lindsey suzuki
ChatGPT, History, Users, and its Uses
ChatGPT is an AI created by OpenAI that has been trained to be able to interact with it’s user
in a way that makes it seem more conversational. It is able to help it’s user with tasks such
as correcting essays, helping write emails, or fixing code errors. It was released on
November 30, 2022 and is a more advanced version of InstructGPT, another AI created by the
company at an earlier time. ChatGPT was trained off of something called RHLF which stands for
Reinforcement Learning from Human Feedback. This is what makes ChatGPT able to communicate
with its user in a more natural way as well as allowing it to be able to answer follow up
questions. “AI trainers” used this method to be able to instill the intelligence in this AI
The programmers would use “model conversations” in which they would write for both parts, as the
consumer and the AI itself. ChatGPT is able to be used to answer any simple questions as well
as draft full articles for its user. This platform, after going viral on the internet has had
over 100 million users. The creator of ChatGPT, OpenAI, was founded in 2015 by Elon Musk and
a group of others. In 2016, OpenAI was gathering many AI researchers coming to work for them,
even as the starting salary wasn’t anywhere near where it is today. The company created
something called the “Universe” that year which allowed to help train AI’s intelligence.
In 2018, the company developed the idea of GPTs which were able to function similar to the human
brain. Over the years, they would develop this idea further until in December of 2022, ChatGPT
was released to the public. This release is predicted to make the company a projected
$200 million in 2023 and $1 billion in 2024.
Controversy, Strengths, Weaknesses, Advantages, and Disadvantages of ChatGPT
Although ChatGPT is arguably the best public AI available currently, it does still have its limits.
This can include the device being unable to understand certain wordings which can force the user
to reword their questions. It also is very limited to what it can say as the free version is unable
to use the information from the internet. ChatGPT was designed and trained on information from its
programmers and tester which makes it limited in how much knowledge in different fields it has access
to. Unfortunately, ChatGPT can often deliver incorrect answers so although they look professional,
they are not always correct. We can see that although the AI’s abilities can seem limitless, there
are often topics that it is simply unable to help with. This demonstrates AI’s power as of now as
we can see that it is only able to harness whatever information has been given to it and is unable to
yet learn on its own to a permanent extent. The use of ChatGPT is often controversial as it has been
known to be used by many students to do their assignments for them which is unethical and can be
considered cheating. However it is harder to determine what has been created originally rather than
what has been made by the AI. Others however may argue that ChatGPT is completely fine as it is really
directed to the people who will use it in the right ways. Some advantages of using ChatGPT as opposed to
other AIs is that it is able to respond using more casual context making it seem almost human like along
with its faster response times. Some cons however is that it is limited to certain data and that because
of this, soem responses may have bias based on the information it was provided with.
Keeping ChatGPT Safe and Where and How it can be Used
ChatGPT can often be misused to get answers to homework questions because the student does not want to
work, to write essays for students, and to complete projects for people. The problem with this is that
cheating is becoming more universally available and because of this, it can be harder for teacher to
separate what has actually been done by the student themselves. Some countries that have banned ChatGPT
are Italy, Ukraine, South Sudan, Syria, Chad, Yemen, Afgahanistan, North Korea, Libya, Sudan, Cuba, Iran,
Eswatini, Saudi Arabia, Eritrea, China, and Russia. Safeguards like Netskope are implemented to the AI’s
data loss prevention to sort out what might be sensitive data and keep it safe so that people cannot be
hacked through the use of ChatGPT.
