AI tools have been trending recently, catching the attention of numerous tech firms. Consequently, Google subsidiary DeepMind says it will release a new chatbot called Sparrow.
DeepMind CEO Demis Hassabis said it delayed the launch to ensure it has features not found in ChatGPT.
More importantly, it will release Sparrow for a “private beta” in 2023. That means only a select group of people would test the program.
What are the features of Google AI chatbot vs. ChatGPT?
Photo Credit: venturebeat.com
ChatGPT took the world by storm last year by creating text resembling manmade content.
Even better, it is easy to use because it only requires simple commands. For example, You provide a simple command, such as, “Can you write a paper about bitcoin?”
Then, ChatGPT would produce an original report about this cryptocurrency that passes plagiarism detection.
Scientists recently discovered that experts struggle to distinguish between AI-generated content from manmade ones.
Yet, ChatGPT has numerous flaws. First, it can only write about topics before 2021.
Second, it could not generate text depicting graphic violence, sex, or other harmful content.
Despite that limitation, it made headlines for providing discriminatory comments and creating code for malware.
Third, the program requires highly detailed instructions to provide quality results. On the other hand, DeepMind has been refining Sparrow to have none of these flaws.
Its research paper says reinforcement learning (RL) to train Sparrow AI. It asked volunteers to look at several model answers.
Then, the researchers asked them which answers they liked the most. Moreover, DeepMind reported the AI tool trained on answers with and without online proof.
The research firm also made an initial set of rules for this Google AI chatbot. Consequently, it would not “make threatening statements and insulting comments.”
More importantly, Google could link Sparrow to the internet so that it would provide sources for its answers.
Conclusion
Many people have been worrying about the negative effects of AI chatbots. For example, the University of the Philippines (UP) faculty called for new policies to prevent the misuse of AI.
On the other hand, the recent World Economic Forum celebrated the positive potential of artificial intelligence.
We will soon have Google Sparrow, but we must prepare for more innovations to thrive in the future. Start by following the latest digital trends at Inquirer Tech.