Oren Ellenbogen

Scaling engineering org is about tweaking for flow, happiness and fulfillment. If your process is not improving one of the above, kill it.

Ilya Sutskever sat on a Toronto panel in 2015 and laid out the entire trajectory of AI in five minutes. The thesis was one slide. Generality. Deep neural networks don't care what problem you give them. If the architecture solves one hard pattern recognition problem, it will solve every hard pattern recognition problem. Eleven years later every breakthrough has come from that single claim. GPT-4, Claude, Gemini, AlphaFold, Sora, AlphaGo. All the same recipe. Backprop, gradient descent, a deep neural network with enough parameters, fed enough data, trained with enough compute. The architecture changed once when Vaswani's team published the transformer in 2017. The training algorithm has not changed at all. The bet that looks obvious in retrospect was hard to make in 2015. AlexNet had crushed ImageNet three years earlier and speech recognition was getting better. That was about the entire deep learning track record. GPT was three years away. AlphaFold was three years away. Stable Diffusion was seven. The dominant academic position was that perception had been solved by deep learning and that reasoning would require a fundamentally different approach. Ilya's claim collapsed perception and reasoning into the same problem. Different inputs, same machinery. The breakthroughs were already implicit in the recipe. The field just had to scale it. The part of the talk people don't quote is his point on optimization. Ilya says there is no mathematical theorem that proves backprop should work on a non-convex objective. He calls it an empirical fact. The whole field still rests on that one observation. We can't explain it eleven years later. The slide was the entire trajectory. He spent the next decade executing it.

The weird thing right now is the public markets don't have access to the growth side of software. Right now the trade is to sell SaaS and buy semis (the raw material of AI). What you don’t have yet in the public markets are the AI native software companies and therefore, you’re comparing the practical values of owning say a Salesforce vs the mythical value of owning a company that’s growing 10x (without having seen the actual financials). And everyone is always going to want the myth. These SaaS stocks aren’t going to trade in a sane fashion until the next generation of AI companies go public and investors can decide how to price a 10% revenue growth company with 30% cash flow vs 300% revenue growth company with negative 100% cash flow and SBC that will blow your mind. Until then, you’re walking hand in hand with your significant other looking over your shoulder. You know the meme.

אני עדיין מנסה לעכל את מה שקרה כאן ב-48 השעות האחרונות. קצת מספרים: מעל 5,100 כניסות לאתר מעל 2,200 חיפושים של ניסויים מעל 1,100 סיכומים בשפה פשוטה מעל 320 צפיות במדריכי שיחה עם הרופא מעל 210 בדיקות ההתאמה המהירות והנתון שהכי ריגש אותי: 62 פניות ישירות לחוקרים ולצוותי הגיוס של הניסויים! אם אפילו חולה אחד/ת יצליחו להיכנס לניסוי מאריך חיים מבין אלה, עשיתי את שלי. תודה שאתם עוזרים לי להפיץ את התקווה הזאת 🙏

someone built an OpenClaw agent that SELLS pool installations on autopilot. finds $500k–$1.2M homes without pools renders a pool in their backyard and mails a before/after postcard.

The so-called human rights champion has no problem reopening an embassy in a country that massacred tens of thousands of protestors, shut down the internet for weeks, and launched ballistic missile attacks at more than 10 countries.

Construction of Medieval Castles still blows my mind. For example, here’s Bamburgh Castle in England from the 11th century!

חברות משלמות מיליוני דולרים על פרסומת של 30 שניות במחצית של הסופר בול. ואז הדבר הבא קרה לנוטלה אתמול בארטמיס 2 באופן ספונטני:

I made a USB-Clawd who gets my attention when Claude Code finishes a response

The Alchemist by Paulo Coelho I have lost count of how many times I’ve read this book over the years. It is a rare gem of a book, and touches the very core of every heart earnestly seeking its own destiny on the journey of life. 13 powerful quotes from the book:

I'm looking to invest in cancer research and breakthrough treatment companies Cancer is humanity's worst enemy, and unfortunately, it's one of the few frontier problems i don't see getting solved in the next 2-3 years.  I'll be able to write checks up to a few million (usually up to a million).  I will work extremely hard to help the founders. I'll leverage my network as much as possible to overcome bureaucracies and help hunt down talent for every company I'm investing in. Things I'm particularly interested in:  1. Accelerating clinical trials: As AI gets smarter and allows novel hypotheses on new drug targets, clinical trials and bureaucracy will remain the biggest bottleneck to fight cacner. 2. Repurposed drugs: the current cost of Phase 3 clinical trials only allows patentable solutions to be economically viable. It resonates that are probably many other treatments and potentially impactful drugs that won't be included in the Standard of Care as they haven't gone through phase 3 clinical trials.  3. Personalized medicine: organoids, digital twin, faster cycle from biopsy -> identifying targets -> creating customised solutions, such as mRNA vaccines, custom Antibody-Drug Conjugates (ADCs), and novel ways of triggering one's immune system.  4. Improving measurement: Currently, the standard PET-CT method does not allow weekly measurement of disease progression. Identifying a superior method to allow weekly measurement of progression could be instrumental in cutting down the time it takes to find a drug the body respond to. Anyone who can connect me to relevant companies will be widely appreciated

Startup idea 💡 Dark mode books.

ההתלהבות מ״התרגיל המבריק״ של הקואליציה היא כמו לפרגן על התחכום לאנשים שמחתימים קשישים על הוראות קבע בטלפון כדי לגנוב להם את הפנסיה שהם משלמים איתה תרופות

In Japanese, “tsundoku” means collecting books and letting them pile up - not for neglect, but for the joy of knowing they're there, full of untold stories. 📍Kinokuniya Book Store, Tokyo

Company meeting in the next 5 years

היום TIME פרסמו כתבת שער על גווין שוטוול, נשיאת SpaceX ומנכ"לית התפעול שלה, תחת הכותרת "Rocket Woman". שוטוול, מהנדסת מכונות בהכשרתה, היא הדמות המרכזית בניהול היומיומי של החברה כבר יותר מ-20 שנה. היא אחראית על הפעילות התפעולית של אלפי שיגורים מוצלחים, פריסת רשת Starlink העולמית, פיתוח Starship והכנת התשתית להחזרת אסטרונאוטים לירח במסגרת תוכנית Artemis. בעולם שבו החזון מגיע לעיתים קרובות מגבר אחד שהוא חזות הכל, שוטוול היא זו שמביאה אותו לידי ביצוע בצורה שיטתית ומקצועית. היא מייצגת דור של נשים שמובילות תעשיות טכנולוגיות מורכבות מתוך עמדות בכירות, ומשלבות מצוינות הנדסית עם ניהול ברמה גבוהה. במיוחד בתקופה הזו, חשוב להביא סיפורים כאלה לקדמת הבמה – כדי שהבנות שלנו יראו מודל ברור של אישה חזקה, מקצועית ומנהיגה בתחום החלל והטכנולוגיה, וידעו שגם להן יש מקום מרכזי בעיצוב העתיד.

Jevons paradox is happening in real time. Companies, especially outside of tech, are realizing that they can now afford to take on software projects that they wouldn’t have been able to tackle before because now AI lets them do so. We’re going to start to use software for all new things in the economy because it’s incrementally cheaper to produce. Marketing teams at big companies will have engineers helping to automate workflows. Engineers in life sciences and healthcare will automate research. Small businesses will hire engineers for the first to build better digital experiences. And as long as AI agents still require a human who understands what to prompt, how to review when an agent goes off the rails, how it guide back, how to maintain the system that was built, how to fix the ongoing bugs, and more, we will still have humans managing these agents. This is why all the advice you get of not going into engineering is wrong. The world is going to increasingly be made up of software, and the people that understand it best will be in a strong economic position. This will happen in other roles as well where output goes up and demand increases.

טייק מרהיב ומרגש של ג'נסן הואנג מנכ"ל אנבידיה מהסופ"ש האחרון על המלחמה עם איראן, ישראל ואיך שהוא רואה את הסיטואציה. לא טריוואלי בשוטף, בטח ובטח לא בתקופה הזאת. כשמנכ"ל החברה הצומחת והגדולה בעולם לונג על ישראל בצורה כזאת מובהקת זה אומר הרבה.

They pulled off the impossible: made a real human video look 100% AI generated 🤣