The head of Google Search warned of a problem that could hinder the existence of artificial intelligence (AI)-based chatbots, namely hallucinations
This was explained by Prabhakar Raghavan, SVP Google and Head of Google Search in an interview with German media. According to him, hallucinations are a problem that can arise in AI.
"Artificial intelligence as we speak today can sometimes produce what can be described as hallucinations," said Raghavan, in an interview with the daily Welt am Sonntag.
The hallucination meant by Raghavan is that AI can act as if it is very convincing but give contrived answers. This is what we are trying to prevent from happening.
"(AI) expresses as if very convincing but gives contrived answers," he added.
In the realm of AI, it seems that Google has lost a step compared to OpenAI, a startup, one of which received an injection of funds from Microsoft. Last November OpenAI released ChatGPT, which until now still amazes people because it can provide answers that are similar to human responses.
Then Alphabet, parent of Google, recently released its AI chatbot called Bard. However, the chatbot gave the wrong answer in a promotional video, which caused Google's stock value to plummet and its valuation plunged by $100 billion.
However, Bard is currently still in the user testing phase, and it is not yet known when it will be made available to the public, we were quoted as saying by Reuters, Tuesday (14/2/2023).
"Of course we feel we have to (release it), but we also feel a big responsibility. We obviously don't want to mislead the public," said Raghavan.