ChatGPT's Trial by Fire

Let’s deal with the sexism charge first. Sharmin Ali is founder of a deep-tech startup called Instoried. She complained to journalists Arpika Bhosale and Aastha Atray Banan in Sunday Mid-day: “At the core of any open AI is a man – not a woman – who inputs ‘variables’

- Minhaz Merchant

IS CHATGPT SEXIST? Does the AI chatbot pose a danger to society? How accurate are ChatGPT’s answers to complex questions?

They tell the AI what words in a sentence are positive or negative, such as happy, sad, angry or joyful. Now add AI’s auto-learning into the mix, and the machine is assigning its own values to new words and framing its responses to users accordingly. These then include the inherent bias of the man, who assigned the first variable, a value which is now multiplied on an unimaginable scale.”

Turn now to the second concern: does ChatGPT pose a danger to human society? Elon Musk, founder of Tesla and SpaceX and reluctant owner of Twitter, thinks it does. Along with a bunch of AI experts, Musk has called for a six-month moratorium on developing more advanced AI systems like GPT-4. The Future of Life Institute issued a letter last month signed by Musk and over 1,000 other experts. The letter warned: “Powerful AI systems should be developed only once we are confident that their effects will be positive and their risks will be manageable.”

Is that paranoid? Many AI specialists think it is. ChatGPT has built in safeguards to block pornography, homophobia, racism and, yes, sexism. ChatGPT comprises more than 100 trillion parameters and is constantly improving, upgrading and updating data.

The latest avatar of ChatGPT was unveiled recently. Bloomberg called it Chat GPT’s “iPhone Moment” which poses a “threat to Google’s sprawling tech empire.”