Mike Ritchie
@thisritchie
Founder @definiteapp (a data team that never sleeps)
It's a massive pain to be "data driven". You need: * a place to store all your data (datalake) * pipelines to get data into the datalake (ETL) * a workspace to munge data and share analysis (BI) * a data team to set all this up and answer questions Skip all of this. Get…
This is how @SlackHQ can lose. I thought @GlueAI looked cool, but never cool enough to switch. This is the sort of shit that makes me reconsider.
The new @SlackHQ API rate limits are absolutely insane: For accessing messages, you can now only make 1 request per minute, with a maximum of 15 messages 🤯 And yes this applies to ALL apps, not just public ones. We can't build internal tools on our OWN data. @Benioff this is…
feels weird to pay $8.49/mo to get hundreds in value what value are you putting on your time?
feels weird to pay $8.49/mo to use a local model
if you include: * picking the right model for the task (e.g. is the context window big enough for the task? does the latency of the model match the use case? etc.) * setting the config (e.g. max token output) * setting the context (e.g. what info does the model need from…
Is prompt engineering really a skill?
anthropic has built-in tools (e.g. text_editor_20250429) that claude4 was **trained** to use. Claude Code uses these tools very well. Cursor built their own tools which claude is not as good at using. docs.anthropic.com/en/docs/agents…
opus-5.x is going to be fucking nuts. The training set from Claude Code is pure gold.
sonnet works better in claude code than cursor... i think this is part of the reason
sonnet often tries to use Anthropic's built-in tools which don't exist in cursor but work fine in claude code docs.anthropic.com/en/docs/agents…
by the way, the science to prompting is knowing what the fuck you’re talking about in the first place
you have my follow sir
I just call it thinking but yeah it’s great
"AI is the next crypto" ≈ "The Industrial Revolution is the next Tulip Mania"