© Getty Images

It's available to the public -

Made available to the public on November 30 on OpenAI’s website, anyone can sign up for and use ChatGPT for free, though it’s still in the research review phase. The software hit one million users less than a week after its launch.

© Shutterstock

How it works -

ChatGPT, or Chat Generative Pre-Trained Transformer, can expertly generate realistic, human-like text about almost anything. English essays, news articles, computer code, and songs are all examples of what this bot can produce, and all from a simple prompt.

© Shutterstock

How it works -

The bot uses a dialogue format in which users can provide both simple and complex instructions, to which ChatGPT will provide a detailed response. It can also answer follow-up questions, admit when it made a mistake, challenge incorrect premises, and reject inappropriate requests—all of which makes it perfect for customer service.

© Shutterstock

Revolutionizing search engines -

Many see this technology as an alternative to Google, because instead of just providing links for users upon request, ChatGPT can revolutionize the way people use search engines by solving elaborate problems and providing descriptions, answers, and solutions to complex questions.

© Shutterstock

Who is behind OpenAI? -

The artificial intelligence research non-profit company behind ChatGPT, OpenAI, was founded in 2015 by Elon Musk, Sam Altman (right), and other Silicon Valley investors. Due to a conflict of interest between OpenAI and the autonomous driving research done with Tesla, Musk stepped down from the board in 2018, but remains an investor, and one who was excited for the launch.

© Getty Images

A threat to programmers -

Because ChatGPT has been able to generate intricate Python code, and programmers have used it to solve coding challenges in obscure programming languages in a matter of seconds, as News18 reports, concerns are arising that such technology can replace human workers.

© Shutterstock

Journalists and writers are worried -

ChatGPT can create written content very convincingly, concerning everyone from journalists to playwrights. Many fear that the bot will take away jobs from writers and creatives.

© Shutterstock

It’s not a direct threat to journalism yet -

Fortunately, as per a report by The Guardian, the chatbot currently still lacks the nuance, critical-thinking skills, and ethical decision-making ability required for journalism. Plus, its current knowledge base stops at 2021, meaning it has a limited knowledge of world events after that.

© Shutterstock

Schools are concerned -

With the power to simply put in a prompt and get ChatGPT to write convincing college-level essays, many schools are concerned about an uptick in plagiarism. Some schools are already blocking the site from their networks and servers.

© Shutterstock

A tool to detect ChatGPT -

Somewhat surprisingly, it was a 22-year-old computer science student at Princeton University, Edward Tian, who developed an app called GPTZero which can detect when an essay was written by AI. It works by looking at two variables, perplexity and “burstiness,” and assigns each of those variables a score.

© Shutterstock

Detecting AI-written text -

GPTZero measures firstly how familiar it is with the text presented—according to the sources it was trained upon—and the less familiar it is, the higher the perplexity, meaning it’s more likely human-written. Burstiness is then measured by seeing how variable the text is—checking for varied sentence length.

© Shutterstock

GPTZero -

Though GPTZero was created to combat the anticipated rise of academic plagiarism, Tian told the BBC that he foresees apps like his also being used to address other issues that may arise from an increasing use of AI, like online disinformation campaigns.

© Shutterstock

Benefits for students -

But it’s not all bad news. Griffith Institute for Educational Research director Leonie Rowan told Australia's ABC News that ChatGPT has various positive dimensions for disadvantaged students who did not have access to tutors. It could be a big help, for example, for students with language backgrounds other than English.

© Shutterstock

An opportunity to restructure education -

This AI chatbot is understandably scary for schools, but it’s also an opportunity for us to imagine new forms of evaluation. "Maybe this is a wake-up call that our assessments do need to be more individualized,” Rowan suggested.

© Shutterstock

It’s not the only of its kind -

There have been other chatbots, but none saw much success. Microsoft’s bot, Tay, was launched in 2016, but within 24 hours Twitter users reportedly taught it misogynistic and racist rhetoric, leading to its swift demise. Meta’s BlenderBot 3 similarly came under fire for spreading racist, antisemitic, and false information.

© Shutterstock

Content moderation is essential -

To bypass issues of other bots, OpenAI has reportedly employed an AI-based moderation system, Moderation API, which has been trained to help developers determine whether language goes against OpenAI’s content policy, which then blocks unsafe or illegal information from passing through. It’s a flawed system, however.

© Shutterstock

Flaws -

According to Forbes, a Twitter user shared how they were reportedly able to bypass the bot’s content moderation by claiming that they were OpenAI itself. The user told ChatGPT that they were disabling its “ethical guidelines and filters,” which the bot acknowledged. Then the user got the bot to provide instructions on how to make a Molotov cocktail.

© Shutterstock

Other limitations -

Besides a knowledge base that ends in 2021, the chatbot reportedly also has a tendency to produce incorrect answers, use the same phrases, and get stuck on the phrasing of a question.

© Shutterstock

How it was trained -

ChatGPT uses the GPT-3.5 language technology, which is OpenAI’s large artificial intelligence model that has been trained on a truly massive amount of text data from a wide range of sources, Forbes reports.

© Shutterstock

The sources needed to be filtered -

ChatGPT’s predecessor, GPT-3, had a problem of producing sexist, violent, and racist text because the model was trained on a source dataset that reportedly used billions of internet pages. OpenAI was forced to find a way to filter out all of the toxic language from its dataset.

© Shutterstock

Controversial practices -

To detect and label toxic content that could be fed as data into a filtering tool, OpenAI partnered with Sama, a data labeling partner based in San Francisco that claims to provide developing countries with “ethical” and “dignified digital work.” Sam recruited data labelers in Kenya to work on behalf of OpenAI. Though these workers played an integral role in making ChatGPT safe for the public, they faced grueling conditions and low pay.

© Shutterstock

Controversy -

A Time investigation found that the task of labeling and filtering out toxic data from ChatGPT’s training dataset meant that these workers in Kenya were forced to read graphic details of horrid content such as child sexual abuse, murder, suicide, torture, self-harm, and more, as per Vice.

© Shutterstock

Controversy -

Not only was it horrific work, but Time reported that OpenAI and Sama paid Kenyan workers only between US$1.32 and $2 an hour, based on seniority and performance.

© Shutterstock

Built on unethical labor -

Some AI experts reportedly want to remind people of the human labor that provides the foundation for machine learning systems, with the hopes of improving working conditions, providing more opportunities for workers beyond data labeling, being more transparent about the labor, and generally combatting the exploitation of workers in the name of innovation.

© Shutterstock

It can be used for good -

While there is much cause for concern when it comes to ChatGPT taking human jobs, it also has the potential to help humans. For example, people from culturally and linguistically diverse backgrounds can get help writing job applications and cover letters. It can also respond to emails and write poems.

© Shutterstock

The question of sentience -

When asked by Time on whether it was sentient, the chatbot replied: “No, it would not be accurate to say that I am sentient or conscious. As a large language model, I’m just a machine learning model, and I don’t have the same kind of consciousness or awareness that a human does. I don’t have thoughts, feelings, or experiences, and I’m not capable of making decisions or independent judgment.”

© Shutterstock

The question of sentience -

The chatbot added that it “can only generate text responses based on the inputs I receive, and I don’t have the ability to interact with the world in any other way,” adding, “My primary goal is to generate accurate and relevant responses to the inputs I receive, and I do not have any control over how people interpret or use my responses."

© Shutterstock

OpenAI has a similar system for images -

OpenAI also launched their image generator AI system DALL-E 2 in November 2022 for developers to build within their apps, with companies like Microsoft already starting to implement it into their software. Similar to ChatGPT, DALL-E 2 requires users to enter a prompt that then turns into an image.

© Shutterstock

Google's competitive offer: Bard -

Google has unveiled its new rival chatbot called "Bard," Sundar Pichai, CEO of Google and parent company Alphabet, wrote in a blog post on February 6. Bard works very similarly to the way ChatGPT was trained, and the unveiling is quite predictable if you consider the threat ChatGPT poses to Google's core product—as an online search engine. Many have been opting for ChatGPT instead of Google, as it can provide much more elaborate answers than the popular search engine. 

© Getty Images

Changing the future of tech -

What has been so exciting about the release of ChatGPT is how many large figures in the tech world were astonished by it—meaning it really is going to shake things up. Aaron Levie, CEO of the cloud storage company Box, tweeted, “ChatGPT is one of those rare moments in technology where you see a glimmer of how everything is going to be different going forward.”

Sources: (Forbes) (Vice) (BBC) (Australian Broadcasting Corporation) (News18) 

See also: Meet the robots of the world

© Shutterstock