@KyleAsay_ for companies in that bucket, it's interesting to think about what the ideal staffing/headcount math for gtm looks like going forward. especially how you pay people.
i have low conviction on model routers - very open to changing my mind but this is a snapshot of my current thoughts
- i don't think it's good to not be aware of what model you're using. coding with LLMs is a skill you develop and getting a feel for models is part of that
- people (at scale) don't have this skill right now which is why a lot of companies are complaining that people are using expensive models for dumb things. a model router promises to solve this without the user having to do anything but i think the issue is missing feedback loops to the user. id rather we figure out how to help users get smarter
- i dont even know how much you can model route when factoring in things like prompt cache. only so much you can do
- their effectiveness is a bit exaggerated by the same dynamic that's impacting everything AI. so many companies desperately searching for opportunities and trying anything. model routing is the one thing models labs cannot do so everyone is jumping on it
very interested to see how these capabilities develop.
actually being proficient in navigating legal/procurement processes has always been an underappreciated gtm lever. far too often contract turns devolve into an intellectual exercise totally removed from closing out a deal.
"don't want to negotiate against ourselves" while digging heels in on pointless shit is negotiating against yourself. if you're not getting ahead of condensing timelines whenever you have the opportunity to do so, you're negotiating against yourself.
Wasting two weeks pushing your MSA knowing their first redline is just going to be inserting their template...
Contract negotiations are like poker games. The right answer depends on knowing your opponent, as much as knowing the law/rules. How good are frontier models at closing deals?
With @micro1_ai we benchmarked frontier models on multiple contract negotiations, across several turns. Rather than individual edits, we assessed the full sequence of judgment calls a lawyer makes across a deal lifecycle.
The headline: no model is close, and there are no standout winners yet.
@thedealdirector@turbopuffer ya..."ai native" get's thrown around a lot for new companies, then you see turbopuffer and are like oh that's what that means
LOOPS...!?
didn't Karpathy introduce this months ago via his /autoresearch project?
isn't that kinda why openclaw exploded?
the /autoresearch idea I found so interesting because IMO it really highlighted the importance of your problem selection + definition and translating that into a measurable/verifiable loop design.
having in seat domain expertise and applied business knowledge can be a form of 'taste' or a hard skill in this context.
somebody on each side needs to step up to close this communication gap and reset the board. can't be letting your guys get in these positions jesus christ lol.
It's back.
AIPCon 10 is about tradecraft: the individual expertise of our partners, amplified by LLMs and the Ontology.
Watch live and see how they are deploying Palantir to streamline farmer services at the USDA, modernize private equity fundraising at Kirkland & Ellis, optimize fleet operations at Hertz, encode 160 years of construction expertise at McCarthy Building, orchestrate supply chains at Nscale, transform parts identification at Parts Town, and help Accenture secure software supply chains, automate deployment, and enable rapid rollbacks with Palantir Apollo.
Streaming live on X and YouTube.
i really think the AI cost discussion is overblown. most companies are still in the discovery phase, which is more efficient with frontier models. but once you find valuable use cases, there are so many ways to drive efficiency... smaller models, tuned models, batch processing
Hmm emails look fine. Are you following up in the same thread?
Assuming there isn’t anything quirky causing them to be blocked / filtered to Spam I’d maybe try
1. Changing the hello@ sender email and name to something that makes it more clear it’s from you v. a system
2. Similarly, change icon to your headshot
3. FU in same thread and maybe work in some options they can just easily just say yes or no to