Everything I publish, in the order I published it: dispatches from Twitter, videos from YouTube and notes from my vault, archived here in full — text, images, video and links — so they outlive the platforms.
922 dispatches since January 2012 · page 24 of 37

I, Alexis, want a better coding tool for my qualitative analysis. …I also want to learn Swift.
Question: WHO? Answer: Me, Alexis Rondeau. Question: Who are you? Answer: I’m many things, but in this context, I’m a qualitative researcher.
As a…
I want to…
Code transcripts
and Compare interviewee segments
So that…
I can deliver
Because…
my clients want to
Unit/Metrics:
“An Insights Development Environment (IDE)”
“For you”
DeepDive is…
Native Mac application
It’s built to be 100% Customizable
“Forever”
Tag fragments
Auto-code
Like DelveTool but…
 







   
 





 
 





 
 


journey
section Discover a missing answer
Create a learning plan: 5: Alexis, Courtney, Client
section Run interviews
Create a interview guide: 3: Alexis, Courtney
Source and schedule participants: 3: Alexis, Courtney
Moderate interviews: 9: Alexis
section Analyse interviews
Read transcripts: 4: Alexis
Code transcripts: 2: Alexis
Re-code transcripts: 3: Alexis
section Create report
Create slide deck: 3: Alexis, Courtney
section Deliver report
Present slide deck: 5: Alexis, Courtney, Client
section Make decision
Present slide deck: 5: Alexis, Courtney, Client
[^1]: Start with WHO not How
Fix limits of conversational memory
Create ChatGPT API System Prompts
2/ & 3/ we don’t have any plans for an affiliate program or a Pro API key at this time (the API is very much in alpha)
Because: Co-editing created the best papers and most impressed students
Shortcodes / Auto-expands
#list → Adds “Make this a list”#md → Adds “Create headings”#nex → Adds “No explanations”Email an das YC Team
DONE Improve: “Chat”-UX
HELP THEM AVOID RUNNING INTO AN ERROR
FIRST TIME / On startup
Later
Business
Pay for extra features
Integrate easy-to-use design system
Expand on privacy policy
We don’t track you, but you shouldn’t trust us by just tell you. Here’s how you can investigate yourself
[^1]: 
TLDR; After getting frustrated, angry even, with how awkward and unreliable the “touch-screen-to-flipt-to-next-page” on the Kindle Classic is, I experimented with different ways to emulate a physical ‘next-page’ button akin to those found on more expensive devices like the Kindle Oasis. My ideal solution turned out to be a simple, twisted rubber band.


Turning/flipping to the next page BREAKS MY READING FLOW because…
Then ONLY HALF (50%) of presses don’t register 🔥🤬🔥

Motion to trigger page turn is effortless
The twist in the band is more reliable





 





Reliability (%) = "Thumb Resting Sweet Spot" + "Thumb Size" + "Twist Segment Hot Spot Size"

This is where you place your thumb most of the time and relax it inbetween page turns. In terms of importance, to me, everything else in terms of usability derives from this. In my case, it’s in the upper-half of the lowest third of the sidebar because of how I hold the Kindle.

This of course heavily depends on your individual thumb. Its size is in direct relation to how many twists your band can have. Once there are too many twists (max 5 for me), the segments become too small.
This is how much vertical space each twist segment provides to register the touch/swipe motion. With each twist, the segments get smaller. For me 5 twists were the maximum. I learned that only the the flat middle-part works reliably when pressed.


I have achieved 90+% RELIABLE PAGE TURNS once I harmonized my personal…
[^1]: https://en.wikipedia.org/wiki/Conductive_elastomer, https://www.amazon.de/-/en/MobiLinyi-pieces-replacement-premium-stylus-6-mm/dp/B0797Y7M24/



It’s a disappointment because:
All the company (and the competitors in the market) can come up with is…
It makes me wonder…
WHO THE FUCK came up with the shitty kick-stand that only works if you have it next to you while working
To make it 100% I would make the following changes:

For the inevitable person at Beurer who will see this and think…
The short of it is:

What if you could use kinesio-tape to help support healing psychological pain as well?

Hypothesis: I believe that if I tape above my physical (and emotional) heart, then I will direct my attention to my heart and its emotions more often…
Basically, have a reusable version of the classic “Hello, my name is” sticker as a button with my name on it and wear it while I’m out and about to show that I’m open to conversation by offering my real name.




My mother died after a long and courageous battle with cancer.
It felt like watching a deadly car-crash in slow-motion. Stretched out over ten agonising, depressing, and feeling-out-of-control years.
I last saw my mother around the 31st of December 2016 when I finished visiting my parents in Spain. She had been very weak and in pain and needed to lay down more then usual. The years of chemo and pain-medication had taken its toll on her body and her spirit. Her mind was still clear and she spoke about money, her brother and me.
I remember the last time I saw her, after we had said good-bye, passing by her window as she lay back into bed. I did not know this was the last time. I figured, after 10 years of ups and downs, this would be another one.
We spoke on the phone a few days later and she sounded very tired and out of it. I still didn’t make much of it, I wasn’t worried more than usual. We spoke one more time when I told her about a cold I had gotten. I remember she wished me well.
Two weeks into January, my father called me and told me they had to move her to the hospital. She had been too weak to stand up. And that he couldn’t take care of her anymore because he, too, was now too weak. He sounded heavy so I started to worry, too. I asked him if I could talk with her but he said she was too weak at the moment. I should try again in a few days.
I asked him if I should fly down to Spain but he refused: “There’s nothing you can do right now.” Then another week passed, he said she had a few good days but slid in and out of consciousness. That’s the first time we decided I was on “stand-by” to fly down.
In the late afternoon of the 5th of February, I got “the phone call”. The call I had dreaded for ten years since My mothers cancer diagnosis 2008. That ONE call in which my dad asked me to buy a ticket and come as soon as I can to see my mother, who was still alive, one last time.
I bought the ticket to fly out of Tegel early next morning to arrive at the hospital before 11am. I remember taking a Taxi to the airport at 6am. Then getting on the airplane feeling still and numb and yet completely functional. Two hours later, at around 9am, I landed in Palma de Mallorca. I switched on my phone and received my fathers WhatsApp message. My mother had passed away over night. And that I needed to come to say good-bye as soon as possible due to hospital regulations.
I took another Taxi and arrived at about 10 or 10:30. I first saw my father who explained to me what happened. She had not been conscious for most of the past days. Had been on pain-management. Her breathing had slowed down last night. He stayed with her for the whole night. Singing her her favourite song “San Francisco” by Scott McKenzie over and over as she passed away.
I had twenty minutes to say good-bye before the hospital had to move her to the morgue.
I entered the room. To the right her hospital bed appeared. I moved closer and saw her body. Eyes closed, her thin hands folded, still.
While it’s been great to have a single venture as my focus, I’m starting to feel too slow about it. Focus is good, but it needs to be balanced with the ability to find something worth 100% of focus. Here I think I’ve had opportunity costs by focusing on one thing too early.
I saw this recently on Instagram and it reminded me of how I had approached my ventures last year. Especially the “double your rate of failure” bit stuck with me. With one venture, I have a rate of 1. I could, and have in the past, parallelized 2 and 3 ventures.
After all… I learned last year, after publishing 10 books, that – if I do everything else right – only 1 out of 10 will actually work and be profitable.

So, here’s a break-down of what I’m thinking:
Starting with the smallest unit, the experiment with its three steps, then sequencing that in one venture, then parallelizing multiple (non-cross-pollinating) ventures to double and triple my rate of failure.
However, this is pretty much an idea where everything always takes the same amount of time. In reality it looks more like this:

So, I do have my experiment board that allowed me to keep track of my experiments before. And I’m not crazy about the Evernote style so far… Should I just re-use my tracking system? Why did I stop using it?
See How To Make Better and Faster Investment-Decisions With Lean Experiments for details ->

Many questions today. But I’ll find the answer!
For every “Greatest Hit”, I had to fail on average about ten times.
In most cases, “missing” means that I didn’t make, what was the right call in hindsight, sooner.
Without further ado, here are slices of my personal humble pie. A selection of my “Biggest Misses” and their biggest lessons:
DreamersWin GmbH (2001)
Next time: If it sounds too good to be true, it probably is. Run!
TigTags and PercentMobile Inc. (2009)
Remember: No matter how well it’s built: If they’re not buying, your offer stinks.
Venture Metrix Inc. (2014)
Make sure that you don’t scale it before you’ve nailed it.
Product-First Marketing (2018)
Once more: No matter how well it’s built: If they’re not buying, your offer stinks.
There are also My “Best Shots” (in case you’re curious.)
As many of you know by now, I strongly believe in the effectiveness of using the scientific method to validate business ideas.
Earlier this year, I began running a bunch of experiments for my ideas and ventures. But, quickly, my experiments became too much to handle on a white-board with stickies. I switched to a Google Sheet but that didn’t have the usability and speed that I loved from using stickies.
So I decided to build a simple experiment tracking system for myself.

I would say it’s pretty basic and covers exactly what I need to keep track of:
Experiments, which comprise of:
Cost tracking
Note that I built the tracking software only once I had run out of space on my physical custom-made Lean Experiment Board

I templated the main three experiment types that I usually run:



Once I have nailed those three, I start running custom experiments for Acquisition, Activation, Retention and Revenue (but not earlier.)
If you look at my dashboard below, I’ve run 43 Experiments for 14 Ventures. As you’ll see, based on my learnings I aborted more than 2/3rds of my Ventures and continued on only a few promising ones.
I think that’s been a huge value for me. To know and to be sure when to persevere and when not.
So far, I’m quiet happy with it. It’s rudimentary but works.
And since the Venture Board was, in itself, an experiment for myself (My theory was that I’ll consistently use it to track my progress and costs) –
I’m persevering.

Let me know if you would like early-bird access to the board. Simply comment below.
Thanks for reading!





Cap color: “Stone” gray Cap style: 39Thirty Logo color: Petrol-green
whole-earth-font-embedded.eps whole-earth-text-as-vector.eps
Bei Raycast gesehen als neues “Workflow” Feature:
Man hat in einer Applikation gerade Text selektiert


Development of the DeepDive
While on vacation at my Dad’s, I bought a tiny note-pad and pen and started carrying it around with me everywhere. I use it to Read to Feed and How to hunt for ideas like an artist.
This tiny notebook has since become a gamechanger for me to quicky capture notes and drawings that I would have otherwise “lost”. I still find that my phone is one of the worst ways to capture ideas quickly and with focus.
Because it’s so small, I’ve taken my notebook on practically every walk I’ve gone on. While I’m walking, when an idea comes up or I read or see something cool, I whip it out, take the pen and write.
The time from “oh, let me note this” to pen-on-paper is less than 3 seconds with this setup. To me, this speed and the flexibility that paper offers (write, draw, illustrate, etc.) is ideal.








Funnily enough, this one started out as a “backup” notepad because I had forgotten my then-current one at home. Because I was going to be out and about the whole day, I went to Muji and bought this notebook on purpose (and the Muji Fountain Pen by “accident”).
I liked the format but – obviously – the fountainpen wasn’t going to be my “out-and-about” combination (still very nice to write with though). So, when I got home, I looked around and picked up the slim “Birdie” ballpoint pen I had bought at Luiban a few weeks earlier. It’s a very nice-writing and nice-looking pen and as I held both in my hand, I saw that the pen might, maybe, could (could it?!) fit in the spine of the notebook.
And fit it did.
As you can see above, it didn’t only fit. The pen also has the right fit length-wise! (I don’t mind the top cap sticking out a bit.)
So, yeah, that wound up being a great combo that I hadn’t even thought of.
Without thinking much about it…, I’m already with my second notepad with this setup
To build on what’s there already and make it better.
Idea is to buy regular clothes, like from Uniqlo, but customize them uniquely and with a high degree of sophistication (think embedded gold rather than silk-screened) so they become something better and new. Add surprising details.
What would be better?
What’s missing?




See Whole Earth 39Thirty Cap (S21 Mini-Project) for details

See "Micro-Note-Taking Set" (S21 Mini-Project) for details


The same idea could also work with off-the-mill, cheap and boring porcellain which would be re-glazed (is that even possible) with better, more interesting colors and textures.

Have a DIY station there, after the cashier with


JamGPT.app
…GPT → (Wenn OpenAI damit okay ist.)

“…GPT”
Lumyn
Let’s make it PERSONAL because it’s coming from US
Alexis
Tobi
Buttercup 🌼
Shuck 🦪
Butter 🧈
FUCK 💥
PUNK
Brainstorm 5 playful, whimsical single-word names for an app that’s currently called “Smoothie”. Do not include “Smoothie” in the suggestion.
Brainstorm 5 playful, whimsical single-word names for an app that’s currently called “Smoothie”. Do not include “Smoothie” in the suggestion. Riff on “Slurpee” if you’d like!
Brainstorm 5 playful, whimsical single-word names for an app that’s currently called “Pogo”. Do not include “Pogo” in the suggestion .
Brainstorm 10 playful, whimsical, fictional single-word names for an app that’s currently called “Smoothie”. Do not include “Smoothie” in the suggestion. Riff on “Slurpee”, “Pogo”, “Jiggy”, “Smoothie”
Brainstorm 10 playful, fictional single-word names for an app that’s currently called “Smoothie”. Do not include “Smoothie” in the suggestion. Riff on “Slurpee”, “Pogo”, “Jiggy”, “Smoothie”, “Floopy”. Do not include real words or names.
Brainstorm 10 playful, fictional single-word names for an app that’s currently called “Smoothie”. Do not include “Smoothie” in the suggestion. Riff on “Pogo”, “Smoothie”, “Floopy”. Do not include real words or names.
3# Brainstorm a list of 5 fictional, made-up, cute, single-word names for a small companion. Riff off of “Flopsy”, “Pogo”, “Floppy”, “Bitzi” (Formatted as a markdown document)
I agree it seems weird if it’s the first contact.
To me it makes a bit of sense because I had built a GPT3-based chat two weeks ago, based on their completion documentation^1
The “format” for this prompt already has everything System prompt, user and assistant.
The following is a conversation with an AI assistant. The assistant is helpful, creative, clever, and very friendly.
Human: Hello, who are you?
AI: I am an AI created by OpenAI. How can I help you today?
Human: Please translate "Hello" to German
AI:
To retain context, I kept the system prompt on top and added the prior prompt/completions to the history stack in-between.
End-user-experience-wise, this worked really, really well. Practically as good as the ChatGPT UI.
As a developer though, it was not fun for obvious string-fiddling-splicing-and-dicing reasons. But again, this worked well and I would have kept it as-is for as long as possible.
So it wasn’t a giant leap when I saw the official format.
<|im_start|>system
You are ChatGPT, a large language model trained by OpenAI. Answer as concisely as possible.
Knowledge cutoff: 2021-09-01
Current date: 2023-03-01<|im_end|>
<|im_start|>user
Hello, who are you?<|im_end|>
<|im_start|>assistant
I am an AI created by OpenAI. How can I help you today?<|im_end|>
<|im_start|>user
Please translate "Hello" to German<|im_end|>
I’ve used that, too, and handling of prior conversations became a bit more manageable.
^1: https://platform.openai.com/docs/guides/completion/conversation
Your YOUness is what makes you a one-in-a-billion event in all of time and space.
Ever.
Your YOUness is endlessly valuable to the world because you exist only once.
It’s all the books you’ve read, people you’ve met, places you have visited, places you’ve lived in, songs you heard, tools you’ve used, events that you took part in, the pain you’ve suffered, the joys you have discovered.
All of that has made you who you are and what you care about.
It has shaped how you look at the world, what you deeply value, what surprising connections you make, what incredible ideas you’ll have.
Your YOUness is is your unique point of view into the world.
It is what you can truly, honestly, effortlessly share with the world.
When I started to think about this, I was unsure what that was for me. I figured, I’d like to explore my amazing MEness.
So, with a few small exercises, I’ve started to collect what makes me uniquely, awesomely, distinctly me.
[^1]
[^1]: Picture source: https://en.wikipedia.org/wiki/Universe