Since Open AI's popular chatbot ChatGPT was launched in November, the tool has inspired a number of unofficial experiments.
This may feel like a throwback to older millennials who used IRC back in the day, only they chatted with presumed humans. Chat GPT is actually an AI robot that is not sentient being and that has caused some eyebrows to be raised.
Despite of people growing up to films showing AI rising above and taking over the world, experts do not believe that a full AI takeover is inevitable.
Shakespeare can be written by an unlimited ways beyond Shakespeare himself, according to Matthew Sag, an Emory University law professor who studies the legal ramifications of training and exploiting the massive language models like ChatGPT.
He shared that while big language models can be incredibly helpful, a very fine line between them and humans still exists.
Massive amounts of data and fancy computational methods enable these AI chat bots like GPT to predict how to put together words in a proficient way. They have been programmed to have a broad vocabulary and ability to comprehend words in context. This enables them to convey extensive knowledge while emulating certain language patterns.
Other major information tech companies, including Google and Meta, have started working on their own impressive language model tools that rely on refined response-generating software that will react to what is inputted into it. In a surprising move, Open AI also developed a user interface that is open to the general public to try it out
Modern attempts to launch chat bots for real-world support have been problematic because they produced peculiar results. In January 2023, the mental health firm Koko faced critic after its founder announced that they had conducted an experiment using GPT-3 to respond to customers. This was not previously revealed to the customers who felt betrayed by this news.
Koko's cofounder Rob Morris, hastened to clear up on Twitter that although users weren't conversing directly with a bot as expected, AI was being used to craft responses.
The contentious DoNotPay business's creator also added that in the future that an AI created "lawyer" will give defendants legit legal advice in courtrooms for traffic cases. The GPT-3 backed chatbot insists that it can be used to assist with resolving customer service issues.
Some researchers are being more cautious with their AI tools in real life use. The efficacy of technology in the judicial field is studied by Daniel Linna Jr., a professor at Northwestern University. He disclosed to Business Insider that he is working on an experimental chatbot named "Rentervention" that will help tenants.
The "Rentervention" bot utilizes Google Dialogueflow technology, which is a similar language modeling tool. Linna confirmed that they are testing out ChatGPT to see if it may help "Rentervention" write better responses than what they are currently using.
Linna believes that ChatGPT has a lot of potential to change the face of customer service, but also does not expect it to be 100% perfect fit for their needs.
This is inline with what Open AI, which also says that their product can produce varied solutions.
