Big Shots, Large Swings, and Hail Marys
You would be hard pressed to find truly great companies in any tech cycle that began as agreeable software businesses. They each began as attempts to own a future that did not exist yet. Regardless of who exactly comes to mind, great companies were predicated on the belief that markets can be remade, supported by economic structures that allowed for niche software solutions to achieve outlier outcomes.
The current crop of founders don’t found like that anymore. Instead of ushering in the golden age of value creation, AI has placed a shroud over our ability to see beyond the present moment. AI has not made ambition impossible, but it has made smaller ambition defensible.
For the first time, it’s incredibly rational, profitable, and even admirable to take on capital to build a startup whose ceiling is visible from the beginning. The collapse of previous barriers to value creation has created a strong facsimile of progress.
The Age of Agreeable Companies
There are real benefits to appearing like you are building something great, especially if the thing you are building references Agents, Automation, Humanoid Robots, etc. In places like San Francisco or New York where job functions carry large social implications, this is especially the case. Capturing consistent mindshare can create life-changing outcomes for founders.
This new relationship between capital, class, and founder outcomes cuts down on the subset of ideas that people think are worth founding, allowing for agreeability to become a hedge for uncertainty.
If I offered you two years of work for 30% chance of a $5M outcome, or ten years of work for 2% chance of a $500M outcome, which are you picking right now? On expected value alone, the second option wins easily. But that's not how founders are choosing right now. For many, the first option feels really appealing. It's still unclear what the future of work and life will look like, and in times of uncertainty, ownership and conformity allow for market participants to attach possibly poor outcomes to a larger group consensus.
This creates an interesting paradox where people are looking to commoditize an industry that is built on big shots, large swings, and hail marys.
The Wrong Thing to Win
It used to be the case that the sheer burden of creating technology allowed specialization to be a defensible moat. By owning the whole technological vertical, corralling the strongest talent, or plainly solving difficult problems, many startups thrived by owning a singular wedge in a particular market.
Salesforce is the company that comes to mind. In 1999, they set out to build CRM software in a category Siebel already owned. Their edge wasn't the feature set or the ease of use — it was the ability to ship one piece of software, running on their own servers, that every customer could use over the internet. At the time, CRM was built and maintained per customer, installed in each company's own data center. The multi-tenant architecture Salesforce invented to escape that model required years of engineering work that no founder could do casually and no incumbent could absorb without dismantling themselves.
This is the shape that birthed the SaaS business. Narrow wedges defended by infrastructure that took years to build and that competitors couldn't easily reproduce. Cloud dissolved the infrastructure half a decade ago. AI is now dissolving the wedge.
Founders have better models, better tooling, faster distribution; products that once took a Series A to validate now take a weekend to create. But the same collapse cuts the other way. Enterprises that used to be your buyers are now well-positioned to build bespoke tools in-house, and the incumbent who used to need years to copy you can ship your wedge as a feature internally. The narrow workflow gets squeezed from both directions, and the strategy that feels like derisking is actually a bet on a kind of defensibility that no longer exists.
Surface Area
The truly epic companies of the late 2020s will have “large surface area.” Surface area is the amount of a customer's life or business a company has permission to operate inside of. If a software vendor holds the trust, context, distribution, data, or the system of record for an industry, then they can create holistic experiences that no one else can offer.
Imagine that you are a software company that sells items online. Through servicing order requests, your company has access to customers' purchasing habits, what they enjoy and what they don't, when they spend and when they don't. In the past, to connect these different stores of information you needed consistent human attention, making this experience unscalable.
In the future, I would imagine that a really great ecommerce experience will be able to proactively recommend items to purchase, provide you insights based on your site activity and the time of year.
This concept applies to many industries, and the best companies will find new ways to combine surfaces that used to live in separate businesses to create new value for customers.
Consequently, creating companies with large surface area actually requires more human interaction, not less. Traditionally, actually building software was the main function of technology companies. As the process of creating complex software becomes less and less involved, value creation and product insights will come mostly from in-person relationships.
It's impossible to automate human interaction, and in my experience getting consistent facetime with customers is actually the strongest unlock for understanding what is worth building. Replicating these interactions across many customers who each have their own unique cases is how great companies will scale.
Closing
It really is a strange time to start something new, and because the equation for success changes almost daily, it's not clear when or what to start. For many, there is an urgency to create something just to stay relevant, and I think that intuition is correct. Building an AI company is a real path to creating wealth, but the companies worth remembering will be the ones built on big shots, large swings, and hail marys.
Extras:
Recent Music Worth Listening To:
Hyper Pop is a genre has been getting lots of eartime for me recently. I was a huge fan of I Love My Computer by Ninajirachi, and HALO by Tiffany Day scratches that same itch. EVERYTHING I'VE EVER WANTED into DOIT4ME is probably my favorite two song run this year.
Jeon Jin Hee, is a Korean artist that is new to me. She has a rare ability to mix strong instrumentation with sublime imagery, truly beautiful. Breathing is an album that explores human emotion month by month, I really like Breathing in May from this album. If you're ever feeling sad throw on that Jeon.
Backstage Hologram by Kuru consistently makes me feel like im a runwaymodel walking in Paris Fashion Week. Its a rare, but elating feeling. FW19 and Shibuya transfer are both great.
Photos Worth Looking At:
NikonF3 w Kodak Tri-X 400, APR 2026
Fujifilm x100vi, MAY 2026
Fujifilm x100vi, MAY 2026
Check out my most recent photos from April and May, and subscribe to my newsletter.