436: Napoleon & Young Leaders, Microsoft + Activision, Jeremy Giffon, OpenAI & Meta, GPU Smuggling, Google+, and The Bear

"How long until we have an action film where the bad guys are selling black-market GPUs to China?"

Right and wrong are not the product of census.

—Terry Goodkind

If Twitter was a spaceship… ¯\_(ツ)_/¯ 💚 🥃

🪡🧵 I guess I gotta do this: Here’s my account on Threads (@Liberty_RPF).

When Mastodon had its moment, a lot of TechTwitter moved there, but Fintwit didn’t budge. When Bluesky had its turn, a lot of cool people moved there, but Fintwit didn’t budge.

But the Threads Tsunami appears to be big enough that a large chunk of Fintwit is now testing the waters and dual-posting (including me)…

It’ll be really interesting to see the final equilibrium because the current situation seems unstable.

🗓️👴🏻🤷‍♂️⏳ It’s strange how there are some people who are about my age (I’m 41), but I don’t feel the same age as them at all. Technically they’re peers, but somehow they feel older and I could almost picture them with a fanny pack and a wood-panelled station wagon.

Meanwhile, in my head, I still feel about 28 (it has been creeping up, until a few years ago, I still felt about 22-24) and like I’m just getting started at life — an impostor any time I have to do adult things or have responsibilities or am considered senior to younger people or whatever.

(Conversely, there are people who are older than me, yet they have so much energy and vitality that they make me feel older than I am…)

I don’t know why I feel this way and if everyone feels the same (even those that don’t look it from the outside) or if it has to do with my strange relationship with time.

Some people have a very clear chronology in their minds. Time is like a straight line and memories and events are neatly positioned on it.

In my head, it’s more like a ball of yarn. Everything is tangled together and I have a hard time knowing when things happened or to ‘feel’ how long ago they were. A lot of it is kind of superimposed and bleeds together.

That’s why I take a lot of notes on anything that I need to be sure about, or I’ll look things up on my calendar or find an email or message with a date on it to be certain.

📏🕰️ Speaking of subjective age and the feeling of time, here’s one more way I’m thinking about it:

Subjective life is logarithmic.

So the passage of 5 years when you are 10yo or 40yo or 80yo is very different subjectively.

It’s a lifetime for the kid (especially since they don’t remember their first few years, so subjectively they’re even younger than their calendar age) and it seems much faster for the 40yo, and I imagine even faster for the 80yo.

If I told my 9yo that within just 10 years he’ll be able to drive and may have his own apartment and such, to me it wouldn’t sound that far in the future, but to him, a decade sound like forever, and may as well not be real (or barely real).

🗣️🗣️ Speaking of time flying (there’s an accidental theme to the intro today, this wasn’t planned!), I had the pleasure of catching up with friend-of-the-show Marc Rubinstein (🏦 💳) yesterday and it was great, as always! It had been way too long since we last spoke.

His Substack is very successful and I’ve linked to it many times, so you probably already know it — but if you don’t, please check it out. I’ve learned a lot from him about the financial sector, banking (and how it goes from crisis to crisis!), payments infrastructure, hedge funds, etc.

💚 🥃 🐇 What has value in your life?

Is it the stuff? Is it the people? Is it the ideas? Is it understanding how things work? Is it discovering new pieces of art that make you feel something? Is it playing pickup basketball with your buddies?

Hey, I’m just asking, I can’t tell you!

Everyone has to figure this out for themselves, it’s crucial because it has an influence on so many other decisions that will then shape your life. You end up in different places based on these forks in the road.

But if you value what I’m sharing here, I hope you’ll become a paid supporter so that you can get even more of it and help me keep this steamboat in the water!

Much appreciated!

---

🏦 💰 Liberty Capital 💳 💴

Are we so used to old leaders that we have trouble taking young ones seriously? 🎥 🎬 👴🏻 🇫🇷

There’s a new trailer for Ridley Scott’s ‘Napoleon’, and as Rian C. Whitton points out, they’ve aged the French general quite a bit:

This trailer shows the spritely 48-year-old Joaquin Phoenix leading the French at Toulon and being crowned emperor.

Napoleon was aged 24 at the siege of Toulon, and 35 at his coronation.

He was only 46 at Waterloo.

Is this because audiences couldn’t suspend disbelief if Napoleon was played by an actor of the actual age of Napoleon during these events?

Are we so used to old men as leaders that it’s harder to believe a 25-year-old general on-screen than lasers from Tie Fighters making sound in the vacuum of space?

It seems like a pretty big societal blind spot. 👀

I remember when I was working at a startup, the founder was in his mid-30s (which seemed old to me since I was early 20s) and told me about how when he got in a room to make a deal, people on the other side almost never assumed he was the CEO. They often first talked to an employee who came along and was older.

Since I was in a visible position at the company at the time, he asked me to find a bio photo where I looked older (because when I was in my 20s, I looked around 15). We edited it, making it black and white and grainy to kind of hide that I was just a kid who couldn’t grow a beard.

It makes me wonder if Mark Zuckerberg would’ve been treated differently in the past 10 years if he looked 50 instead of 20 🤔 (not that I’m saying that his youthful appearance is the only factor at play, but I do think it’s *a* factor).

🎮 Microsoft + Activision: Wait, is this happening…? 🇺🇸🇬🇧👩🏻‍⚖️

A federal judge in San Francisco has denied the Federal Trade Commission’s motion for a preliminary injunction to stop Microsoft from completing its acquisition of video game publisher Activision Blizzard.

The deal isn’t completely in the clear, though. The FTC can now bring the decision to the U.S. Court of Appeals for the Ninth Circuit, and the two companies must find a way forward to resolve opposition from the Competition and Markets Authority in the United Kingdom.

On the UK side:

Microsoft and the UK’s Competition and Markets Authority (CMA) have both agreed to pause their legal battle over the proposed Activision Blizzard acquisition in order to further negotiate. [...]

“After today’s court decision in the U.S., our focus now turns back to the UK. While we ultimately disagree with the CMA’s concerns, we are considering how the transaction might be modified in order to address those concerns in a way that is acceptable to the CMA,” says Microsoft president Brad Smith

There are some potential ways for Microsoft to do the deal even if the UK doesn’t budge by finding some convoluted way to sell UK rights to some other entity or something of the sort (which would suck for them, but better than no deal globally because of one relatively small country).

I admit I kind of stopped paying attention to the details of this saga when it seemed like it wasn’t going to happen, but like one of those Army of Darkness undead, it just won’t die 🤷‍♀️

🕵️‍♂️ Interview: Jeremy Giffon 🔎💰

This was a really good interview by Patrick O’Shaughnessy (🍀):

• 🎧 Jeremy Giffon - Special Situations in Private Markets

There are many highlights, including the part where he says that we shouldn’t be afraid to ask about the basics and focus on understanding a business (what it does, how it makes money) because the most interesting businesses can be strange and unusual and may not make sense at first glance.

Conversely, if everything is super easy to understand and plug into a spreadsheet, that’s probably a sign that they are fairly commoditized.

I also like the discussion about patience, the ability to say no to things, pass on them and wait for better and more obvious opportunities:

"We've been debating this for three weeks, ergo, we should not buy it. It's not obvious. It doesn't smack in the face."

And this comes back to incentives again, which is you have to have the ability, incentive-wise, to be able to say no. If you're being forced to deploy capital or whatever, you trick yourself into doing these deals. But the good ones, you're just like, yes, I get it. Let's do it. This makes a lot of sense. I think part of that is just having the abundance mindset of being like if things are too hard, just let them go, it's fine. [...]

Another favorite question for managers is, how long have you gone without deploying money? The best I've ever heard is six years. […] Tiny did two years, which was incredibly hard. I can't fathom six years. I mean six years, you're doing nothing. You literally just do nothing. […]

Buffett and some other folks, I think, have done that, but it's so difficult. I find you don't really need to think about all this stuff. If your default frame is just not to do anything and you have to be compelled out of your seat. Because if you are, then you're going to have this visceral hurdle rate, which is, this worth my time? And if you're excited about it, then you do it and if you're not, you don't, and I find that you can really avoid doing bad deals that way.

Giffon seems like someone who has the ability to figure out what actually matters or not and which games are worth even playing in the first place.

That’s very important because even if you have a powerful vehicle, it won’t help you if you’re moving in the wrong direction.

🤖 Sarah Silverman and other authors are suing OpenAI and Meta for copyright infringement 🍿 🧑🏻‍💼⚖️👩🏻‍⚖️

The suits alleges, among other things, that OpenAI’s ChatGPT and Meta’s LLaMA were trained on illegally-acquired datasets containing their works, which they say were acquired from “shadow library” websites like Bibliotik, Library Genesis, Z-Library, and others, noting the books are “available in bulk via torrent systems.” […]

In the OpenAI suit, the trio offers exhibits showing that when prompted, ChatGPT will summarize their books, infringing on their copyrights. [...]

The claim says the chatbot never bothered to “reproduce any of the copyright management information Plaintiffs included with their published works.” [...]

As for the separate lawsuit against Meta, it alleges the authors’ books were accessible in datasets Meta used to train its LLaMA models [...]

In both claims, the authors say that they “did not consent to the use of their copyrighted books as training material” for the companies’ AI models. (Source)

Grab your popcorn, this is going to keep happening, and considering how non-deterministic and stochastic the justice system is, anything could be the outcome of this.

I’ve shared my thoughts on this a few times, so I won’t repeat them here, but if you’re curious ➡️ here’s a convenient link and another. ⬅️

🇨🇳🔐☁️ ‘U.S. Looks to Restrict China’s Access to Cloud Computing’ + How to Reduce GPU Smuggling 🚔

If you thought the U.S. chip ban was the end of it, think again. The goals remain the same, but the implementation is clearly being tweaked and adjusted as China makes moves to adapt to the original ban:

The new rule, if adopted, would likely require U.S. cloud-service providers such as Amazon.com and Microsoft to seek U.S. government permission before they provide cloud-computing services that use advanced artificial-intelligence chips to Chinese customers [...]

“If any Chinese company wanted access to Nvidia A100, they could do that from any cloud service provider. That’s totally legal,” said Emily Weinstein, a research fellow at Georgetown Center for Security and Emerging Technology. (Source)

To be honest, it’s kind of surprising that this wasn’t part of the original ban, but maybe they figured that even if Chinese entities used GPUs in the cloud, their access could be monitored and revoked at the flip of a switch, while physical GPUs being in China are bells that can’t be unrung. 🔔

Friend-of-the-show Jordan Schneider (🇨🇳🗣️) has a great piece with potential improvements to the system:

• How to Enforce The Chip Ban (With one weird trick! Well, a few weird tricks...)

Unlike other dual-use items, cutting-edge chips are manufactured in the millions. They can also usually fit in a shoebox, making smuggling possible even at the scale required to build modern supercomputers (thousands of chips).

And despite the controls’ focus on physical exports, most Chinese technology firms access chips virtually using services offered by cloud computing companies. These services are not monitored to prevent usage by blacklisted foreign entities under the United States’ current system of safeguards.

This part seems to be getting addresses! But how about the GPU smuggling:

How should the United States address these gaps while preserving its global dominance in supercomputing? First: The BIS needs to patch current controls by addressing the physical smuggling of chips. [...]

we propose a random chip sampling program. What would this look like? First, require that sellers of controlled chips register their sales with the BIS using unique IDs assigned during manufacturing. Then place requirements on sellers to notify the BIS about any secondhand sales and any cases where chips are destroyed or lost.

Finally, to stop chips from leaking into China, conduct randomized end-use checks to confirm the current registered owner of a chip actually has it in their possession. The BIS has historically done this with larger devices, such as lithography machines (a category of manufacturing equipment crucial for producing advanced chips). But given that millions of controlled chips are sold each year, BIS officials couldn’t hope to keep tabs on each one manually. Instead, by checking the location of a small but random share of all sold chips, the BIS would catch (or better yet, deter) any supercomputer-scale smuggling of chips to China. […]

At first glance, such a program might appear impossibly ambitious due to the sheer size of the semiconductor market. However, though there are countless chip companies and products available on the market, likely less than a dozen companies currently sell chips powerful enough to be subject to export controls.

How long until we have an action film where the bad guys are selling black-market GPUs to China? 🤔

📜 Remember Google+?

A blast from the past: "Google+ drew 90 million “users” six months after it launched in 2011 but eventually shut down because no one spent any time on it."

I remember when Google told every team that they had to support G+ with their products and *everything* had to be social in some way. Not that it was all a waste because many aspects of this remain to this day and have no doubt helped Google’s other products, but this was not a small deployment of resources.

To be very clear because I know the inference that could be made from posting the above: I’m not making a comment on Threads. I just saw a mention of G+ in an article and it reminded me of it, as I had not thought of it in years. I’m not saying that Threads will turn out like G+. In fact, if I was betting on the outcome, I’d bet on success (where success is not defined as killing Twitter, but just as having a large and engaged user base and being a great source of targeted signal for ads and text for machine learning).

👨🏻‍🔧 ‘How to do Great Work’: A new essay by Paul Graham 👩‍💻

I won’t even try to summarize this one. It’s one of Paul’s longer ones, but it packs a lot of wisdom.

---

🧪🔬 Liberty Labs 🧬 🔭

Reading at the beach (in the shade, of course) 💚 🥃

📘 Book rec: ‘Outlive’ by Dr. Peter Attia 🏋️‍♂️

I haven’t even finished reading it but I can already recommend you this book.

I’ve been listening to Peter’s podcast since it began, so I’m very familiar with all the themes in the book.

And yet!

It somehow works very differently in book form.

Rather than being scattered over dozens of hours of conversations that I may listen to over months or years, it’s all very logically organized and there’s relatively no fat, which makes each chapter punch very hard.

I’m already considering making some life changes because of this book that I wasn’t making based on hearing a lot of the same information on podcasts.

It’s also the first paper book that I read in a while, and it’s making me fall in love with paper books again. Having a physical artifact makes me want to read more.

In recent times, I had been reading largely ebooks because I want to have the highlights and notes, but now that I use Readwise and OCR my highlights, I feel like I’m getting the best of both worlds.

I’ll likely do a longer review when I’m done reading all of it, but even if the second half of the book was just blank pages, I’d still feel I had gotten a ton of value from it.

⌚️🚶‍♀️AI Model can identify Parkinson’s disease years before clinical diagnosis from smartwatch movement data

By identifying the risk of developing the disease sooner, maybe some neuroprotective treatments could be developed to avoid it getting to a point where symptoms are visible, or at least delaying that phase:

Machine learning models trained using accelerometry data achieved better test performance in distinguishing both clinically diagnosed Parkinson’s disease (n = 153) (area under precision recall curve (AUPRC) 0.14 ± 0.04) and prodromal Parkinson’s disease (n = 113) up to 7 years pre-diagnosis (AUPRC 0.07 ± 0.03) from the general population (n = 33,009) compared with all other modalities tested (genetics: AUPRC = 0.01 ± 0.00, P = 2.2 × 10⁻³; lifestyle: AUPRC = 0.03 ± 0.04, P = 2.5 × 10⁻³; blood biochemistry: AUPRC = 0.01 ± 0.00, P = 4.1 × 10⁻³; prodromal signs: AUPRC = 0.01 ± 0.00, P = 3.6 × 10⁻³). Accelerometry is a potentially important, low-cost screening tool for determining people at risk of developing Parkinson’s disease and identifying participants for clinical trials of neuroprotective treatments.

Will there be an option to enroll in this type of research on the Apple Watch?

More details here.

It’s not the first study of this type that I write about (see Edition #352 for a way to diagnose Parkinson’s and Multiple Sclerosis video of someone’s gait).

The unintuitive math that shows why Mercury is the closest neighbor to other planets in the solar system 🪐🔭

This is just a neat TIL! Science FTW!

---

🎨 🎭 Liberty Studio 👩‍🎨 🎥

Roman Emperor Caligula’s 2000-Year-Old Sapphire Ring

Look that that craftsmanship! Source.

🐻 🧑🏽‍🍳🔪🌯 ‘The Bear’ Season 2 (Post-Mortem Review) 📺

If you’re curious about this show, my season 1 impressions are here (with very light spoilers).

Below, I’ll share a few thoughts and highlights on each episode of S2. There are spoilers, though relatively light and it’s more a character-based show than plot-based. But be warned! 🚨

I’m very happy to report that something rare has happened: Season 2 of a show that started strong in the first season not only matched, but surpassed the original!

It feels like they reduced the anxiety and stress by about 20%, which is nice, but some episodes are still *intense*.

My fave scenes are like in episode 2 when two characters go off and do a side project. Carm and Sydney working on the menu, or Carm and Claire hanging out.

I love the tour of real restaurants in Chicago in episode 3.

I was getting some ‘Treme’ flashbacks (I love that show, watched it multiple times), and I think it’s very cool to pay homage to the people *really* doing what the actors are pretending to do.

Episode 4, which takes place mostly in Copenhagen, was one of my favorites. It shows a lot of great stuff about building skills for a craft and what it takes to be the best. Some great founder mentality in there!

And even the B-plot about Tina at culinary school worked for me. When she takes the huge gyuto Japanese kitchen knife and crushes her work at school and then kills it at the karaoke — very satisfying!

I like the character of Claire. She’s smart and no BS and seems likethe least neurotic person on the whole show, which is a nice change of pace.

The real standouts are episodes 6 (Fishes) and 7 (Forks). Both are very different, but brilliant in their own ways.

Fishes stars Jamie Lee Curtis, and she’s incredible. I wouldn’t be surprised if this was an Emmy performance for her — she certainly deserves it.

Bob Odenkirk is also very good. The whole edifice, the cast, the writing, the structure of the episode are on a knife’s edge the whole time but somehow they make it work.

It feels like a cornerstone episode for the show to understand who these people are and what shaped them.

But the real winner for me is the next one, Forks!

I can't believe they did an Alinea episode. And a Richie episode too! It was one of my very favorite episodes of TV from the past few years. So good that I think I will re-watch it as a standalone.

The layers of meaning in this go pretty deep, but even just on the surface, it was a pleasure to watch. The way it's filmed is also gorgeous. The contrasts between the gritty environments, the clinical kitchen, and the beautifully organic and strange dishes. Very well done.

These two episodes cement 'The Bear' as one of my favorite shows right now.

This is getting long, so I’ll drop the per-episode thoughts and just say that the final episodes are also very strong. I love the quieter moments like that discussion under the table. The walk-in fridge plot device was a bit much, but I’ll forgive it.

The ending makes me wonder where they’ll go for S3, but I’m in!