News! We just launched “The General Podcast” 🎙️
It’s a ‘hostless' show that pairs tech and business leaders together for honest, behind-the-scenes conversations about building:
🎙️ Ep 1: Katie Dill (@lil_dill) x Randy Hunt (@randyjhunt)
🎙️ Ep 2: A healthcare journalist x a Pulitzer Prize-winning author
🎙️ Ep 3: Two rising media founders
🎙️ Ep 4: Two legendary COOs
And many more. So why are we doing this?
The best conversations don’t happen in interviews. They happen between peers—people solving similar problems, or shaped by shared experience across an entire career.
We built a show around that. Listen 👇
The fact that larger models keep impriving when trained repeatedly on the same data with the right recipe means more compute can make existing data go further.
Great read from @industriaalist and the Qlabs team.
AI should get better as people use it. @TrajectoryLabs is building the tools to make that happen.
Proud to partner with @rronak_ and the team and excited to see this launch. Let's keep building 💪
Imagine if everyone had the tools to shape intelligence itself. That day is today.
Today we launch the Trajectory platform - the easiest and most intuitive tool to post-train models of your own. The foundation for continual learning.
With just a few agent prompts, you get access to a full creative suite to train, eval, and deploy trillion parameter models.
On top of that, we believe in a future where you don’t just shape your model once, it also keeps learning on its own from the world around it.
We’re releasing two research previews.
1. Continual grading scales your taste, so your model learns what good looks like from far more than you could ever grade by hand.
2. GS1, a new algorithm that lets models learn directly from what they do in production.
We’ve now raised 60M, with a new round led by @Sequoia at a 300M valuation to democratize the most powerful technology of today - training models - so that the lawyers, doctors, and engineers of the world can own their own intelligence.
You know what good looks like, now go build it. Own your own value function with Trajectory.
The first public workshop on Lanyon at the Vibecheck Hackathon in SF.
So many interesting projects! From using Lanyon to model articulation constraints in robots, to producing a full flight simulator, to modeling integrable systems, to solving the Klein-Gordon equation!
130K one-person businesses did $1M+ in revenue last year, up 11% (via Forbes). AI is making the solo company a real thing, and it still needs a finance department. Proud to back @CollectiveFin and @hoomanradfar as they build it. Congrats on the new round and massive growth!
It's been awesome building with @hoomanradfar and the entire @CollectiveFin team over the last few years.
To the people who worry AI will destroy jobs - here is an example of a AI making it easier and more approachable to be an entrepreneur and earn a living based on your unique skills and creativity (instead of the ability to organize a back office finance function and manage compliance and tax). (cc @thegp)
RoTo (return on tokens) is the best way to understand the health of a software business today.
RoTo = $ revenue captured / $ inference spend. For every dollar spent on inference, how much revenue does your application capture?
More from @phineasb on RoTo as the new standard for measuring AI businesses:
Founder email to @theGP team today about what it's like to build with us:
"I wanted to send a quick note as we wrap up our talent engagement. We ironed out the last details with ___, and she starts Wednesday. With that, you have supported the hiring of four candidates, half our FTEs, all PhDs. You helped me attract people with specific technical experience where my network was limited, and get me where I needed to be."
When people say they "help" with recruiting, ask if this is what they mean.
It’s been awesome building with @capitaltruist and the whole @falconer_ai team.
Don’t burn tokens making up new shit everytime there’s a question— the best answers are inside your company already — @falconer_ai brings you the answers you need when you need them.
Every company is about to have more AI agents than employees, and both are useless without knowledge they can trust. Dave built that layer at Stripe and Uber. Falconer brings it to everyone including Fireworks Thatch, Payabli and many more. Proud to be backing this team