Google's Bard launches in Europe and Brazil

Bard claims to offer "more coherent, consistent and creative responses" compared to ChatGPT through its use of "causal reasoning." Like ChatGPT, Bard allows users to have open-ended conversations and answer complex questions.
 

By Matthaios Tsimitakis
July 13, 2023
You can download this article in PDF format here!
Bard positions itself as a rival to OpenAI's popular ChatGPT, claiming to offer more creative and coherent responses. Bard allows users to have open-ended conversations and get answers to complex questions across a wide range of topics, similar to ChatGPT. 

"Starting today, you can collaborate with Bard in over 40 languages, including Arabic, Chinese, German, Hindi and Spanish," Google senior product director, Jack Krawczyk said in a blog post.

Bard was trained on a smaller dataset than ChatGPT, consisting of 1.5 trillion word embeddings compared to ChatGPT's 175 billion words. The company claims this makes Bard more environmentally sustainable while still delivering high performance. However, experts say Bard still has catching up to do to match ChatGPT's scale and capabilities. OpenAI's model has been tested by over 175 million users, while Bard is just launching.

"We need to be vigilant about the use of LLMs like Bard. These technologies have the potential to be very powerful, and we need to make sure that they are used in a way that respects people's privacy" said Marietje Schaapveld of Privacy International in a statement. The European Commission also stressed the importance of Google's compliance with GDPR. "We welcome the launch of Bard in Europe, and we are glad that Google has taken the time to address our concerns about privacy. We will continue to monitor the company's compliance with the GDPR, and we will work with them to ensure that Bard is used in a responsible way."

As large language models like ChatGPT and Bard become more common, experts say developing interpretability and oversight will be crucial to ensuring they develop responsibly. The differences in the causal reasoning approaches of Bard and Claude could also stimulate further innovation in the field. But for now, OpenAI's enormous lead in scale and data puts ChatGPT firmly at the front of the pack.

Here is a list of pros and cons between the three rivals in the LLM race

Pros of Bard:
• Bard's creators claim it provides more creative and coherent responses than ChatGPT as it can research current information such as recent events. This could make it more useful for applications like content generation
• Bard has a smaller carbon footprint than ChatGPT since it was trained on fewer parameters. This makes it a more environmentally sustainable option.

Cons of Bard:
• Bard is still in its early stages and has not been tested at the same scale as ChatGPT, which has over 175 million users. It remains to be seen how well Bard performs in real-world applications.  
• Bard has a more limited knowledge base than ChatGPT since it was trained on less data. This could make it provide incorrect answers at times, especially for complex or specialized questions.

Claude, another AI language model from Anthropic, claims to offer "more factual, coherent and consistent responses" than ChatGPT through its use of causal reasoning and a "causal filter." However, Claude has not yet launched publicly. 

ChatGPT's pros include its large knowledge base, high performance, and OpenAI's resources for continued improvement. Cons include the potential for generating inaccurate, unsafe, or biased responses.