Moontower #270

Friends,

Happy Mother’s Day all. I’ll keep it short today with a few quick hits of useful things, stuff that picked me ears up, or otherwise found enjoyable.

✍🏽AI is not magic. It’s just math (12 min read)
Khe Hy

“A non-technical breakdown of how ChatGPT actually works”

This is a cool post even a middle-schooler can follow. Khe geeked out on Andrej Karpathy’s video How LLMs Work. The video is 3.5 hours so Khe distilled major points into a story. Readers might remember a 1-hr Karpathy vid I shared over a year ago called Intro to LLMs. He’s amazing at explaining the tech.

3D Printer Fun

It took about 90 min of assembly, downloading the app, getting the Bambu A1 mini 3D printer online, and loading the filament — but I did nothing — the 11 and 9 year olds managed it all on their own. That alone, made me a fan of this printer (ty Dave Nadig for the rec!).

They got the printer set up by Wednesday.

By Thursday, Zak sold his classmate this fish for $1:

next to a Milano cookie for scale

By Thursday night he rolled his eyes at me for inserting arithmetic into everything. By now he doesn’t complain, he just works it out to shut me up.

Me: How much material does it take to make a fish?

Kid: 10 grams

Me: The filament cost about $20, how much filament do we have?

Kid: 1 kilo

Me: So how much does it cost to make the fish?

Kid: $.20…I make 80 cents.

Me: What about the electricity?

Kid: Oh come on dad, I’m not paying for electricity.

Me: Ok fine, but you need to know the cost at least. [Quick Google search on how many watts the Bambu mini takes.]

It takes about 60. That means it uses 60 watts per hour.

Kid: How much does electricity cost?

Me: Here it costs about $.40 per kilowatt/hr. This is much higher than most places by the way. It’s more like $.11 where [uncle] lives. How long does it take to print the fish?

Kid: Half an hour.

Me: So what’s the electricity bill for the fish?

Kid: $2.40

Me: Bruh

Kid: [Laughs] Uhh, 2.4 cents

Me: Bruh, half an hour

Kid: [Laughs] 1.2 cents

Me: Cool. I cover the electricity, you use the profits to buy filament.

[We had 2 tangent discussions. He brought up the idea of higher-margin items. He has a friend selling a 30-gram fidget for $3 and he noted that it was the “same profit per gram”.

The other discussion was about heat and watts. We talked about one of the best YouTube videos we ever watched together about the discovery of the blue led. We learned that most of the energy, about 85%, in an incandescent bulb is turned to heat. Very inefficient. While LEDs, light-emitting-diodes, are way more efficient but also that they are an astounding and relatively recent innovation because it took so long to crack the mystery of the blue led. You’ll remember from the 80s that we had red and green LEDs in all kinds of electronics but we didn’t get white LED light bulbs until about 30 years ago. We needed blue. The blue led saga deserves a movie.]

Anyway, I’m sharing this because parents always tell me they like this stuff. I apologize to your kids in advance:

🚸Games and Word Problems For Kids

[Someone I’m close with who escaped the world of drug-dealing joked that my kid selling stuff in school (he’s not supposed to do that) is a gateway. That we bought filament by the kilo and Zak sold the fish an implied price of 5 kilos is a bit too on the nose.]

Full-body MRIs getting cheaper

I mentioned last week that Yinh and I are getting our full-body MRI scans soon. When Function acquired Ezra they slashed the price of scans for Function members. Since we had our scans booked before the acquisition they actually refunded us the difference. The price was lower by $499 which is the annual cost to be a Function member. Nice to see synergy on day 1!

This is the link to the scans.

[FD: Long Function; it was both a cash and stock acquisition of Ezra. I pay full price for the services though.]


Money Angle

🎙️Recap of The Sloan Sports Analytics Conference (Bet The Process podcast)

Jeff Ma and Rufus Wainright attend the conference every year. They talk about what panels and meetings stood out. It’s a great podcast in general because you pick up insights in passing just as they banter. The grout of the conversation always oozes gambling wisdom. When they echo the trope “if your model shows that you’re like a 10% edge, you probably should like re-evaluate your model” it reminded me of the pick-off story Kevin Muir told…hey mate, check your price.

I was doing a consult call recently with an option group and we were talking about those moments when you see an option structure trade at a strange price and how a surprising proportion of the time you come away thinking…”Oh, I see why they’d do that. I like their side.” It’s usually someone buying [selling] something that everyone would agree looks optically high [low] but the weird print snaps your mind out of autopilot to think about the scenario harder — only to come to the same conclusion as the aggressor — the underlying distribution is doesn’t conform to what conventional surfaces presume. [A good example was the emergence of the “teepee” vol surface ahead of earnings my friends and I referred to it. Voladynamics calls it the “W”].

Money Angle for Masochists

A new vid for the options audience.

Demonstrating the new Volatility Visualizer tool in the context of my IBIT May 60c position.

It covers what I see in the vol surface now to make me want IBIT vol using:

  • The DASHBOARD view to get a snapshot of overall vol levels and term structure.
  • The REAL Tool to compare implied volatility (IV) against realized volatility (RV).

We also discuss the misleading nature of realized vol after high-volatility periods, and how that impacts volatility risk premium (VRP) readings. I show how VOL CONES can help correct these distortions manually — and preview a coming feature that will do it automatically.

⏱️ Chapters
00:00 – Introduction
02:33 – Exploring the Moontower AI App
05:57 – Understanding Volatility and Z-Scores
09:45 – Using the Volatility Visualizer Tool
12:08 – Current Position and Delta Management
16:03 – Analyzing IBIT’s Volatility Trends
20:41 – Comparing Implied and Realized Volatility
25:36 – Evaluating the Market’s Volatility Expectations
30:34 – Final Thoughts on IBIT Options Trading

From My Actual Life

My mother-in-law who lives with us is 1 of 12 (all from 1 mother!). Most of the siblings and their families live in the Bay Area so we are expecting over 50 people today for festivities. Including my mother-in-law’s mother.

For my mom, love you, I’ll call you in a bit. And happy Mother’s Day AND happy birthday to my sis. Good luck in pickleball today 😛

Stay Groovy

☮️


Moontower Weekly Recap

Posts:

YouTube:

🎥demonstrating the new Volatility Visualizer tool in moontower.ai

Leave a Reply