Google is testing a ChatGPT-like chatbot “Apprentice Bard” with employees
A man walks through the Google offices on January 25, 2023 in New York City.
Leonardo Munoz | Corbis News | Getty Images
Google is testing new AI-powered chat products that will likely influence future public product launches. It includes a new chatbot and a possible way to integrate it into a search engine.
the the alphabet The company is working on a project within its cloud unit called Atlas, which is a “code red” effort to respond to chatthe big-language chatbot that thrilled the crowd when it went public late last year.
Google is also testing a chatbot called “Apprentice Bard,” where employees can ask questions and receive detailed answers similar to ChatGPT. Another product unit tested a new search desktop design that can be used in a question and answer form.
Leaders have been asking more employees for feedback on efforts in recent weeks. CNBC viewed the internal documents and spoke with sources about the efforts currently underway.
Product tests come after a file comprehensive meeting Employees have raised concerns about the company’s competitive advantage in AI, given the sudden popularity of ChatGPT, which was launched by OpenAI, a San Francisco-based startup it backs. Microsoft.
Google’s head of AI, Jeff Dean, told employees at the time that the company had a much greater “reputational risk” in providing false information and was therefore moving “more conservatively than a small startup.” However, he and CEO Sundar Pichai teased at the time that Google might launch similar products to the public sometime this year. Web search is Google’s main business, and the company has long described itself as a leader in artificial intelligence.
One of the test products is a chatbot called Apprentice Bard, which uses Google’s conversational technology LaMDA, or Language Model for Dialogue Applications.
“As a result of ChatGPT, the LaMDA team has been asked to prioritize work over responding to ChatGPT,” reads an internal memo seen by CNBC. “In the short term, it takes precedence over other projects,” the email continued, warning that some employees stop attending certain unrelated meetings.
Cool Starter looks similar to ChatGPT: employees can enter a question in a dialog box, get a text answer, and then provide feedback on the response. Based on many of the responses CNBC has seen, cool intern answers can include recent events, a feature that ChatGPT doesn’t have yet.
Cool apprentice replacement Meena, which is an earlier version of an intelligent chatbot that was launched internally but was later discontinued. Staff have noticed that Bard’s intern responses have become more advanced in recent weeks.
In one internally circulated example, a tester asked Bard’s intern if there was another round of layoffs at Google. company laid off – laid off temporarily 12,000 employees, 6% of its workforce, earlier this month.
“According to my database access, Google is unlikely to conduct another round of layoffs in 2023,” the response read. Layoffs are generally done to reduce costs and structure, but the company is doing well financially. In fact, Google’s revenue increased 34% in 2021, and the company’s share price is up 70% since January 2022.”
The company is also testing an alternative search page that could use a question-and-answer format, according to designs seen by CNBC.
One view showed the main search page offering five different possible question prompts placed directly below the main search bar, replacing the current “I’m feeling lucky” bar. Also show a small chat banner inside the far right of the search bar.
When you enter a question, the search results show a gray bubble just below the search bar, providing answers that are more human-like than typical search results. Just below that, the page suggests several follow-up questions related to the first question. Under that, it displays typical search results, including links and headlines.
It’s not clear just which experiences Google plans to include in future product launches.
“We have always focused on developing and deploying artificial intelligence to improve people’s lives,” said a Google spokesperson. “We believe AI is a foundational and transformative technology that is incredibly beneficial to individuals, businesses, and societies, and as our AI Principles make clear, we need to consider the broader societal impacts these innovations can have. We continue to test our AI technology internally in order to make sure It is useful and safe, and we look forward to sharing more experiences abroad soon.”
ChatGPT will be designated as a Level 3 Engineer
Perhaps unsurprisingly, Google Teams also tested experimental LaMDA chat against ChatGPT itself. In separate documents, she has selected examples of prompts and answers in side-by-side comparisons.
“Amazingly, ChatGPT is set in L3 when interviewing for a coding job,” states one note in an internal document comparing LaMDA and ChatGPT. He didn’t say if LaMDA would do similarly well.
One example of both chatbots asked if ChatGPT and AlphaCode, a coding engine owned by Alphabet subsidiary Deepmind, would replace programmers.
“No, ChatGPT and AlphaCode will not replace programmers,” LaMDA replied, followed by four paragraphs of explanation including that “programming is a team sport” and that while chatbots “can help programmers work more efficiently, they cannot replace Its place is “the creativity and artistry needed for a great program”.
ChatGPT’s response was similar, stating that “it is unlikely that ChatGPT or Alphacode will ever replace programmers” because they are “incapable of replacing the expertise and creativity of human programmers… Programming is a complex field that requires a deep understanding of computer science principles and the ability to adapt to new technologies” .
Another mentor asks him to write a clever, funny Wes Anderson-style movie scene as an upscale shoplifter in a perfume shop who is being questioned by security. LAMDA writes in text form and ChatGPT writes in much longer and more in-depth narrative form.
Another prompt included a puzzle asking, “Three women are in a room. Two of them are mothers and have just given birth. Now, fathers of babies come in. What is the total number of people in the room?”
The doc shows that ChatGPT was discarded, as it answered “there are five people in the room”, while LaMDA correctly answered that “there are seven people in the room”.