Ryan Seawell

Software investing

[Adam @ RepVue @thaAdamLittle]

Targets up - attainment down

Targets down - attainment up

Targets up - attainment up :)

Targets down - attainment down :(

[Ryan Seawell @LewistownCap]

I'm not sure how $UBER loses internationally. (1) In most international markets, Uber now leads in delivery and mobility. (2) As a result, in delivery you can't beat them on ETAs/pricing (they have highest market share/density) and users also get mobility benefits on top through Uber One -- $DASH does not seem to be taking any share from Uber on the whole internationally b/c of this, and they are trying. (3) In mobility Uber will be partnering with BYD/Geely internationally at some point I assume. These are the lowest cost EVs in the world, cheaper than $TSLA. I'm not sure how Tesla competes effectively. (4) That leaves Waymo, but Uber has big advantages over Waymo in terms of existing user base, combining delivery/mobility, lower ETAs, same or lower fares. Uber's going to be launching AVs at same time as Waymo in London, Tokyo, etc.

63% of Uber revenue is international pf for Delivery Hero. What am I missing?

[Ryan Seawell @LewistownCap]

@TerminalMiles really good take, i agree

[Ryan Seawell @LewistownCap]

Things that could help re-rate $UBER: (1) London. Both Uber and Waymo are launching AVs in London at roughly the same time (Uber with Wayve and Apollo Go). Uber is already widely used in London for mobility/delivery. So it will be a heads-up battle with Uber theoretically having a huge initial advantage. If Uber continues to dominate share in London post AVs, that could be a strong signal that Uber is going to fare well once AVs are more widespread on the platform. Also Apollo Go is launching in London with its RT6 vehicle, which only costs ~$35k (cheaper than whatever Waymo is going to use -- I-Pace, Ojai or IONIQ), (2) same for Tokyo, (3) Nuro launch in the SF Bay Area in H2'26. Assuming this goes well (it's supposed to be Bay Area wide, unsupervised) that would be the first US, non-Waymo AV partner on the platform with a robust offering, (4) NVIDIA launching AVs in H1'27 in Bay Area and LA. If NVIDIA names BYD (or Geely) as one of its international AV partners for its Uber robotaxi program, that would also be interesting. That would give Uber a price advantage (on top of everything else) internationally over Tesla and perhaps Waymo as well. Both BYD and Geely are on NVIDIA Hyperion, making this a distinct possibility, (5) maybe Uber shows a ton of traction with hotel bookings, adds a "flights" product to the app, shows significant broader momentum re Uber One -- cements itself as the all-things travel app...Others?

[Nicolas Bustamante @nicbstme]

My increasingly strong opinion is that compute is the biggest moat in AI.

Scientific breakthroughs don’t stay secret for long. Papers get published, researchers move, ideas leak, open source catches up. Give it enough time and almost everyone has access to roughly the same algorithms.

What doesn’t diffuse nearly as fast is the ability to run 10x bigger training runs, collect 10x more RL experience, serve billions of inference tokens, and iterate faster because you can afford thousands of experiments in parallel.

Scaling compute is brutally hard. It’s GPUs, power, substations, networking, cooling, land, permits, supply chains… and hundreds of billions of dollars. You can’t just decide to build that overnight.

The more I look at the industry, the more it feels like the frontier is less about having one genius idea and more about who can compound compute the fastest. Scaling laws reward whoever can keep scaling. That’s a very hard moat to replicate imo!

[Ryan Seawell @LewistownCap]

@thaAdamLittle Looks like AWS is up strongly?

[Ryan Seawell @LewistownCap]

@thaAdamLittle $VEEV $APPF

[Ryan Seawell @LewistownCap]

@andrew__reed If you "save as" and change the file type to PDF, it works every time + is instantaneous

[Zach Abrams @zcabrams]

It feels like gambling is going to become the opioid-like crisis of the 2020s

[Jaynit @jaynitx]

Kobe Bryant: "Failure doesn't exist, it's a figment of your imagination"

An interviewer asks: "Are you someone who loves to win or hates to lose?"

Kobe responds:

"I'm neither. I play to figure things out. I play to learn something. Because if you play with a fear of failure or you play with the will to win that supersedes fear, I think it's a weakness either way. If you play with fear of failing, you'll capitulate to that fear. If you play with the sense of 'I want to win, I want to win,' then you have the fear of what happens if you don't. But if you find common ground in the center, you're unfazed by either. That enables you to stay in the moment and not feel anything other than what's in front of you."

The interviewer asks: "How did you become someone who doesn't seem afraid of failing?"

Kobe responds:

"What does failure mean? It doesn't exist. It's a figment of your imagination."

He explains with an analogy:

"Let's use happy endings. Everybody wants a happy ending, right? Snow White finds her prince and lives happily ever after. Well, I call BS on that because two months later, they had an argument and he's sleeping on the couch. The point is: the story continues. So if you fail on Monday, the only way it's a failure is if you decide to not progress from that. If I fail today, I'm going to learn something from that failure and try again on Tuesday. That's why failure doesn't exist."

The interviewer asks: "If you finished your career without a championship, would you have looked at that as a failure?"

Kobe:

"No. I would look at it as being extremely disappointed, because I had a dream and goals I wanted to accomplish. If I didn't accomplish those goals, I'd have to ask myself why. Poor leadership? Failure to communicate with my teammates? Lack of preparation? Those would be reasons why I didn't win. So I'd have to analyze that. And as I evolved post-basketball into business, those same weaknesses would reveal themselves there too. If I don't learn from that, I'm going to struggle again."

He concludes:

"I can take those situations and learn from them and have them make me a better person later in life. But if I don't take that stuff and apply it someplace else, that's failing. The worst possible thing you can ever do is to stop. It's to not learn."

[Ryan Seawell @LewistownCap]

$ADSK is example of the 'seats' problem. AI is going to really improve Revit, for example. (1) You'll genAI designs natively in Revit (show me 10 examples of a 2-story farmhouse optimized for this lot). (2) You'll automate the process of 2D drawings that are spit out from your 3D architectural models (this is a problem today). (3) You can wire up an entire design for MEP (mechanical, electrical, plumbing) in one shot. (4) It will even understand your past portfolio ("use the same railing on this balcony that we used for 123 front st")

But what does this do to seats? I've read that ~1/3rd of Revit users are MEP specialists. What happens when the need for MEP specialists almost entirely goes away? I also read that automated 2D drawings could give architects ~30-50% of their time back. What happens then? And what happens to AutoCAD subscriptions (where all the 2D drawings happen, and still a big revenue line for ADSK)?

Can Autodesk just charge 100% more for Revit? Not sure they can without a customer revolt. Esp when there are a growing number of other options, and they've already been ramping pricing aggressively. But maybe they can, I don't know.

There seems like a pretty clear path to retaining dominance for Autodesk, which is just baking in this new AI functionality for a reasonable (or no initial) uplift in pricing. It's just a question if you are willing to weather a very painful few years of stalled growth, gross margin compression and multiple contraction. That's a big ask for mgmt, most couldn't/wouldn't navigate that well.

As an investor, perhaps then you catch them back on the upswing once they've re-cemented themselves with customers and growth can re-inflect (with maybe even greater lock-in and pricing power) + multiple expansion? But this doesn't seem like a "snap-back" situation, more like a protracted drawdown and trough.

[Brent Beshore @BrentBeshore]

21 days later, my opinion has completely changed with the introduction of Claude Cowork. I'll work on a longer update, but here are a couple off-the-cuff observations:

Last year we @PermanentEquity started dozens of agentic AI experiments, led by a talented technologist. All failed expectations, with only a few mild successes. Most experiments were 100+ hours of work over 3+ months. As I explained in the annual letter, we shut down the efforts in December.

Claude Cowork comes out Jan. 12th and I ignore it. I see the early adopters and charlatans doing their indiscriminate evangelism thing. I have high skepticism.

A couple friends I highly respect start talking about it. That's interesting. A few more people who aren't early adopters historically start chirping. Now I'm more interested.

We start playing around with Cowork on Monday. By Wednesday two of our top projects from last year were done. What failed with 100+ hours over 3 months led by a tech professional took a couple no-code private equity scrubs 20 minutes to complete flawlessly.

Since then we've started running dozens more experiments to great success. Not always perfect, but always good and quickly getting better.

The future is here. The implications are real.

[Kevin Gee @kevg1412]

New Fed Chair Nominee Kevin Warsh on how long it will take to see productivity gains in AI

[Ryan Seawell @LewistownCap]

@buccocapital Look, the heart wants what the heart wants. I’m sorry!

[Ryan Seawell @LewistownCap]

This seems like a very real trend $VEEV

[Bucket Shop Capital @bucketshopcap]

I bought 3 software stocks since the beginning of the year & I already regret it. IDK how you guys cover this stuff professionally, but I admire your fortitude I guess.

[Paul Graham @paulg]

I met today with the founder of Starcloud and I realized this is going to be one of the biggest engineering projects of our era. When you look at the tradeoffs, it seems inevitable that all the GPUs are going to live in space.

[Ryan Seawell @LewistownCap]

Operating Leverage as a Source of Market Inefficiency https://t.co/Qkdpd3CclO

[Ryan Seawell @LewistownCap]

@thaAdamLittle 1A, 2B, 3D, 4C