💯 agree with this. Most enterprise problems can be solved without frontier models as cost / efficiency is key. They could even solve a large part of their needs with open source models, however I do see a lot of enterprises caring about the fact that open source models are mostly Chinese and so they will opt for less capable models from the top labs
I hear you, that’s my exact problem too. It’s not so much that I don’t know what to build it’s more than I can build a thousands things but want to build something that is worth building.
That’s when you need to connect with like minded people over a beer and shoot the sh1t.
Next time I’m heading to the West coast I’ll give you a shout. Or we can have a beer over zoom
@DavidOndrej1 I posted this just the other day...
"It feels like people are comparing a clean Hermes install to a heavily customised OpenClaw setup and maybe not making a fair comparison Could this be the reason Hemes feels 'more stable'?"
@BenjaminDEKR Not sure I agree
It's not like Elon to give up, so I could see Grok becoming a frontier model and SpaceXAI being a hardware / compute provider at the same time.
What am I missing?
SpaceX deal seems like a good use of hardware that is sitting idle.
My question is "why" is it sitting idle?
What is Grok missing that Anthropic and OpenAI have?
SpaceX/Grok has the power of Elon + enormous compute available, so is it AI talent that is missing?
It feels like people are comparing a clean Hermes install to a heavily customised OpenClaw setup and maybe not making a fair comparison
Could this be the reason Hemes feels 'more stable'?
Effective today, we are:
1) Doubling Claude Code’s 5-hour rate limits for Pro, Max, and Team plans;
2) Removing the peak hours limit reduction on Claude Code for Pro and Max plans; and
3) Substantially raising our API rate limits for Opus models.