Bard, Google’s alternative to ChatGPT, is now accessible to a larger audience thanks to the company’s debut of the generative AI chatbot in more than 40 languages and, following an initial delay due to data privacy concerns, its arrival in the European Union (EU).
A number of additional features were also added to Bard by the internet behemoth, but some are initially only accessible in English.
In what appeared to be a hasty response to the meteoric rise of ChatGPT, a super intelligent search engine and chatbot that relies on large language models (LLMs) to produce original content from straightforward instructions, Google initially teased Bard back in February. The creator of ChatGPT is OpenAI, a firm heavily backed by Google’s rival Microsoft.
Although Bard first became available for early access in English in the United States and the United Kingdom in March, the first waitlist was finally eliminated in May with a global launch that included support for almost 180 countries as well as additional languages including Japanese and Korean.
However, the EU has been notably absent thus far, with Google delaying the rollout there after a privacy regulator raised concerns. When businesses use Ireland as their European headquarters, the Irish Data Protection Commission (DPC), which oversees data protection in the EU region, said that while Google had informed the DPC of its plans to launch Bard in the EU, it hadn’t given the regulator enough information to allay its concerns regarding data privacy.
It appears that Google has now provided the DPC with what it was looking for with today’s debut.
In a blog post, Bard product lead Jack Krawczyk and VP of engineering Amarnag Subramanya stated that they had “proactively engaged with experts, policymakers, and privacy regulators on this expansion.”
As a matter of fact, Google has hailed its most recent version as its “biggest expansion to date,” extending it out to the majority of the world with support for Hindi, Arabic, Spanish, Chinese, and five more languages. Bard is now accessible in Brazil in addition to the EU.
Fine-tuning
New features that improve Bard’s responses and boost the chatbot’s potential for productivity coincide with the expansion. Early in May, several were telegraphed and given a preview, but today is the day they go public.
Users can now choose from five different options to alter the tone and manner of Bard’s responses: “simple,” “long,” “short,” “professional,” or “casual.” The toggle, which is initially only available in English, modifies Bard’s standard responses to a prompt to fit the user’s chosen tone and writing style.
A new text-to-speech AI capability allows Bard to now vocalize its responses, among other things. By selecting the new sound symbol next to a prompt, the chatbot’s auditory responses, which support more than 40 languages, can be accessible.
Bard may now export code to more locations, notably Python code to Replit, a browser-based integrated development environment, which increases efficiency.
Users can upload photographs with prompts (currently only in English) and Bard will review the image. photographs can be used in prompts.
Users may now pin, rename, and pick up recent discussions with Bard thanks to new features. Additionally, links now make it simpler for others outside of Bard’s circle to get his responses.
According to Krawczyk and Subramanya’s writings, human creativity is fueled by curiosity and imagination. “That’s why we created Bard: to help you explore that curiosity, enhance your imagination, and eventually get your ideas off the ground — not just by answering your questions, but by helping you build on them,” the authors write.
Early on in Bard’s development, Google battled significantly to match the level of responses provided by competing chatbots like ChatGPT. Even Google personnel called the chatbot “worse than useless” and a “pathological liar” since it provided factually wrong replies with made-up sources. (At the start of Bard, the company’s shares briefly fell by 8%.)
Google, though, asserts that Bard is growing in quantifiable ways, especially in subjects like math and programming. Additionally, it now has more features, like the capacity to explain code, organize data in a table, and surface images in responses. These extensions come from both Google’s own apps and services as well as third-party partners like Adobe.
However, Bloomberg’s investigation last week indicated that the people who instruct Bard are frequently overworked and underpaid, which is yet another bad sign for Google. Some contractors work for as little as $14 an hour, with little training, and are required to finish intricate audits of Bard quickly.
Bloomberg’s report comes in the wake of an Insider investigation published in April, which revealed that Bard testing contractors were not given enough time to confirm and double-check the chatbot’s most accurate response. It appears that nothing has changed in that regard.
