585: Going Deep on Nvidia Q2, Genie 3 Dreams, Amazon's Business Anomaly, Global Payments, AI Bubble?, ByteDance vs Meta, Google's Automated Bug Hunter, and James Hetfield

"The sheer logistical nightmare"

It takes as much energy to wish as it does to plan.

—Eleanor Roosevelt

😴💭🌐 Real-time world generation like DeepMind’s Genie 3 reminds me of dreaming, where a reality is constructed on the fly as you move through it.

There are few true constraints. Anything could happen next (and it sometimes does).

The starting point could be any weird thought or image, and then you roll with it.

One of my favorite emergent use cases for Genie 3 is “stepping into a painting”, which turns a piece of art into a navigable 3D scene. But there are so many use cases, and I’m sure that countless emergent capabilities are just waiting to be discovered.

I highly recommend watching the compilation of videos I made for a sample of what DeepMind researchers have shared. It’s incredible 👆 and I don’t use that word lightly, this will be a big milestone dot on the “Progress of AI” timeline in history books.

🥩🍖🍗🧊 We went to Costco, so I finally had a chance to try my new vacuum sealer.

I like it so far. It’s more space-efficient and reduces freezer burn by keeping air off the food, which should help preserve flavor and texture. We don’t plan to go that far, but people report vacuum-sealed meat lasting 2–3 years without issues.

The real test will be to see if we notice an improvement in flavor/texture/freshness as we cook over the coming months.

On the convenience front, I packed the ground meat pretty flat so that it thaws faster in the fridge (more surface area). I haven’t done it yet, but you can also pack things directly with spices or marinades as a future time-saver.

---

🏦 💰 Business & Investing 💳 💴

🩻🔍🔥 Going Deep on Nvidia Q2 & The AI Gold Rush 🤖🕵️‍♂️

I know people get blasé pretty quickly about everything these days, but I still find it truly incredible that Nvidia went from ˜$32bn in trailing-12-month (TTM) revenues in Q2 of 2023 to ˜$165bn TTM in Q2 of this year. And over the same period, EBITDA margins went from the 30s to the 60s 🤯

The sheer logistical nightmare of scaling a company that sells extremely complex physical products this fast deserves more attention and accolades. I want an in-depth documentary on Nvidia's internal and external supply chain over the past couple of years, and what mountains they must have had to move to make it happen.

Let’s have a quick look at Q2’s numbers before we dive all the way in: