After schools, colleges, and workplaces, ChatGPT has now cropped up in the courtrooms as well. The new, super-intelligent chatbot, has recently been in news in India after the Punjab and Haryana High Court, in a first, sought the response of the chatbot while dealing with a bail plea in a murder case. The incident occurred when the bench of Justice Anoop Chitkara was hearing a bail plea in the crime that took place in June 2020 and has earned mixed reactions. While another judge has praised Chitkara, the incident opens up a world of questions about the potential uses and implications of ChatGPT.
But what is ChatGPT?
A conspicuous little pink tap on the official OpenAI website transports one to the fantastical world of ChatCPT, the newest AI craze to have hit the world of internet and communication. As one digs in deeper, telltale prompts ask for your phone number and go on to remind you that your conversations will be read. The company also reminds you to not “divulge personal information” to the chatbot. And that’s it. Your journey with ChatGPT has now begun.
ChatGPT is an artificial-intelligence-driven chatbot developed by OpenAI and launched in November 2022. Built based on OpenAI's GPT-3.5 and GPT-4 families of large language models, Chat GPT has been fine-tuned using both supervised and reinforcement learning techniques. It builds on OpenAI’s previous model ‘InstructGPT’ which could understand instructions and deliver responses. Its successor ChatGPT focuses more on dialogue.
What is it, really?
The tech jargon aside, Chat GPT is basically a chatbot that relies on “natural language” processing to have more convincingly human-like conversations with humans and other chatbots.
The tool can respond to questions and assist human users with mechanical and even cognitive tasks like composing emails, essays, code and even poetry. According to Open AI, using the “dialogue format makes it possible for ChatGPT to answer followup questions, admit its mistakes, challenge incorrect premises, and reject inappropriate requests”.
The chatbot has been designed to answer all questions one could possibly have - from planning the best birthday gift for a teen to explaining quantum computing in lay terms, ChatGPT can be used for a milieu of jobs that were previously only considered the domain of humans. This means that upon being asked the correct questions, ChatGPT can help one draft an article for a newspaper, write a research thesis or an essay on William Shakespeare’s writing techniques, or understand how a fission reactor works.
What is OpenAI and what has Elon Musk got to do with it?
Founded in San Francisco in 2015 by the whos-who of the tech world - Sam Altman, Reid Hoffman, Greg Brockman, Jessica Livingston, Elon Musk, Ilya Sutskever, Peter Thiel and others, Open AI is a self proclaimed “research and deployment” organisation that woks for taking the benefits of the “Artificial general intelligence” to the majority of the masses. Twitter CEO Elon Musk resigned from the board in 2018 citing potential conflict of interest with his other ventures. "I didn’t agree with some of what the OpenAI team wanted to do,” Musk had later said of the exit. “I’m not trying to be anyone’s saviour. I am just trying to help in my own way,” Musk had said.
What are the potential uses of ChatGPT?
As outlined above briefly, ChatGPT can be used for two primary purposes - content creation and programming/coding. Recently, the High Courts of Punjab and Haryana refereed to ChatGPT to assess the global view on bail when the assault is laced with cruelty. The incident occurred when the bench of Justice Anoop Chitkara was hearing a bail plea in a murder case.
Chat GPT can be used to generate essays, write answers to questions in examinations, write essays, resumes, news and research articles, letters and perform coding and programming tasks. It can generate ideas for theme parties, and school science projects. It can be used to write or debug code.
But what about ethics?
Research in Artificial Intelligence has always been rife with the ethical conundrum over its use (read misuse) in the world of humans. Today, when AI-based talking tools are a reality, it is perhaps time to revisit some of the ethical considerations that ChatGPT represents.
According to researchers and internet activists, one of the biggest ethical objections against ChatGPT is the bias in training data. Since ChatGPT is a language model trained on specific data sets, biases in the training data sets are bound to reflect skews, particularly toward the marginalised communities, which will then reflect in the model’s output or responses.
The other major concern is its potential unethical use such as for generating responses to a test or for impersonating someone or spreading misinformation. Its ability to impersonate human speech means it can be misused for malicious intent.
And then, there is the aspect of privacy and how a language model chatbot can be used to get information from users that had not specifically consented to give. Data is gold today and almost every week, we hear of a new data breach in some part of the world, making personal data of thousands of people across the world vulnerable to misuse by private profit-making companies selling products and services. In speaking to ChatGPT, it is possible that users divulge personal information, behaviour patterns, or biases that can fetch top dollar in the international data mining market today.
How will ChatGPT impact the job market?
Apart from the ethical concerns, tools like ChatGPT also pose another threat - to employment and the human job market. While robots taking over the world is an overused sci-fi film trope, the fact is that artificial intelligence has already started making its presence felt in workplaces.
According to a report in Fortune, 42 percent of Americans fear that AI will make harder for them to get jobs and that seven out of 10 believe that AI will replace humans in jobs roles that focus heavily on skills role processing and data entry, coding, media and PR communications. Ironically, it can also perform the principal task of an HR, which is to streamline hiring and execute firing.
However, most believe that for now, ChatGPT can only compete with humans in writing skills and programming. When it comes to critical thinking and making cognitive decisions and analyses, humans still seem to be on top. The question perhaps is, for how long? Additionally, in a country like India where entry-level writing and programming jobs employ millions of people, could the emergence of a more cost-effective, time-efficient copywriter or programmer who doesn’t complain about long work hours or the weak office coffee render this workforce rather undesirable for companies? According to research, the chatbot is great at customer care and tech support and delivers well when quizzed on corporate strategy. Users have nevertheless reported that the model errs on simple math questions sometimes, is prone to generate misinformation and also throws up the occasional incorrect code.
The research is vigorously catching up. Last year, researchers at Italy's University of Naples Federico II started developing a robot that could serve drinks and also engage in an emotional conversation with its guests. Artificial Intelligence told like ChatGPT has immense scope in remote healthcare, connected car dashboards that can be used to connect self-driving cars, and many other potential industries.
When I asked ChatGPT if it would take away human jobs, it took a split second longer to respond. As if it was thinking. “As an artificial intelligence language model, I don't actually take away jobs. Instead, my purpose is to augment and enhance human capabilities, making certain tasks more efficient and freeing up time for humans to focus on other important tasks,” it said. It further gave me a list of all the sectors in which AI is helping humans customer service, healthcare, finance, and more.
Those in support of AI and products like Chat GPT claim that while tools like ChatGPT will definitely make some tasks automated, new job roles will also be created that require human skills and expertise in working alongside AI. ChatGPT assistants, for instance, are already a category of professionals that firms deem necessary. Soon there will be job roles for ChatGPT strategists as well and more.
Nevertheless, studies show a linkage between the introduction of robots and the disappearance of jobs in firms in India, China, Vietnam, Thailand, Malaysia, Indonesia, Turkey, Nigeria and Brazil. In terms of sector, only the Information Technology (IT) sector exhibits a positive relationship between higher employment and adoption of new tech in forms. As per the 2022 paper titled Are Robots stealing jobs? Empirical evidence from 10 developing countries, sectors other than IT have exhibited a negative relationship, meaning higher technology adoption has led to a decrease in employment.
Ultimately, the impact of AI on jobs is complex and depends on many factors, including the specific industry, the type of job, and how organizations choose to implement AI technology. But one thing is sure, AI is here to stay it is the responsibility of programmers, tech entrepreneurs, and politicians to ensure safe and ethical usage of this potentially dystopian technology.