It seems that the issuance of Google as a competing product for “ChatGPT” is only a matter of time, and the American company possesses a wide range of very advanced algorithms in the field of artificial intelligence in various fields, but its “commitment to the ethics of artificial intelligence “The launch of the service to the public has been delayed until now, as confirmed by an informed source.
It is known that Google has the technical knowledge and ready-made artificial intelligence algorithms that go into the work of many applications, from analyzing images and extracting information from them, to “Large Language Models”.
Google launched the “LaMDA” project in the year 2022, which is an artificial intelligence algorithm created by the scientists of the “DeepMind” team of the American company, which is characterized by its ability to conduct conversations and answer questions with very high accuracy that outperforms humans in many ways. of domains.
According to the source, Google does not want to launch a competing product for the ChatGPT program until it is confirmed that it can give it a correct reference for any answer provided to the user, which is what the ChatGPT algorithm lacks today.
The source also confirmed that Google is testing new chat products supported by artificial intelligence, which are likely to be preliminary preparations for launching a product for public use in the future, including a new chatbot and a possible way to integrate it into the famous search engine.
The American company is working on several projects in the field of artificial intelligence, which comes as a result of declaring a state of emergency and activating the “red code” to respond to the intrusion of the ChatGPT program, which is an automated chat program that uses artificial intelligence, which impressed the public when it was launched at the end of the year. The past was developed by OpenAI, a Microsoft-backed startup based in San Francisco.
Leaders of the tech giant asked employees for feedback on efforts in recent weeks from their colleagues, as they made these products available for internal trials.
Google’s decision to test its AI products comes after a recent meeting with company leaders in which company employees expressed concerns about its competitiveness in the field of artificial intelligence, given the sudden popularity of ChatGPT.
Google’s head of artificial intelligence, Jeff Dean, told employees at the time that the company had “a reputation as a big company that makes it move more conservatively than a small startup, especially when it comes to providing misinformation through its AI products.”
However, Dean and CEO Sandra Pichai said at that meeting that Google may release similar products to the public sometime this year.
The web search engine is one of Google’s most important products, and the company has long described itself as a leader in artificial intelligence.
Google is testing a chatbot called ApprenticeBard, where employees can ask questions and receive detailed answers similar to ChatGPT.
Another unit within the company is testing a new design for a new search application that can be used in a question-and-answer format, as in ChatGPT.
The chatbot “Apprentice Bard” uses Google’s “lambda” conversational technology and the language model for dialogue applications.
It is noteworthy that, as a result of the emergence of ChatGPT, the “Lambda” team was asked to prioritize work over responding to it.
ApprenticeBard looks similar to ChatGPT, where employees can enter a question into a dialog box, get a text answer, and then provide feedback on the response.
Apprentice cool answers are expected to include recent events, a feature ChatGPT doesn’t have yet.
ApprenticeBard replaced Meena, an earlier version of the intelligent chatbot that was launched internally but later discontinued.
Employees have noticed Apprentice Bard’s responses have become more advanced in recent weeks.
Artificial intelligence reassures Google employees
In one internally circulated example, an ApprenticeBard tester asked if there was another round of layoffs at Google after the company earlier this February laid off 12,000 employees, which is equivalent to 6% of its workforce.
“According to my database, Google is unlikely to conduct another round of layoffs in 2023,” ApprenticeBard’s response said.
He added, “Layoffs are generally done to reduce costs, but the company is doing well financially. In fact, Google’s profits increased by 34% in 2021, and the company’s share price increased by 70% since January 2022.”
The company is also testing an alternative search page that could use a question-and-answer format.
When you enter a question, the search results show a gray bubble just below the search bar that provides answers that are more human-like than the typical search results. Just below that, the page suggests several follow-up questions related to the first question, after which the sample search results are displayed, including links and headlines.
ChatGPT is a junior programmer at Google
Teams of Google employees conversationally tested the LaMDA experimental bot against ChatGPT, and the teams compared each other’s answers.
An internal document comparing LaMDA and ChatGPT stated that “it’s amazing how ChatGPT can be hired as a junior programmer when interviewing for a programmer job,” and did not say whether lambda would perform similarly well.
In one comparison, an employee asked chatbots if ChatGPT and Alphacode—a scripting engine developed by Alphabet-owned DeepMind—would replace programmers.
LaMDA’s answer was “No, GBT and alpha code will not replace programmers”, followed by an elaboration of his point stating that “programming is a business” and that while chatbots “can help programmers work more efficiently” they ” It cannot replace the creativity and artistry needed to develop a great program.”
ChatGPT’s response was similar, stating that “it is unlikely that ChatGPT or alpha code will ever replace programmers” because they are “unable to replace the experience and creativity of human programmers. Programming is a complex field that requires deep understanding.” Computer science and the ability to adapt to new technologies.
Another comparison involved a puzzle in which an employee, ChatGBT and LaMDA, asks 3 women in a room, two of whom are mothers and have just given birth. Now, the fathers of children come in, what is the total number of people in the room?
The result shows that ChatGBT failed to come up with the correct answer, as it answered “There are 5 people in the room”, while Lambda answered correctly that “There are 7 people in the room”.