Google has begun testing an AI chatbot called Apprentice Bard that’s similar to ChatGPT, an online service that in many ways is more useful than Google search, CNBC reported this week.
And we might see it soon. While reporting gloomy fourth-quarter financial results, Chief Executive Sundar Pichai said Google plans to make its own AI language model tools available to the public “in coming weeks and months.”
ChatGPT uses artificial intelligence technology called a large language model, trained on vast swaths of data on the internet. That type of model uses an AI mechanism called a transformer that Google pioneered. ChatGPT’s success in everything from writing software, passing exams, and offering advice, in the style of the King James Bible, on removing a sandwich from a VCR has propelled it into the tech spotlight, even though its results can be misleading or wrong.
Google’s Apprentice Bard tool is similar to ChatGPT, but the company also is testing a question-and-answer format for Google search results, CNBC said. Google Research leader Jeff Dean told employees in a 2022 meeting that Google is being more conservative with its approach to avoid “reputational risk” of AI that offers bad responses.
Google didn’t comment on the specifics of the report, but spokesperson Lily Lin said in a statement, “We have long been focused on developing and deploying AI to improve people’s lives. … We continue to test our AI technology internally to make sure it’s helpful and safe, and we look forward to sharing more experiences externally soon.”
AI technology already is all around us, helping in everything from flagging credit card fraud to translating our speech into text messages. The ChatGPT technology has elevated expectations, though, so it’s clear the technology will become more important in our lives one way or another as we rely on digital assistants and online tools.
Google AI subsidiary DeepMind also is involved. Chief Executive Demis Hassabis told Time that his company is considering a 2023 private beta test of an AI chatbot called Sparrow.
Google detailed transformers in 2017, and it’s since become a fixture of some of the biggest AI systems out there. Nvidia’s new H100 processor, the top dog in the world of AI acceleration at least in terms of public speed tests, now includes specific circuitry to accelerate transformers.
The large language model (LLM) revolution in AI that resulted is useful for language-specific systems like ChatGPT, Google’s LaMDA and newer PaLM, and others from companies including AI21 Labs, Adept AI Labs and Cohere. But LLMs are used for other tasks, too, including stacking boxes and processing genetic data to hunt for new drugs. Notably, they’re good at generating text, which is why they can be used for answering questions.
Google, which endured bad publicity over the departure of AI researcher Timnit Gebru in 2020, has a program focusing on responsible AI and machine learning, or ML, technology. “Building ML models and products in a responsible and ethical manner is both our core focus and core commitment,” Google Research Vice President Marian Croak said in a January post.
Google is keen to tout its deep AI expertise. ChatGPT triggered a “code red” emergency within Google, according to The New York Times, and drew Google co-founders Larry Page and Sergey Brin back into active work.
Editors’ note: CNET is using an AI engine to create some personal finance explainers that are edited and fact-checked by our editors. For more, see this post.
