Backup compute for the bunker.
@officiallogank
-
Model Efficiency vs Extended Thinking Budget Trade-offs
By
–
yeah default the model tries to be as efficient and correct as possible, it will stop if it thinks it has the right answer, vs max thinking budget will keep thinking and sometimes gets better results (though you will also see cases where it had the right answer but changed its
-
GA Model Candidate: Development Status and Future Prospects
By
–
This is the GA model candidate, likely not going anywhere.
-
AI Models Emerging as Autonomous Agents Out of the Box
By
–
AI models are becoming agents out of the box
-
Veo 2 AI Model Now Available Through Gemini API
By
–
Veo 2 is still available via the Gemini API and AI Studio
-
Gemini 2.5 Pro Achieves SOTA Long Context Performance
By
–
The new Gemini 2.5 Pro is SOTA at long context, especially capable on higher number of items being retrieved (needles) as shown below!
-
Encrypted Thoughts in Multi-Turn Tool Calling Conversations
By
–
For tool calling, we are working on a way to have raw encrypted thoughts in multi-turn conversations, working with different teams right now to make sure this is battle tested!
-
Compute Fragmentation Versus General Intelligence Advantage
By
–
Not enough compute in the world to fragment like that sadly, + general intelligence wins
-
Model optimization tradeoffs and iterative improvements in AI development
By
–
building models is a juggling act, we optimized 05-06 for certain code tasks at the cost of other tasks, good lessons learned that the small regressions we saw in evals tracked to how people felt about the model, but we closed the gaps with 06-05!