper into a writing template: 10 minutes. ~600 words. 1 highlighter. 1. Find a best paper in your field. 2. Look for 1 sentence that describes: • Context (C) • Problem (P) • Method (M) • Result (R) • Contribution (K) 3. Get the skeleton and the style.

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

Instead of watching one hour of Netflix tonight, watch… NO! Just take the time off and watch/read/do whatever you want. Your work is hard enough without influencers telling you what dumb video will make you more intelligent. Enjoy your well-deserved break.

000

What matters more to you? Execution? or Product? Hit me up in the comments.

000

Understand that some non-pleasant parts still teach you. Never trade in learning opportunities for automation.

100

6. As you redo the process improve reference info 7. Redo only the parts of the process valuable to you 8. If AI does the parts not valuable to you → Review output 9. Review the quality of your process output This is how you can build valuable systems.

100

How to build a stellar AI-supported system: 1. Understand your start (A) and end points (B) 2. Capture the process of how you get from A to B 3. Review and examine your capture for improvements 4. Build checklists with AI for repeatable tasks 5. Structure where you save reference information

100

7. Feed rules into AI to generate the artifact → Iterate 8. Vet the final artifact based on audience/your standards In this case, the artifact matters more than the process. But the process still has some human iteration.

100

How to build a stellar AI artifact: 1. Do your own research on the topic 2. Gather the output quality criteria 3. Find examples of top benchmarks 4. Understand the artifact's audience 5. Iterate quality criteria to match expectations 6. Create rules based on the new criteria

100

Content mass production is killing creativity. AI has made us excellent at producing content. Chattie just lurves to help out. And we gobbled it up. The thing is: You don't understand the sh*t you didn't do. You cared about the artifact. You ignored the process. Two ways to fix this:

AI lets anyone perform intelligence. That must feel great if you're dumb. But it is so scary when you're smart. Who can you trust when you actually need deep advice? Everyone pretends to be an expert. We debate permissions. While AI builds our new reality.

Computation has outpaced governance. Everyone is mass-producing output. But was output always the goal? What about process? Wearing an exoskeleton to the gym will let you lift heavy. But will you build muscle? Nope.

These books are free from AI interferences. Coherent, edited, long-form reasoning. Better training for their models.

100

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.

Your drafts prove you have the time to write. Time lets you write those drafts. Permission lets you publish them. Grant yourself the second one.

000

Research is not coding. Writing is not an algorithm. Research is writing is thinking. Thinking needs creativity. Thinking needs friction. Thinking needs space. You can’t optimize this.

Don't phone in the work. Then you’re pulling the exact same stunt.

000

→ 1985: Published Surely You're Joking, Mr. Feynman!, turning lab stories into a 500,000-copy bestseller → 1986: Dropped a rubber O-ring into ice water on live television to expose NASA's management breakdown If you're using AI. Build actual rigorous systems. Understand them first.

100

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

• Contradictions between 2 competing research schools • Testing your thesis against known edge cases One churns out automated commodity slop. The other builds defensible intellectual property.

000

Bad AI lit review: • Generic prompt with automated review • Cherry-picked old studies randomly selected • Ignoring 20 years of theoretical debate in between Good AI lit review: • 50 hand-annotated PDFs from your private library run w/ a local model

100

• survive rejection. The defensive posture won you tenure. Now, it keeps your best thinking invisible.

It took a 5 years to master your discipline. It only took 3 anonymous reviewers to bury it. The institution trained you to impress them. Yet, you lack the clarity to feed your work to the world. You write to: • hide behind passive verbs. • cover every edge case. • avoid peer criticism.

Lennart Nacke, PhD @lennartnacke.com · 22/08/2026

When you hit tenure: What did the month after feel like?

000

Lennart Nacke, PhD @lennartnacke.com · 21/08/2026

Where do you personally draw the line on what AI is allowed to do in your work, and what already crosses it?

000

Lennart Nacke, PhD @lennartnacke.com · 12/08/2026

How anyone can get over the fear of criticism (this is what I did): 1. Publish about something you have a real opinion on 2. Collect the criticism that you get 3. Notice how little of it draws blood 4. Compare that sting to another year of keeping your knowledge silent Simple but powerful.

Relentless obsession on one restart at a time is required to improve each.

000

Your momentum only has 3 bottlenecks: 1. Restart size 2. Restart speed 3. Restart shame At any time, you must shrink one of these (and put every grand comeback plan on a Later list.) Create a smaller first post, get a faster return → zero apologies.

100

notice you had time to read this. The institution already took enough hours from you. Don’t donate the leftovers to more input.

000

Stop collecting another reading list. Stop bookmarking ways to write faster. Stop watching one more video about deep f*cking work. Stop consuming strategies for the paper you aren't writing. Just open the draft. And if your objection is that you don’t have time,

100

If the question matters, it has to win a block of time before the day starts performing urgency.

000

Most research days fail in a boring way. Not because the idea is weak. Because the first clean hour gets spent on work that feels responsible and leaves nothing behind. A meeting happens. An email gets answered. A form gets filed. The paper does not move.

100

They are necessary. They push your ideas past peer review. What looks too polished to the paranoid naked eye often looks like competence to a serious reader. Don’t dumb your prose down to perform humanity. Keep the rigour. Keep the clarity. It's your name on the paper, not Claude's.

020

Anyone who says that clear structure means it was written by AI is announcing they don’t read careful research. They don't appreciate the beauty in the mundane. Topic sentences, signposting, and clean abstracts existed long before chatbots came along.

Clarify the claim in speech first. Lay it bare. The page will stop fighting you.

000

Brutal truth: You can’t write the paper you can’t explain out loud. And yes, you can’t explain what you still only sort of understand. When you can't figure out your methods section, it's not a formatting problem. That is your mind telling you the design is still fuzzy.

Funders, collaborators, students, clients, and journalists do not search your hard drive. They rarely sift through archives. They encounter what you made easy to find.

000

Your expertise is already expensive to build. It becomes valuable when someone else can find it in time of need. Papers archive your work for a field. Public writing puts a grip on it.

One system. Repeated. Simple. Novelty feels like progress. Repetition creates your infrastructure.

000

The researchers getting real leverage from AI are not prompting all day. They're running agents. Hermes, Claude, Codex. Boring on purpose. Same literature dig on Monday. Same reviewer pushback script on Wednesday. Same clarity pass before coauthor review on Friday.

100

That is not a branding problem. It is a thinking problem.

000