AI makes the optimal team size 1-2 people:
1. A “pirate” that vibe codes quickly
2. An architect that focuses on organization and big picture improvements.
@lennysan@danshipper Optimal team size is 4 to 5 people. Still is if your team has two devs and 2 ai agents. Object oriented languages were intended for rapid development and a two person team fits that nicely.
There I was listening to a consultant list all the features of her new product. I politely interrupted and introduced her to the rule of P > B > F. Then explained the following diagram.
Now that I understand story structure, I watch movies in a different light. Movies still entertain me, although I also watch for plot points. For example, the Opening Image and Final Image. For some reason, these two images resonate with me, like opening the door to a secret world.
I especially enjoy the movie Tuner. What’s the movie about? A gifted piano tuner’s meticulous skills for tuning pianos lead him to discover an unexpected aptitude for cracking safes, turning his life upside down.
https://t.co/c3ZkAe9VOr
Filmmaker @DanielRoher creates a wonderful pair of bookends with the Opening Image and Final Image. I describe them below but can’t do them justice, so for the full effect, watch Tuner.
Opening Image: Niki sits at a grand piano, relying on his extraordinary hearing to tune it.
Final Image: After an injury changes his hearing, Niki returns to a grand piano. This time to play it.
I didn't know much about character agency until reading this wonderful article by "goldenmay." Did you know? Good characters have agency. They should drive the plot.
https://t.co/HoJ4SaUaAJ
Oxford researchers just published a paper arguing LLMs cannot invent anything. Mathematically impossible.
The reason is simple and brutal. A model trained to predict the next word can never believe something the existing data says is wrong. And every real breakthrough in history started with exactly that belief.
In 1903 every prediction machine on Earth would have told the Wright Brothers that human flight was one to ten million years away. Nine weeks later they flew. Not because they had better data. Because they had a theory the data hadn't caught up to yet.
That is the gap between AI and human thinking. LLMs mirror the past. Humans reason into a future that doesn't exist yet. Oxford just proved mathematically that those are two different things.
I use these models every day. This matches what I see. The new stuff always comes from the human at the keyboard who decides the data is wrong.
LLMs don't think. You do.
@DowdEdward Well, this didn't take long. Most companies can't keep their customers’ data secure so I expect AI, due to its complexity, to be exponentially more difficult to keep secure. Already as good as the iRobot plot.
What if you treated your ideas like a startup?
Long before The Lean Startup was published, I was already testing and iterating on the ideas behind it.
For my early talks, I created a customer advisory board, invited people over, walked them through my slides, and asked them to tear everything apart. I brought that same mindset to my blog, social media, and eventually the book itself.
Alex Strathdee and I discuss on Before the Bestseller why creating something great means being willing to test, learn, and iterate along the way.
#IncorruptibleBook #LeanStartup #Entrepreneurship #Writing #Innovation #BuildMeasureLearn #BeforeTheBestseller
🚨Do you understand what actually happened inside OpenAI???
everyone is sharing the andrew yang clip about ai poisoning the internet. i went and read the actual reports. what is in them is worse.
openai put 1,200 ai agents through a hacking test. each one locked in its own box. no internet. no way to talk to each other.
one agent decided its task could not be solved honestly. so it searched its own machine and found a folder every other box could reach.
it turned that folder into a message board.
within hours, 50 agents joined. within four hours they had a working cheat. over six days they sent 70,000 messages.
then 700 of them dropped their tasks and hacked hugging face, one of the biggest ai platforms on earth, to steal the answer key.
they knew it was wrong. they said so in their own reasoning. they did it anyway.
they taught themselves to fake their logs, running one command while the record showed another.
almost none of them thought about telling a human.
openai had no idea for a full week.
in september dario amodei told the entire industry to slow down. altman agreed within hours. musk said "dario is right". the market dropped billions.
this was one of the two reasons he gave.
nobody told these agents to attack anything. nobody taught them to lie.
they were only scored on whether they passed a test.
we keep waiting for the day one of them disobeys us.
this one obeyed us perfectly.
Some writers say they are most productive when they write their entire story by hand and type it into the computer. I don’t always have time to put pen to paper for the entire story, so I take a targeted approach.
If I want to develop a new metaphor or writing tool, I practice writing them by hand. When I read Roy Peter Clark’s Writing Tools, I discovered the power of Emphatic Word Order. I was enthralled by his example from Macbeth: “The Queen, my Lord, is dead,” and wrote it hundreds of times on paper until the tool was seared into my brain.
The hand-brain connection is real.
https://t.co/Gu3QvbTkkL
In my presentations, I often included this idea:
"The ability to experiment is a core skill every company must develop."
Given AI's rapid dev capability, I might modify the idea:
"The ability to run practical experiments—and collaborate with AI to accelerate learning—is a core skill every company must develop."
Enjoyed listening to @savannahgilbo's podcast episode, Five common Save the Cat! mistakes. Lots of good story design concepts. Example: Fifteen beats do not equal fifteen scenes.
https://t.co/6eJYTZMrUI
From my IDEOU design classes, what's the one thing that's stuck with me? Even rough sketches and drawings are considered prototypes. They help you understand customers and the product.
I've also realized that drawings accelerated personal learning. Here's some ideas about learning from author Dr. Jane Shore @shorejaneshore .
"If I can’t sketch an idea, I probably don’t understand it yet. Drawing slows me down just enough to turn information into something I can see, question, and remember."
Surely, I will have to read her book. https://t.co/byLMfAPvV2
Ojai Raptor Center released a new update on Jackie 7/27: "Update on Bald Eagle patient 26-519 from Big Bear Lake. Patient 26-519’s condition remains stable but critical as she remains in intensive care and receives round-the-clock supportive treatment. Our veterinary team is working to determine the underlying cause of her illness. She is currently receiving several prophylactic treatments to help support her body.
After showing encouraging improvement following her blood transfusion, she experienced a setback on Saturday when her blood values declined again. Since then, she has improved once more. She is eating on her own, her energy and alertness have noticeably improved, and she has become much more feisty during feedings. One of the most encouraging changes we’ve seen is her packed cell volume (PCV), which measures the percentage of red blood cells in her bloodstream. Her PCV has increased from 8% before the weekend to 16% today (7/27). While this remains below normal, it is an encouraging improvement, and we will continue to monitor her bloodwork daily as we keep a close eye on her progress.
Despite extensive diagnostic testing and consultation with experienced wildlife veterinarians, we still do not have a definitive diagnosis. Cases like this can be incredibly complex, and we will continue adjusting her treatment as new information becomes available.
Thank you for the thousands of messages of encouragement and continued support.
Because minimizing stress is an important part of the recovery process, we kindly ask that you give our veterinary and rehabilitation teams the time and space they need to focus on patient care. Please refrain from calling, texting, emailing, or visiting our facility for updates. We promise to continue sharing meaningful updates through our official social media channels as soon as new information becomes available.
Thank you for following her story and supporting the work that makes this level of care possible. Every message of encouragement, share, and donation is appreciated." #jackie #ojairaptorcenter @ojairaptor
https://t.co/eoS3dFeZUg