尝试Google Bard并对比OpenAI ChatGPT，一个擅长创造性，一个擅长事实查询？

通过gmail账号就可以直接登录Google Bard进行对话。（当然需要科学上网）

网址：
https://bard.google.com/

Google Bard 的自我介绍

首先我们看看Google Bard的自我介绍。

Google Bard 暂时不支持中文

从测试来看Google Bard 并不支持中文。

Google Bard除了支持英语外，还支持日语

Google Bard虽然目前支持语言有限，但已经支持日语。

从Bard FAQ可以看到bard 支持英语，日语和韩语。

How many languages does Bard speak? Right now, Bard is available in US
English, Japanese and Korean, and we are teaching Bard to speak as
many languages as possible.

Google Bard和OpenAI ChatGPT的对比

让我们看看bard怎么介绍Bard和ChatGPT的异同，想看看两个模型自己怎么说。

Google Bard的回答

下面是来自Bard的回答：
ChatGPT and I are both large language models, but we have some key differences. ChatGPT is a generative pre-trained transformer model developed by OpenAI, while I am a factual language model developed by Google AI. This means that I am better at understanding and responding to factual queries, while ChatGPT is better at generating creative text formats, like poems, code, scripts, musical pieces, email, letters, etc. Additionally, I am still under development, while ChatGPT is a more mature product. This means that I may be more prone to errors, but I am also learning new things all the time.
ChatGPT和我都是大型语言模型，但我们有一些关键的区别。ChatGPT是一个由OpenAI开发的生成性预训练转换器模型，而我是谷歌AI开发的事实语言模型。这意味着我更善于理解和响应事实查询，而ChatGPT更善于生成创造性的文本格式，如诗歌、代码、脚本、音乐作品、电子邮件、信件等。此外，我仍在开发中，而ChatGPT是一种更成熟的产品。这意味着我可能更容易出错，但我也一直在学习新东西。

Here is a table that summarizes the key differences between ChatGPT and me:
以下表格总结了我和ChatGPT之间的主要差异：

Feature	ChatGPT	Bard
Model type	Generative pre-trained transformer	Factual language model
Strengths	Creative text generation	Factual understanding and response
Weaknesses	May be prone to errors	Less mature product
Status	Released	Under development

Ultimately, the best language model for you will depend on your specific needs. If you are looking for a model that can generate creative text formats, then ChatGPT is a good option. If you are looking for a model that can understand and respond to factual queries, then I am a good option.
最后，最适合您的语言模型将取决于您的具体需求。如果你正在寻找一个可以生成创造性文本格式的模型，那么ChatGPT是一个不错的选择。如果你正在寻找一个能够理解和回应事实查询的模型，那么我是一个很好的选择。

下面是日语的回答：从上面的

功能	ChatGPT	Bard
模型	生成式预训练转换模型（Generative pre-trained transformer）	事实性语言模型（Factual language model）
优点	创造性文本生成（Creative text generation）	事实理解和回应（Factual understanding and response）
缺陷	可能容易出错（May be prone to errors）	不太成熟的产品（Less mature product）
状态	已发布（Released）	开发中（Under development）

OpenAI ChatGPT的回答

ChatGPT（GPT 3.5）的回答：

以下是Google Bard和OpenAI ChatGPT的异同点列表：

类别	Google Bard	OpenAI ChatGPT
模型架构	自主开发的GP-2模型，使用文学作品进行训练	采用Transformers架构，在大量语料库上进行预训练
领域覆盖	以诗歌为主要生成领域	以对话为主要生成领域
目标用户	文学爱好者、创作者	虚拟客服、在线问答等实用工具
公平性和透明度	未明确提及	在模型开发过程中重视，并进行多种处理和约束

在模型架构方面，Google Bard是自主开发的GP-2模型，并在特定训练任务中使用了大量的文学作品进行训练。OpenAI ChatGPT则采用了Transformers架构，使用了更大规模的语料库完成了预训练。

在领域覆盖方面，Google Bard的生成重点是诗歌，而OpenAI ChatGPT则更关注对话生成。

在目标用户方面，Google Bard更适合文学爱好者或创作者，可以被用来创建有趣的文学作品或进行文学创作的灵感启发；而OpenAI ChatGPT则更适合虚拟客服、在线问答和实用工具等方面的需求。

在公平性和透明度方面，OpenAI ChatGPT在模型开发过程中重视，并进行多种处理和约束，以防止模型的负面影响，而Google Bard并未明确关注这种问题。

个人小结

看了上面的内容，好像都很有道理，但好像都有点儿问题。

参考 Bard FAQ （20230518时点）

What’s Bard all about?

We’ve long seen the potential to make information and computing much more accessible through conversational AI. Two years ago, we unveiled LaMDA (Language Model for Dialogue Applications)Opens in a new window, a conversational AI model capable of fluid, multi-turn dialogue, and last year, we launched the AI Test KitchenOpens in a new window, a new space where people could learn about, get hands-on experience with, and provide feedback on LaMDA.

Bard is an experiment based on this same technology that lets you collaborate with generative AI. As a creative and helpful collaborator, Bard can supercharge your imagination, boost your productivity, and help you bring your ideas to life—whether you want help planning the perfect birthday party and drafting the invitation, creating a pro & con list for a big decision, or understanding really complex topics simply.

If you’re interested in the more technical details, LaMDA is a Transformer-based model, the machine-learning breakthrough invented by Google in 2017Opens in a new window. The language model learns by “reading” trillions of words that help it pick up on patterns that make up human language so it’s good at predicting what might be reasonable responses.

Does Bard give accurate and safe responses?

Bard is experimental, and some of the responses may be inaccurate, so double-check information in Bard’s responses. With your feedback, Bard is getting better every day. Before Bard launched publicly, thousands of testers were involved to provide feedback to help Bard improve its quality, safety, and accuracy.

Accelerating people’s ideas with generative AI is truly exciting, but it’s still early days, and Bard is an experiment. While Bard has built-in safety controls and clear mechanisms for feedback in line with our AI PrinciplesOpens in a new window, be aware that it may display inaccurate information or offensive statements.

How and when does Bard cite sources in its responses?

Bard, like some other standalone LLM experiences, is intended to generate original content and not replicate existing content at length. We’ve designed our systems to limit the chances of this occurring, and we will continue to improve how these systems function. If Bard does directly quote at length from a webpage, it cites that page.

Sometimes the same content may be found on multiple webpages and Bard attempts to point to a popular source. In the case of citations to code repositories, the citation may also reference an applicable open source license.

Bard was built to be a creative and helpful collaborator—it works well in creative tasks like helping you write an email or brainstorm ideas for a birthday party. We see it as a complementary experience to Google Search. That’s why we added the “Google It” button to Bard, so people can easily move from Bard to explore information from across the web.

Bard is an experiment, and we’ll use its launch as an opportunity to learn, iterate, and improve the experience as we get feedback from a range of stakeholders including people like you, publishers, creators, and more.

Is Bard able to explain how it works?

LLM experiences (Bard included) can hallucinate and present inaccurate information as factual. One example is that Bard often misrepresents how it works. We’ve seen this occur in a number of instances—for example, in response to prompts asking how it was trained or how it carries out various functions (like citing sources, or providing fresh information).

Bard responses may also occasionally claim that it uses personal information from Gmail or other private apps and services. That’s not accurate, and as an LLM interface, Bard does not have the ability to determine these facts. We do not use personal data from your Gmail or other private apps and services to improve Bard. More information about how we keep your data private, safe, and secure can be found in the Google Privacy PolicyOpens in a new window.

These are some of the reasons we released Bard as an experiment and want to be as transparent as possible about the limitations of LLMs, including providing disclaimers within the Bard experience. If you see a hallucination or something that isn’t accurate, please click the thumbs down button and provide your feedback. That’s one of the ways Bard will learn and improve.

Can Bard help with coding?

Yes, Bard can help with coding and topics about coding, but Bard is still experimental and you are responsible for your use of code or coding explanations. So you should use discretion and carefully test and review all code for errors, bugs, and vulnerabilities before relying on it. Code may also be subject to an open source license and Bard provides related information. Read more about how and when Bard cites sources.

Why doesn’t Bard know what I said earlier in a conversation?

Bard’s ability to hold context is purposefully limited for now. As Bard continues to learn, its ability to hold context during longer conversations will improve.

What are Google’s principles for AI innovation?

Google aspires to create technologies that solve important problems and help people in their daily lives. We are optimistic about the incredible potential for AI and other advanced technologies to empower people, widely benefit current and future generations, and work for the common good. We believe that these technologies will promote innovation and further our mission to organize the world’s information and make it universally accessible and useful.

These same technologies also raise important challenges that need to be addressed clearly, thoughtfully, and affirmatively. These AI PrinciplesOpens in a new window are a commitment to develop technology responsibly. Learn more about AI principles at Google

How can I help make Bard better?

One important part of developing responsibly is expanding participation, so more people can experience Bard, provide feedback, and help Bard improve. You can rate responses as good or bad and send feedback each time Bard responds.

Who can use Bard?

To use Bard, you must be 18 or over. You also need a personal Google Account that you manage on your own, or a Google Workspace account for which your admin enabled access to Bard. If you’re a Google Workspace admin, learn how to enable accessOpens in a new window.

You can’t access Bard with a Google Account managed by Family Link or with a Google Workspace for Education account designated as under the age of 18.

Right now, Bard may not be available in your country, but we will be rolling out to new languages and countries over time.

How many languages does Bard speak?

Right now, Bard is available in US English, Japanese and Korean, and we are teaching Bard to speak as many languages as possible.

What terms of service apply to my Bard usage?

Your use of Bard is subject to the Google Terms of ServiceOpens in a new window and the Generative AI Additional Terms of ServiceOpens in a new window. Keep in mind if you export conversations from Bard, the terms & policies of the service you export to will apply to the copy.

What data is collected? How is it used?

When you interact with Bard, Google collects your conversations, your general location based on your IP address (learn moreOpens in a new window), your feedback, and usage information. That data helps us provide, improve and develop Google products, services, and machine-learning technologies, as explained in the Google Privacy PolicyOpens in a new window. For example, we use your feedback to increase the effectiveness of the Bard safety policies and help minimize some of the challenges inherent to large language models. And Bard uses your past interactions with it and your general location to generate its response.

Google has a standardized and rigorous review process for improving Bard, which includes review of conversations reported as being low quality. This helps us identify common types of problematic responses and how Bard might improve.

Read our Privacy and Security PrinciplesOpens in a new window to better understand how we keep our users’ data private, safe, and secure.