The edge in lean, conviction-driven investing isn't access. It's walking into the first call already fluent in the market, the comps, and the competitive position.
That prep is expensive. At our size, it either crowds out founder time or doesn't get done.
So we built Flight Deck. What it does, and what we don't let it do: https://t.co/LuEEJ0X6g4
H/T @fredwilson, who recently described a strikingly similar operating model
Sophisticated investors are pulling back from software entirely. The reasoning is coherent. The conclusion is wrong.
"Software is risky" is not the same as "all software is equally risky."
Enterprise infra with deep workflow lock-in is not facing the same AI displacement risk as thin SMB SaaS. Software as a monolithic term is no longer relevant. Nuance matters.
I first met @MikayelK, @Wirestock's CEO, for a walk along Marina Boulevard. Before even talking about Wirestock, we talked about what AI actually looks like if LLMs aren't the whole story. Compute-intensive language models are powerful, but for real-world use cases, alternative architectures may win. The next wave of AI won't run on text alone. Vision models, robotics, world models. And every one of them has the same dependency: proprietary, high-quality multimodal data that synthetic generation simply cannot replicate. Wirestock is built for that world.
The internet has been scraped. Text data is largely tapped out, and every major AI lab knows it. The next frontier of model capability isn't more text - it's multimodal: image, video, motion, the messy visual complexity of the real world. And unlike text, that data can't be crawled. It has to be created, curated, and rights-cleared.
@supercruisecap is proud of what Wirestock has accomplished and are excited to support their next phase of growth!
Yesterday, we announced a $23 million Series A led by Nava Ventures. This funding will enable us to continue scaling our data capabilities to meet the growing demand from the world's leading AI labs for premium, ethically sourced multimodal training data.
Read more on our blog: https://t.co/e2ob34EPPa
We just raised $23M Series A led by @NavaVenturesVC to continue building the infrastructure that the next generation of AI models will be trained on.
Read the full story on @TechCrunch: https://t.co/d8umALeVQ4
1) I’m excited to announce @EqualVentures’s 5th Annual Emerging Manager Circle Summit
8 years ago, this was just a few of us swapping stories, trying to help each other. I couldn’t imagine what EMC would become
Now, we need events like EMC more than ever
https://t.co/7Z2QycjJhl
Around this point in Stripe's history, top talent had vested and VCs were actively recruiting anyone with Stripe DNA
Comparatively, you're not seeing that from OAI or Anthropic yet. The comp is too good, the secondaries are flowing, and AI is the only game in town
When the Labs start spinning out founders or VCs are exclusively hiring alums, start questioning where we are in the cycle
This is incredibly well laid out and I don’t think enough people are focused on it yet
World models are the next frontier of AI and literally where rubber meets the road for AI
https://t.co/GinKDE8g6z
It's been 2 years since my best friend and partner, Chris Martin (not from Coldplay), joined me at @supercruisecap
I can't even describe the positive impact he's had
There really is nothing better than working with your friend each and every day 🙌🙏
The vertical SaaS pitch: "Deep workflows create defensibility."
The reality: You need dominant market share for a data moat. Vertical startups don't have that. Your competitor has similar data from their slice of the market.
Meanwhile AI just made building vertical software 10x faster. Your 2-year head start now faces 20 competitors who shipped in 3 months.
And those "extra" engineers at the incumbent getting 2x more productive? They're building your exact vertical product on top of a platform with actual distribution.
The only real moat left is technical differentiation. A genuinely better model. Faster inference. A proprietary data pipeline.
"We're vertical" is not a strategy. It's just market selection.
@credistick@PeterJ_Walker What happens to the secondary market if these 5-10 companies do go public?
All the secondaries markets companies of the past tell you volume dries up, it’s not reallocated to 11-30
The top 10 private companies capture 76% of all secondary demand.
That's not a market. That's a crowded trade with a nice name.
At a dinner the other week, @PeterJ_Walker shared Cartа's data, which makes the bifurcation impossible to ignore: parabolic valuations at the top, hollowing of the middle.
Some of the best companies aren't invisible because they're bad. They're invisible because the funds got too big to see them or lack the courage to be different.
More thoughts on this in the comments.