Bard 可以一定程度上替代 GPT4V 了

ChatGPT 4V 不太好访问,对网络要求很高,而且现在停止订阅了,这个时候不妨尝试一下 bard ,上了 gemini 以后,能力有大提升,能访问网络,还能识别图片,对网络要求也不是很高。

就是别输入敏感信息,会有人工审查(页面上有提示)

gemini终于生出来了?

这显然还是PaLM。

2023.12.06

Bard is getting its biggest upgrade yet with Gemini Pro

  • What: Starting today, we’re introducing Gemini Pro in Bard, for Bard’s biggest upgrade yet. We’ve specifically tuned Gemini Pro in Bard to be far more capable at things like understanding and summarizing, reasoning, coding, and planning.
    You can try out Bard with Gemini Pro for text-based prompts, with support for other modalities coming soon. It will be available in English in more than 170 countries and territories to start, and come to more languages and places, like Europe, in the near future.

https://bard.google.com/updates

即使用英文问,他自己也是这么说的:

Which basic model are you basing it on?
My primary base model is currently Pathways Language Model 2 (PaLM 2), which was unveiled by Google AI in late 2022. PaLM 2 is a powerful and versatile large language model (LLM) built upon the Transformer neural network architecture. It offers significant improvements in performance and efficiency compared to its predecessor, PaLM.
However, my development also incorporates elements from other models, including:
LaMDA: This was the initial foundation for Bard, and some aspects of its architecture and training data still contribute to my capabilities.
Other LLMs from Google AI: These models provide additional training data and insights that help me learn and improve across a wider range of tasks.
External datasets and research: My knowledge base continues to grow through continuous learning from various sources, including publicly available datasets and ongoing research in the field of natural language processing.
Overall, while PaLM 2 serves as the primary framework for my operation, I benefit from a combination of technologies and knowledge sources to achieve my current capabilities.
This allows me to continually learn and evolve to become a more comprehensive and informative language model.

测试后,还是那个智商欠费的样子