A bit of shade being thrown on OpenEuroLLM for only having a small investment of โฌ54m by the Euro Commission, but don't forget @deepseek_ai spent a lot less, and look what happened there...
So I'm wondering how much of this new o3-mini model was influenced by @deepseek_ai R1, or if this was already in the works and we've yet to see the real R1 impact in @OpenAI GPT models...
OpenAI just dropped o3-mini and itโs insane!
It's their fastest, smartest, and most cost-effective reasoning model yet and people are going absolutely crazy over it.
10 wild examples:
1. "create a snake game, where snakes compete with each other"
I do realize that other AI providers do this but you can opt out (for a price).
Also I do realize you can operate Deepseek offline but that doesn't mean it won't phone home when it goes online to be updated.
I'm seeing lots of good things about @deepseek_ai but as the old saying goes "if something is free, you are likely the product". It concerns me that your data and inputs might be sent "home" to Deepseek's servers at some point.
Can any of the gigabrains confirm this?
HMU on Threads by @instagram. Sign up process is super slick. Gonna be interesting to see how much market share this takes from @Twitter https://t.co/Sqh6LxARsT
Currently testing out @synthesiaIO for a business intro presentation. So far the UI is friendly but it's taking far longer to transpose the existing @powerpoint than it should do due to the import feature not being rich enough.