3 ms·

I started trying out 3.7 Flash this week and it is competitive with opus/fable and also FAST. It is getting work done that anthropic models were struggling wit

I started trying out 3.7 Flash this week and it is competitive with opus/fable and also FAST. It is getting work done that anthropic models were struggling with and the speed with which it does is quite a bit noticeably faster. Beginning to think Google is a dark horse in this race and some of Anthropic's "everything feels janky and rushed" karma is going to catch up.

re-thc 24d ago
> Beginning to think Google is a dark horse in this race Google 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.
- colechristensen 24d ago
  Hype that burned out pretty quickly, it's hard to speak to the size and significance of old hype, I never felt it. Every 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.
cousinbryce 24d ago
I use Gemini because I feel like Google will win the AI race, and it’s Good Enough
- throwaway219450 24d ago
  I 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.
  - spacebanana7 24d ago
    I 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. Gemini’s integration with maps and search is more important for Google.
    - cousinbryce 17d ago
      That’s a good point, and probably why it’s a multimodal model. I wouldn’t be surprised if it’s over index on code written by Googlers.
- baq 23d ago
  My long term base case, too, but they need to install Demis as CEO and I don’t think either side is ready.

re-thc 24d ago

> Beginning to think Google is a dark horse in this race Google 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.

colechristensen 24d ago
Hype that burned out pretty quickly, it's hard to speak to the size and significance of old hype, I never felt it. Every 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.

cousinbryce 24d ago

I use Gemini because I feel like Google will win the AI race, and it’s Good Enough

throwaway219450 24d ago
I 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.
- spacebanana7 24d ago
  I 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. Gemini’s integration with maps and search is more important for Google.
  - cousinbryce 17d ago
    That’s a good point, and probably why it’s a multimodal model. I wouldn’t be surprised if it’s over index on code written by Googlers.
baq 23d ago
My long term base case, too, but they need to install Demis as CEO and I don’t think either side is ready.