Sarah (37), a married woman in South London decided to file for divorce this January after taking advice from the latest chatbot in the news, the ChatGPT. According to news reports on her filing, Sarah was not satisfied with the advice received from her friends as to whether to continue her marriage or not, and she finally turned to the ChatGPT chatbot for the ‘verdict’. “You can decide for yourself about divorce. But the reason for the new relationship should be examined by yourself. Really be honest with yourself” ChatGPT suggested dutifully, as per sources.
And Sarah filed.
Later, in mid-January, the British Newspaper Daily Mail compared relationship advice of its own Agony Aunt with ChatGPT’s advice. Daily Mail’s conclusion?
“While neither our real-life Agony Aunt nor the AI chatbot had simple, immediate solutions to most of the dating dilemmas, they did have very similar answers. All answers carefully reflected on considering individual circumstances, thinking through tough decisions and even consulting with a therapist when things are too tough.”
Daily Mail cautions: “But as we know, chatbots can get it wrong, and have been linked to several problems including biases around gender, race and culture..” (from Dailymail.co.uk)
Amidst the gloomier news on the pandemic, the Ukraine war, tech industry downturn, overhanging global recession, and exploding unemployment, ChatGPT stands as a ray of sunshine.
ChatGPT (short for Chat Generative Pretrained Transformer), an AI model based on the powerful Reinforcement Learning from Human Feedback (RLHF) algorithm, is often and simply called a chatbot, though it mimics more advanced human-like text conversations / responses in real time. The Elon Musk backed ChatGPT was released by OpenAI in November 2022.
Despite being simply labeled as a chatbot, it immediately became clear that ChatGPT (now in its third version) is not just “another chatbot.” The salience of ChatGPT, unlike many other AI models, is its impressive linguistic power that is derived from the incalculable number of words across the internet. According to BBC Science Focus, “It is one of the largest and most powerful language processing AI models to date, with 175 billion parameters.”
More than a million users signed up with ChatGPT in a month’s time. Its processing system was so overloaded, the message “That model is currently overloaded with other requests. You can retry your request..” flashed frequently. Though it can create AI images and videos, ChatGPT’s comprehensive textual responses are incomparable and cannot be mimicked by any other AI interface (presently).
As a result, ChatGPT is now considered as one of the most impressive technological developments to have arrived in the technological landscape. Some view it as the new Google! ChatGPT’s technological breakthrough stunned novices and experts alike. Its impact was such, OpenAI founder Sam Altman reportedly met Microsoft for a $10 billion funding round (from ChatGPT’s total valuation of $29 billion).
ChatGPT has clawed into the minds of many creators and teams due to its incomparable speed in completion of routine tasks. For instance, some software developers say that programs for which they spend hours coding and compiling, are outputted by ChatGPT in a few seconds. In fact, one viral trend in today’s internet and social media are the posts and videos by troubled software developers who feel threatened – and could be pushed to obsolescence – by ChatGPT. So are copywriters and freelancers. Still, what is important to note is that ChatGPT is a threat mostly for entry level positions in these and not for experienced, senior level performers who stand above the basic AI level processing of chatbots. AI YouTube Creator and E-Book Author @IgorPogany, in his “The AI Advantage” Youtube channel, comments: “As we move into the future it is becoming increasingly clear that AI will replace a lot of jobs that we’re used to these days. it should be pointed out that for all these jobs, it’s really the entry-level positions that will be replaced first.”
Importantly, it is still far from the reality that even advanced chatbots such as ChatGPT would replace more relevant, non-routine tasks needed to be performed by humans such as clearly interpreting the intended meanings of bodies of texts by going beyond its literal meaning, nuanced expressions, psycho-analysis of an individual or develop an entire software piece for an institution having first studied the processes taking place within the institution (the “system study”). In that, the chatbots – while being efficient- would work in the automation of processes but they have to rely on the vision created only by its human master.
