All Things Venture

Hey friends - so I want to experiment with a new form factor for this newsletter. I don’t have as much time anymore to write the longer form essays, but I do want to keep sharing my thoughts with the world. Most of this is for selfish reasons. I need to find a better way to reflect, think about, and synthesize everything I learn in a given week and I also want to be more consistent in doing so. As I’ve thought about the greats in any field, both historical and modern, consistency is a recurring theme. Just the sheer ability to do a singular thing, any thing really, day in day out, has a compounding effect. I’m aiming to start with weekly reflections, shared on Sundays and we’ll see where this goes. The reflections will largely focus on tech & VC, and for the time being mainly focused on AI applications and fintech. Let’s dive right in.

Klarna goes Public - Klarna went public on Thursday this week after a 20+ year journey to defy the odds. Klarna raised around ~$222 million in primary, and existing investors sold about $1.2 billion of exisiting shares. I think the Klarna story is interesting for a few reasons:

1. Fintech is Not Winner Take All: This should be obvious by now, but every where you look it’s clear that fintech businesses, especially those that are involved in money movement (i.e payments, lending, banking, insurance) are not winner take all markets. Where there’s a Klarna, there’s also Affirm. Where there’s Stripe, there’s also Adyen. Where there’s Ramp, there’s also Brex, not to mention even an Airwallex. Where there’s a Rain, there’s a Reap. Where there’s a Bitcoin, there’s an Ethereum or Solana. Where there’s an Itau, there’s a Nubank. Deel vs Remote vs Rippling, the list goes on. In fact, in fintech, it actually may pay to be second and draft off the market education of the leaders, even if they operate in different geographies (case in point Klarna, Adyen, Brex, etc.)

2. Regional Dominance to Global Winner: While certainly the exception to the rule, Klarna is an example of regional dominance being parlayed into a global outcome. The dominant venture narrative is unequivicoally that the US is the best place in the world to build a tech company. At $100bn in GMV, $2.8bn in revenue, with nearly 100m consumers Klarna is clearly a narrative violation. Moreover in Sweden, where Klarna began, 82% of the adult population were active Klarna consumers in 2024. In the US, that number is 10%. I call these figures out mainly to make the argument that human behavior is universal amongst peer groups. Consumers like choice, they like flexibility, and they like credit on fair terms. This is the basis of Klarna, but wired for the digital age.

3. So What? The so what for this busines is tricky and I’m going to think about it a little bit more, but my initial reaction is two fold. 1) In the context of investing in fintech focused in the US, counter intuitively the Klarna story shows that it may be more important to look externally to determine the next leg of innovation 2) If Klarna is an example of a consumer payments/lending business wired for the digital age, I can’t help but think there’s going to be a similar business wired for the agentic age. If we are living in a world where both human to agent, and agent to agent interactions only increase. How will money move in those scenarios? Will the default checkout infrastructure (i.e Klarna, Stripe, Adyen, Affirm, etc.) persist? Or will something new emerge? I’m not sure to be honest but this will definitely be an area to watch in the coming months/years

Brain Co. Announced - There’s a growing trend within venture behind investing in services businesses. The AI rollup market is emblematic of this, as is the excitement within the personal injury space with businesses like EvenUp, Supio, Tavryn, and Finch. Brain Co. is another business prosecuting the AI services theme, but at the highest level. Brain Co. was co-founded by Elad Gil & Jared Kushner, and includes a who’s who of silicon valley investors: Brian Armstrong, Patrick Collison, Reid Hoffman, Nikesh Arora. The stated goal of Brain Co is “to develop AI applications for both corporations and governments, offering a central platform through which institutions can interact with AI systems. Unlike many AI startups pursuing niche markets, Brain Co. has no plans to specialize in a single industry, instead positioning itself as a bridge between emerging AI capabilities and large-scale institutional needs.” I read this statement and think a few things:

• This is just a competitor to Palantir

• Even if it is just a competitor to Palantir, that’s great because Palantir is growing +50% YoY at a ~$4bn run rate and is viewed as one of the only games in town

• Kushner’s global reach and political connections are undoubtedly helpful

• Goverments across the world, similar to our own, have hundreds of millions if not billions of dollars earmarked to adopt AI

• From a competitive standpoint, intermediation makes sense becuase governments won’t want to have single points of failure with any model provider. If your national security apparatus, or healthcare system, or tax agency all run on OpenAI - you have massive counterparty risk and a single point of failure

• Application development is a function of technical talent, and while we’re increasingly in a talent diffuse envrionement - there is still significant concentration in Silicon Valley. Do the governments of South Africa, the Netherlands, or Japan have the talent to create, maintain, and upgrade their own custom application across OpenAI, Anthropic, DeepMind, etc.? The answer is no

• So Brain Co, has a pocket of opportunity to aggregate the necessary talent, aggregate government contracts on a global basis, they have the top down connections to kick start their distribution, they’re likely 1-2 degrees removed from even the most far flung institutions, and they can pitch themselves as a model agnostic service provider, a Switzerland of sorts. - even if they are a Palantir 2.0. Seems like a good trade to me

What Happened to Jasper AI? One of the traits that is emerging within the AI era of applications, is that outside of the businesses selling services to the foundation labs, business growth seems to emerge bottoms up. In 2021, the poster boy of AI applications was Jasper AI. Toward the of 2022, they announced a massive $125m Series A round led by Insight Partners. Details were scant, but they shared that in 18 months they had scaled to 70K paying customers. Assuming that pricing then was roughly what it is now ($60 per seat per month), this means they were at $50m ARR. Since then, they’d had a tough go of it. ChatGPT was launched in November 2022, a month after they announced their fundraise, and in all likelihood nuked their growth. By mid-summer 2023 the original CEO had stepped down, and in an fundraising market where growth rounds occur within months of one another, Jasper has not publicly announced a significant downstream round. I think Jasper is instructive because it shows just how “flighty” revenue can be in the AI era, as well as the importance of going multi-product as quickly as possible. If you look at Jasper’s website today, it positions itself as an AI platform for SMBs and the Mid-Market. A single place where you can create custom AI applications, create marketing content, and search company documents. It’s still positioned for marketers, but it generally feels like a Glean meets ChatGPT meets Canva, which is probably because each of those companies in their own right is trying to be Glean meets ChatGPT meets Canva. I think at the end of the day, what we now know, 3 years into the AI application era is that product defensibility is much lower than it’s ever been. Motivated teams can and will copy the feature that you have, but there is a solve to this, and that solve is the business model. You can’t and never will be able to compete with free, and as long as you have human attention - you will be able to find a way to monetize. The question from there really just becomes, is there enough depth and value in the attention that you have? An AI application that captures the attention of high school students better have a lot of depth, because high school seniors have very little money. Conversely, an AI application that captures the attention of Fortune 500 CEOs? A fraction of the relative depth is needed because there are hundreds of brands willing to pay millions of dollars for that constituent’s attention. Fundamentally it makes me think, is the AI era going to follow the modern internet? Free applications subsidized by ad driven business models? Or will it follow the current paradigm of streaming? Subscription based walled gardens with inevitable price increases to fund the next iteration of R&D or profitability milestones? Again - something to follow and think about over the coming months and years.

Hope you guys enjoyed the post! Per usual, I’m open to feedback and would love to hear from you. Drop a note in the comments and don’t be afraid to share with others!