According to reports from The New York Times, concerns have been raised within Google over the recent introduction of OpenAI’s AI chatbot ChatGPT. The Times now reports that Google intends to “show a version of its search engine with chatbot characteristics this year” and to introduce more than 20 AI-powered projects.
Even as recently as December, we learned that Google executives were concerned that, while making significant investments in AI technology, implementing it too quickly may damage the company’s brand. However, things are evolving swiftly. Google said earlier this morning that it will be cutting more than 12,000 jobs and prioritising AI as a field of particular importance.
No specific date has been given for the release of Google’s AI search demo, but other initiatives from the slide show that the Times was able to view appear to be scheduled to go live during its annual I/O event in May, which has previously seen the introduction of features like Duplex and Google Glass.
The situation is so critical that the Times also claims that Google founders Larry Page and Sergey Brin, who continue to hold the majority of shares in Alphabet, the parent company of Google, have “offered advice” to business executives, approved strategies, and presented ideas during a meeting to discuss ChatGPT last month. It states that Sundar Pichai, CEO of Google/Alphabet, extended an offer for them to interact at this level after leaving their day-to-day positions in 2019.
Google will focus “getting facts right, guaranteeing safety, and getting rid of disinformation,” according to the chatbot search demo, in an effort to address the problem of AI replying to requests confidently and clearly with inaccurate information. It is also examining ways to expedite review procedures that are meant to examine the technology to see if it is operating in a fair and ethical manner.
A panel of executives, including Jeff Dean, who oversees the company’s research and AI division, presented the report’s new product launches, which include an image generation studio that “creates and edits images,” an app for testing product prototypes, and a set of tools called MakerSuite that other companies can use to develop AI prototypes directly from a browser window. The company is also developing two more tools: Colab + Android Studio, which aids in building mobile apps, and PaLM-Coder 2, which is a code generating tool akin to Microsoft’s GitHub Copilot product.
When it comes to the introduction of fresh AI products, Google has tread gingerly in recent years. After firing two well-known researchers in the field, Timnit Gebru and Margaret Mitchell, the business found itself at the heart of a discussion about the morality of artificial intelligence. The duo outlined objections to AI language models, pointing out issues like their inclination to exaggerate biases in their training data and portray misleading information as fact.
Even though Google is reputed to have the most sophisticated AI research, it has only tested software that has extremely severe constraints. For instance, the company’s AI Test Kitchen app gives users access to text and image creation tools that are comparable to OpenAI’s DALL-E and ChatGPT. Google, however, severely limits what users can ask of these programmes. The company has already demonstrated some of its own chat-intensive AI technologies, including a private demo of a system resembling ChatGPT in 2021.
But with the introduction of OpenAI’s ChatGPT and ominous predictions about Google’s impending death, it appears the firm is changing its strategy. Google has previously claimed that it has refrained from releasing specific AI products due to the risk of “reputational damage.” It appears that the reputation it wishes to avoid right now is that of being obsolete.
