I give it a year until we see a new breed of AI native private equity firms that acquire companies just so they can move their workflows from Claude to open source Chinese models and flip them.
we are offering discounted tokens and certainty on capacity availability in exchange for 1-3 year commits.
we expect that the world will feel increasingly capacity constrained for the next while, as models continue to get much more useful.
@HankCouture we use tailscale and gitops-managed EKS
was a bit to setup but trivial to deploy new Claude-coded apps
and the platform stack itself is built to be claude-codeable
curious that both Anthropic and OpenAI are moving natively (in billing atl) into other clouds
thought they were playing to disrupt the clouds as they move into hosted services (i.e. workspaces and managed agents) as inference commoditizes and continues to race to the bottom
some scribbles on possible motivations:
- perhaps still good comp adv to let someone else run the VMs and they want to own higher abstractions (sensical for Anthropic w/ no metal, but for OpenAI w/ stargate buildout?)
- perhaps this is a concession they’re giving to get compute
- or maybe the agreements don’t exist, this is sama’s way to make every cloud ceo assume he’s got a deal w/ their competitors and get better deals w/ them and then lev over msft?
- anthropic has had Claude in GCP/AWS for a while actually, maybe that was bleeding OpenAI w/ enterprise? committed spend + startup credits go brrrr ig
we have updated our partnership with microsoft.
microsoft will remain our primary cloud partner, but we are now able to make our products and services available across all clouds.
will continue to provide them with models and products until 2032, and a revenue share through 2030.
@dennizor ultimately this looks more like a data warehouse even when it presents as an app builder
lot of solved data eng problems underneath: branching, lineage, metadata
with that, API/CLI/MCP over robust multimodal tractional store w/ ample metadata is codegen
oh heck yeah
we’re just exploring building something in the space
opinionated data stack but w/ agent-friendly declarative everything and operational apps as the end goal, not just BI
would love to chat if you can share more about your project
still a prototype but these are exactly the use cases we want to specialize for
If you are seriously AGI-pilled, then one weird implication in the limit is that “talent” seemingly stops mattering as much for company success. It just becomes a game of hard power: access to the very best AI models, compute, data, land, etc.
TBPN has been acquired by OpenAI
The world is changing quickly but TBPN will stay the same. Live every weekday just with a lot more resources.
Thank you to everyone that has been a part of this journey big or small. We are 17 months in and unironically just getting started.
everyone's dunking on lovable going horizontal. imo too much focus on output and no attention on input
input modes > output modes
engineers get magic from claude code because code is precise by nature
non-technical builders haven't had a lingua franca with the tools, but if lovable speaks spreadsheet... biz logic already encoded in formulas and column names is a plenty precise prompt
file-to-app is the killer launch
Introducing Lovable for more general tasks.
Lovable has always been for building apps. Today it also becomes your data scientist, your business analyst, your deck builder, and your marketing assistant.
This is a big step toward what Lovable is becoming: a general-purpose co-founder that can do anything.
See examples below.
@shiftj we’re doing this at scale for PE funds + their portfolios
still moving like a software company, progressively automating the internals
we’re hiring
Chicago-based @tractorbeamai is hiring engineers
Role sits at the intersection of AI strategy and implementation
They're building out of out Fulton Market office
https://t.co/48Pd7EDbNs
but actually, if we kind of accept that "an AI Agent is basically a for loop", then what's the lispy version of this?
> an AI Agent is a foldL where the accumulator is a context window, and the reducer is an LLM DetermineNextStep + a switch statement on how to handle it