Joseph T. Hasselmann

Outdoor enthusiast, amateur cook, and founder of Evercrest Capital Partners @evercrestlp

[Joseph T. Hasselmann @jthassel]

Consensus is narrower. It is how the capital market reacts to what is already zeigeist. It updates in steps, round to round, narrative to narrative. The job of a GP is to see evidence the capital market has not yet priced, and to enter before consensus forms around it. 5/9

[Joseph T. Hasselmann @jthassel]

General Compute just placed a 9-figure order for @SambaNovaAI SN50 chips and will be the first neocloud deploying them.

This is one of the most important emerging dynamics in AI infra. Inference is becoming its own category entirely, given specific requirements around:
• token economics
• memory architecture
• power efficiency
• latency and throughput
• agent-to-agent tooling

CoreWeave/NVIDIA was the first landmark pairing in AI infra. Our bet at @evercrestlp is that GC + Samba is the next.

https://t.co/fijFEP0rri

[Joseph T. Hasselmann @jthassel]

@george_matelich ngl that’s a sad looking sando

[Joseph T. Hasselmann @jthassel]

Spot on. And we’re hiring for this type of role right now.

[Pavel Prata @pavelprata]

Imho, solo GPs probably won’t need a traditional CoS in the same way going forward, because most ops workflows (scheduling, CRM, admin coordination, etc.) will get heavily streamlined through agents/workflows over the next few years.

I think EMs will increasingly need what I’d call “gravity makers.”

I don’t even have a precise JD for these ppl yet, but you kinda know them when you meet them.

Usually they are:

> hungry college kids or young operators actively rejecting the traditional 10-year career path
> choosing venture intentionally, not because it’s “cool”
> obsessed with curiosity, focus, execution
> terminally online and monitoring the market daily
> consuming information from X/Substack/@tbpn/@MTSlive + old books/research papers at the same time
> hunters, not planters: they roll up sleeves and go online + IRL to find the most interesting founders before everyone else
> deeply plugged into emerging communities and niche ecosystems
> building media, distribution, automations, workflows from day one even with almost no resources
> AI-native and constantly experimenting with tools/workflows
> good at turning raw info/data into actual insights
> developing taste and chasing authenticity
> good writers who can articulate observations that later become contrarian theses
> able to jump from “build a Notion LP portal” to “make a G Sheet + Claude fund model” in the same afternoon
> low ego, humble, trust the process

And with the right GP, these ppl eventually become force multipliers that make 1+1=3 instead of just 1+1=1.5.

Because in a world where competition for access among funds is insane, the most important thing from day one is building enough "gravity" that the best founders naturally come to you first.

All of this probably sounds blurry and hard to define.

But ppl who are deep in the ecosystem probably know exactly the type I’m talking about.

[George Matelich @george_matelich]

Today, I’m excited to share that we’ve raised a $4.5 Million seed round led by @2048vc and @RangeVC with participation from @btv_vc.

[Joseph T. Hasselmann @jthassel]

Neither produces the full picture alone. Vaz is right that picking is the job. And concentration is the other coefficient that matters. Success in both requires an apparatus built for it. 9/9

[Joseph T. Hasselmann @jthassel]

Vaz's "Narrow Path" is the clearest articulation I've read of the fissures in the seed model. The position concentration math alone settles the argument.
One addition worth making, because it changes how you act on his conclusion... 1/9

[Lucas Vaz @lucasbagnocvaz]

https://t.co/wkTd7Thw67

[Joseph T. Hasselmann @jthassel]

It also clarifies the structure of the work. Evidence discovery is two things. An investor's perspective does the financial and market work. An operator's perspective does the categorical work: reading customer pull, technical adoption, talent density from inside the pattern. 8/9

[Joseph T. Hasselmann @jthassel]

Three implications: per-seat pricing is structurally broken. The value prop that sells is leverage, not headcount reduction. And the real bottleneck is change management, not model capability.

[Joseph T. Hasselmann @jthassel]

The agent transition is playing out a bit differently than many expected. A few observations, and fewer conclusions...a 🧵:

[Joseph T. Hasselmann @jthassel]

...The companies bridging that complexity gap for non-technical buyers are quietly building the most defensible positions in the market.