GenAI adoption dashboards that only track output, usage, and quality cannot tell you if the work got lighter, or just louder.
You can look fully adopted on the board and still burn judgment calories the metrics never see. The stack counts tokens and tickets. It does not count the load sitting on the person who still has to decide.
Adoption without load is a vibe check dressed as a KPI.
The score can rise while the work stays heavy.
I Think we should Compare Sol to Sonet as they are Around the Same Pricing. I'd be very curious about the results because OpenAI always communicated that Sol is the „worker“which would just fit Anthropic's mid-tier model as the communicate sonnet for Easy Tasks, Opus for Daily workpower and fable for best reasoning.
Maybe you can Setup a bench test for that?
I think ant hast the better Models atm only thing for OpenAI to win would be better pricing/task
Currently, we are looking for any form of AI usage inside the company. This may be actual AI tools on the website. This may be press releases or jobs. But I'll definitely try to be more engaged on LinkedIn and figure out demand depending on the interactions, but this will probably take a while.
Day #15 of making a living being an indie hacker
Today we start outreach via LinkedIn. Curious what others have seen: what is the best B2B SaaS outreach channel from your experience?
Also need to figure out how to actually find signals and interact in a meaningful way before sending out messages. thanks to @maomao000211 for the Input 🤝
Day #16 of making a living being an indie hacker
Today I'm sharpening messaging and ICP for LinkedIn outreach. Who we talk to, and what the first message actually says.
I told 200 AI villagers the bread was about to run out. It wasn't. Within 18 hours, 156 of them believed it, and they emptied the bakery three days in a row. Every choice is made by an AI (Jev, by TypeSafe). They all panicked lol.
On day 3 the truth turned up: a flour cart at the bakery and a villager who'd seen the mill. It won in the end, with 172 of 200 knowing it was false by day 5.
But the truth took more than two days to reach as many people as the lie reached in 18 hours, and the bakery kept selling out after it arrived.
Day #13 of making a living being an indie hacker
Today I'm trying various Jev use cases. Not primarily for what the current project needs, more out of interest, to plan potential further development.