I use Gemini because I feel like Google will win the AI race, and it’s Good Enough
replyMy long term base case, too, but they need to install Demis as CEO and I don’t think either side is ready.
replyI want to believe this, but every time I try Gemini coding assistance within Colab it's utterly dire. Code gen in a cell is OK, but things fall apart when you try to get into a feedback loop. The system prompt/harness fails to inform the agent about what it can and can't do, or does and doesn't have access to. It will confidently tell you it's done a thing, and then you ask, it admits can't actually do that but will happily try and fail again. Very frustrating, because I really like Colab as a platform for little reproducible experiments that may or may not require CUDA.
replyI suspect Gemini doesn’t focus that much on coding, at least compared to frontier labs. It’s not part of Gemini’s core business case.
replyGemini’s integration with maps and search is more important for Google.
> Beginning to think Google is a dark horse in this race
replyGoogle was so hyped up early Gemini 3 era (only some months ago). And now dark horse? The TPU takeover almost crashed nvidia and everyone else.
Hype that burned out pretty quickly, it's hard to speak to the size and significance of old hype, I never felt it.
replyEvery time I personally tried Gemini models up until last week they simply couldn't do the long complex tasks I'd being doing with Anthropic models for many months.