Google is not about Let Microsoft or whoever else win its search crown without a fight. The company announced today that it will launch a chatbot called Bard “in the coming weeks.” The launch appears to be a response to ChatGPT, the blockbuster AI chatbot developed by startup OpenAI with funding from Microsoft.
Google CEO Sundar Pichai wrote in a blog post that Bard, already available to “trusted testers,” aims to put “the breadth of world knowledge” behind a conversational interface. It uses a smaller version of a powerful AI model called LaMDA, first announced by Google in May 2021, based on similar technology to ChatGPT. Google says this will allow it to offer chatbots to more users and gather feedback to help address challenges around the quality and accuracy of chatbot responses.
Both Google and OpenAI build their bots on text-generating software that, while eloquent, is easy to fake and can replicate unsavory speech styles collected online. The need to eliminate these flaws, and the fact that such software cannot be easily updated with new information, challenges hopes of building powerful and profitable new products based on the technology, including suggestions that chatbots could reshape web search.
Notably, Pichai did not announce plans to integrate Bard into Google’s profitable search box. Instead, he demonstrates a novel and careful approach to using underlying artificial intelligence techniques to augment traditional search. For questions that do not have a consistent answer, Google will synthesize responses that reflect dissenting opinions.
For example, the query “Is it better to learn the piano or to learn the guitar?” would come across “Some say the piano is easier to learn because of the more natural finger and hand movements… Others say it’s easier to learn chords on the guitar.” Pichai also stated , Google plans to make the underlying technology available to developers via an API, just as OpenAI did with ChatGPT, but no timetable was provided.
The heady excitement sparked by ChatGPT has led to speculation that Google is facing a serious challenge to its web search dominance for the first time in years. Microsoft, which recently invested about $10 billion in OpenAI, will hold a media event tomorrow to introduce its collaboration with the creator of ChatGPT, which is believed to be related to a new feature on the company’s second-ranked search engine, Bing. Sam Altman, CEO of OpenAI sent a photo His conversation with Microsoft CEO Satya Nadella shortly after Google’s announcement.
Quietly launched by OpenAI last November, ChatGPT has become an internet sensation. Its ability to answer complex questions with apparent coherence and clarity has many users dreaming of a revolution in education, business and everyday life. But some AI experts advise caution, noting that the tool doesn’t understand the information it’s given and is inherently prone to making things up.
For some of Google’s AI experts, the situation could be particularly vexing, since the company’s researchers developed some of the technology behind ChatGPT — something Pichai alluded to in a Google blog post. “Six years ago, we repositioned the company around AI,” Pichai wrote. “Since then, we have continued to invest in AI across the board.” He named Google’s AI research arm and the work of DeepMind, the British AI startup Google acquired in 2014.
ChatGPT is built on top of GPT, an AI model called a Translator, originally invented by Google, that takes a string of text and predicts what will happen next. OpenAI rose to prominence for publicly demonstrating how feeding massive amounts of data into Transformer models and boosting the computer power to run them yielded systems that excel at generating speech or images. ChatGPT improves on GPT by having humans provide feedback to different answers output by another AI model fine-tuned.
By its own admission, Google has chosen to tread carefully when adding the technology behind LaMDA to its products. In addition to creating the illusion of misinformation, AI models trained on text scraped from the web are prone to exhibit racial and gender bias and recurring hate language.
Those limitations were highlighted by Google researchers in a 2020 draft research paper advocating caution with text-generation technology, angering some executives and leading the company to fire two prominent ethical AI researchers Timnit Gebru and Margaret Mitchell.
Other Google researchers working on the technology behind LaMDA were frustrated by Google’s indecision and left the company to build startups that leverage the same technology. The advent of ChatGPT seems to have inspired the company to speed up its timeline for pushing text generation capabilities into its products.
