What is ChatGTP?
ChatGPT (Chat Generative Pre-trained Transformer) is a chatbot launched by OpenAI in November 2022. It is built on top of OpenAI’s GPT-3 family of large language models and is fine-tuned (an approach to transfer learning[2]) with both supervised and reinforcement learning techniques.
ChatGPT was launched as a prototype on November 30, 2022, and quickly garnered attention for its detailed responses and articulate answers across many domains of knowledge.
These days ChatGPT is a viral sensation that is taking the internet by storm.
From writing essays to composing poetry and limericks, artificial intelligence is the most accessible tool of its kind to date – and anyone can start up a conversation with the chatbot.
But while its popularity is soaring amid reports that OpenAI, the company behind ChatGPT, is closing in on a $10bn investment from Microsoft, there are fears the technology could cause more harm than good.
Human or AI?
“To determine if something was written by a human or an AI, you can look for the absence of personal experiences or emotions, check for inconsistency in writing style, and watch for the use of filler words or repetitive phrases. These may be signs that the text was generated by an AI.”
That’s what ChatGPT told an AP reporter when asked how to tell the difference.
OpenAI said in a human-written statement this week that it plans to work with educators as it learns from how people are experimenting with ChatGPT in the real world.
“We don’t want ChatGPT to be used for misleading purposes in schools or anywhere else, so we’re already developing mitigations to help anyone identify text generated by that system,” the company said.
Does this threaten Google?
There’s been some speculation that ChatGPT could upend the internet search business now dominated by Google, but the tech giant has been working on similar technology for years — it’s just more cautious about releasing it in the wild.
It was Google that helped jumpstart the trend for ever-bigger, ever-smarter AI language models that could be “pre-trained” on a wide body of writings. In 2018 the company introduced a system known as BERT that uses a “transformer” technique that compares words across a sentence to predict meaning and context. Some of those advances are now baked into Google searches.
But there’s no question that successive iterations of GPT — which stands for Generative Pre-trained Transformer — are having an impact. Microsoft has invested at least $1 billion in OpenAI and has an exclusive license to use GPT-3.
Why are Schools blocking it?
Ask the new artificial intelligence tool ChatGPT to write an essay about the cause of the American Civil War and you can watch it churn out a persuasive term paper in a matter of seconds.
That’s one reason why New York City school officials a few days back started blocking the impressive but controversial writing tool that can generate paragraphs of human-like text.
Schools in New York City have already banned ChatGPT amid fears pupils could use the technology to do their homework, while the way that the AI learns means that it has absorbed harmful material which exists on the web.
The decision by the largest U.S. school district to restrict the ChatGPT website on school devices and networks could have ripple effects on other schools, and teachers are scrambling to figure out how to prevent cheating. The creators of ChatGPT say they’re also looking for ways to detect misuse.
What are ChatGPT’s Limitations?
1). Wrong Answers
ChatGPT is a large language model that is continuously being trained to increase the accuracy of responses. However, since it is a brand-new tech, the model has not yet undergone sufficient training. Therefore, the AI chatbot may provide wrong answers. Due to this, StackOverflow has banned ChatGPT, saying: “Overall, because the average rate of getting correct answers from ChatGPT is too low, the posting of answers created by ChatGPT is substantially harmful to our site and to users who are asking or looking for correct answers.”
2). Limitations in Training Data and Bias Issues
Like many AI models, ChatGPT has limitations in its training data. Both the constraints in training data and bias in the data can create a negative impact on the model’s output. In fact, ChatGPT has shown bias when it comes to training minority data groups. Therefore, it’s important to improve the data transparency of the models to reduce bias in this technology.
3). Sustainability
On Twitter, there is a conversation thread regarding how many Graphics Processing Units (GPUs) are required to run ChatGPT. The takeaway here is that running ChatGPT is highly expensive. Given that ChatGPT is a free product, a lot of questions are currently being raised on how sustainable the technology is in the long run.
