Google issued an urgent warning about ChatGPT – the new search bot that has taken the internet by storm.
ChatGPT is an advanced AI bot trained by OpenAI and interacts in a conversational manner.
It is an online service that answers the questions of people who register.
On OpenAI’s website the company says “We trained a model called ChatGPT that interacts in a conversational fashion.”
“The chat format makes it possible for ChatGPT to answer questions that are related to each other while admitting mistakes, challenging incorrect facts, and rejecting unacceptable requests.”
Google vice president Prabhakar Raghavan has warned tech users that the bot gives completely fictitious answers.
He told for German newspaper Welt Am Sonntag: “This is a kind of artificial intelligence that sometimes talks about things we call hallucinations.”
“The machine’s answers are at once convincing and fictitious.”
He added that: “The large language models behind this technology make it impossible for humans to monitor any inappropriate system behavior.”
“But we want to test it on a large scale so we’re happy with the metrics we use to check the answers with facts.”
“We’re considering how to integrate these options into our search functions, especially for questions that don’t have just one answer.”
ChatGPT can alternate the tone of messages according to McAfee research, messages that can appear to be romantic and persuade someone to provide personal information or money.
McAfee says that with the increasing popularity of AI, especially tools like ChatGPT, the chances of us receiving machine-generated information are high.