David Joffe βŒ¨πŸ–ŒπŸŽ¨ (@d_joffe) header
David Joffe βŒ¨πŸ–ŒπŸŽ¨ (@d_joffe) profile picture

Software Engineer Cofounder @TshwaneDJe HLT Created TLex,tlTerm C++ GameDev VR GPU Art:Pro-revival of tradition of beauty&skill Edu dictionaries Reason AI Aging

Western Cape, South Africa
Joined June 2010
Conceptual mockup (not a product) Exploring ideas around an interface for GPU-accelerated molecular simulation, aimed at longevity & age-related disease research (e.g. Alzheimer’s, cancer) (In addition to my TLex/tlTerm work) Visuals are illustrative, NOT biologically accurate
6
1
157
127,453
Making dictionaries with TLex Lexicography Software (Dutch) ...
2
93
TLex (Dutch)
πŸ€– Made with AI
1
1
16
English
πŸ€– Made with AI
2
12
"I'm worried, my wife's last online says 41,964,012 milliseconds ago" Investigations: "Sir, you said you spent 18,540,238.2 milliseconds at the store, but CCTV shows you spent 9,310,104.3 milliseconds, can you explain the difference?"
1
45
Disputes: Airline: "Sir, you were 34,035 milliseconds too late for check-in" Client: "No, we were here 49,201 milliseconds before cutoff!" Or: "My AI agent requested ticket cancellation 49,034.2 milliseconds before cutoff! Your system took 98,025.1 milliseconds to respond"
1
28
The above may become slightly prescient in an increasingly AI/agent/robot-driven world. CVs/Resumes: "Software Engineer with over 946,080,000,000 milliseconds experience" Job requirements: At least 157,788,000,000 milliseconds Python programming
1
24
New trend idea, what if we made millisecond resolution the norm for everything? "I got 16,887,408 milliseconds of sleep 😭" "They say you should get at least 28,800,000 milliseconds of sleep a night!" "Honey, hurry! Flight check-in closes in just 956,539.1 milliseconds" πŸ™ƒ
I got 4 hr 41 min of sleep 😭
1
79
David Joffe βŒ¨πŸ–ŒπŸŽ¨ retweeted
I got 4 hr 41 min of sleep 😭
bf complaining that the You Can See Everything screening is going past his bedtime while bryan johnson sits two rows ahead of us. sir you are in the presence of the man who invented going to bed.
201
289
33
27,111
1,348,705
Why interruptions hit hard Deep focus is fragile. One β€œquick question” can shatter the whole mental model, and rebuilding it can take longer than the interruption itself. If it’s not urgent, and I'm focusing, please let me finish my focus block. Who else relates? 😊
πŸ€– Made with AI
1
41
David Joffe βŒ¨πŸ–ŒπŸŽ¨ retweeted
Por si te interesa, el cΓ³digo fuente original del Apolo 11, cuyo desarrollo liderΓ³ Margaret Hamilton junto a su equipo en el MIT, puede consultarse hoy en GitHub. EstΓ‘ en ensamblador y con los comentarios originales: github.com/chrislgarry/Apoll…
7
15
181
9,180
David Joffe βŒ¨πŸ–ŒπŸŽ¨ retweeted
Am I a madman or messiah? I’m not going to feign humility or pretend I’ve stumbled into this. For 30 years I’ve worked to create something of enduring value for the human race. My primary education has come from reading biographies. You can read a book on World War II and learn how the Germans did this, and the Americans did that. Autobiographies, by contrast, preserve the high resolution nuance that broad histories discard, capturing the raw psychology and lived friction of the era. Biographies speak to me because they represent the rare artifacts that survived the brutal compression of time. Every generation’s existence is compressed by around 90%. That means, 90% of people and things that seemed immortal during their time are erased from history in a single generation. It’s simply too much information for humans to carry forward. Three generations of compression means that 99.9% of human activity is forgotten, leaving only one in a thousand figures, ideas, or works to cross the threshold into cultural memory. Humans solve this information overload by encoding entire epochs into a single person, event, work, institution or idea. These distilled archetypes become required information for the species to tell its own story. Going deep into biographies in my 20s, I could see these patterns: an infinite depth of human experience continuously compressed into durable memetic code. I felt inspired by this idea and by the people I was learning about. I specifically admired those, who in their time and place, were able to identify the central defining issue of their time. They didn’t do what others were doing. They didn’t by default pursue the status games of their era. They sought out what I call Zeros. Schopenhauer drew this distinction: > Talent hits a target no one else can hit. (First principle thinking) > Genius hits a target no one else can see. (Zeroth principle thinking) I wondered: > what will be the Zero of my lifetime? > what is currently invisible yet will be the only thing history will see? > can I find it and build it? I’ve been pursuing this goal since I was 21 years old. Without having any real talent, struggling as a student, having a humble upbringing, and no world connections, I set a goal of becoming independently wealthy by age 30. Then, with money, I reasoned, I’d set my focus on becoming a Zero detective and a Zero builder. At age 34, I sold Braintree Venmo in 2013 for $800M, making $300M of that. It was now time to become a zero detective. The problem was that I was cooked as a human. The fourteen years of entrepreneurship, young children, existentially wrestling with my spirituality, and a challenging relationship left me chronically depressed, overweight, inflamed, recently divorced, and without meaning. First, I’d have to rebuild myself to stabilize. Health became my stabilization. I didn't know it then, but the thing I did to survive would turn out to be the thing I was looking for. I obsessively got to work searching for a Zero. I became a voracious reader and thought about it constantly. I hosted a dozen dinners across the U.S., posing the question to attendees, β€œImagine we’re in 2050 and life is extraordinary, what did we focus on in 2016 to make that happen?” I’d carefully write down everyone’s ideas and then after the dinner series, I put them down on a piece of paper, drew a box around them, and then told myself I couldn’t do anything inside the box. Still, I couldn’t see, feel or hear a Zero. I then did a thought experiment that did it for me. I imagined being present in the year 2500 and sat there listening to them converse. They were talking about homo sapiens who lived in the early 21st century and they expressed their gratitude saying β€œWe are grateful for homo sapiens of the 21st century who did __________ and __________.” What do they say? What would survive 500 years of compression? I thought about this question for years. Then one day the answer came to me, they say: β€œWe are grateful for homo sapiens of the 21st century who gave birth to superintelligence and figured out they’d no longer die.” Our century compressed into two artifacts. Superintelligence will be built whether or not I’m around. The other artifact was wide open. It would be a turning point in intelligent life. Transitioning from everyone inevitably dies (YOLO) to we are the first generation who won’t die. Changing everything about how we think about life and death, other people, the planet and our precious, fragile existence. Don’t Die is the Zero. A tectonic shift in human values from being willing to kill themselves and others for power, wealth and status, to valuing existence itself above all other rewards. Wanting to live is the most ordinary desire a human has, which makes it a strange thing to call invisible. Very few actually want to die, at least today. Yet almost nobody organizes their life around not dying because for all of history it was unachievable. It makes more rational sense to pursue immediate hedonic, status and power objectives and to lean on the elaborate philosophies the species has made to make peace with death. The era of Don’t Die will commence in 2030. New age reversal therapies will emerge and AI automation will run our lives, removing the biggest blockers of health: discipline and effort. Death acceptance will vanish as fast as body positivity did with GLP-1s. (The irony here is that if Don’t Die is a Zero, and it comes to pass, then the 90% compression that has been the rule may stop being the rule.) Kate and I are building for this. We do serious experimentation on health. We are performance artists. We are memetic representations. We are building a philosophy and ideology. We are building the infrastructure for the species to turn its attention and power to Don’t Die. If you’ve ever felt inspired, like me, by people of the past, who were willing to be hated, ridiculed, and dismissed in their pursuit of a Zero, the above should give you motivation. You will have to suffer through a prolonged period trying to find and build a Zero but the reward is being of service to the long arc of intelligent existence in this part of the galaxy.
158
54
18
942
84,013
I don't know if it's a function of training but LLMs sometimes feel like 'status quo machines', particularly when I like to try rethink so much of the status quo that is bad. It wants sometimes to "explain" bad systems to me when l am discussing to rethink bad systems
1
17
Thinking of Pretoria (Tshwane) power issues, I previously helped develop 3D virtual electrical substation simulator and line inspection software for training, inspection and maintenance, and now wonder if these could be helpful still for helping reduce these outages?
1
1
2
481
Fraud is everywhere and yet almost every app is basically not really designed correctly to help detect and handle this correctly. I have ideas why, and how to improve it...
1
2
62
David Joffe βŒ¨πŸ–ŒπŸŽ¨ retweeted
Making dictionaries with TLex Lexicography Software (Dutch) ...
1
2
93
Are we basically going from "9 out of 10 dentists agree" to "9 out of 10 AIs agree"? I built RunAI Studio, a multi-pane app allowing one to query multiple models at once. One prompt, multiple AIs, side-by-side.
1
3
73
Saw some online productivity advice that one should have a "not-to-do list". So I'm starting a "not-to-do list" and the first item I'm putting on it is: 1. Not to do: Create a not-to-do list Seem right?🀣
1
61
The second item on my Not-To-Do List: 2. Waste time making weird meta-recursive jokes about Not-To-Do Lists.
40