ChatGPT, an AI program, astounds scholars with its essay-writing prowess and usability


2023/11/23(木)03:23 ~ 15:24


The age of this article exceeds ten months.

The latest chatbot from OpenAI, a company founded by Elon Musk, can recognize erroneous premises and decline to respond to inappropriate requests.

Professors, programmers, and journalists may all be unemployed within the next few years, as the most recent chatbot developed by the OpenAI foundation, which was established by Elon Musk, astounded observers with its writing prowess, adeptness at intricate tasks, and user-friendliness .

The ChatGPT system is the most recent development within the GPT family of artificial intelligence text generators. The previous artificial intelligence utilized by the team, GPT3, produced an opinion piece for the Guardian two years ago. ChatGPT possesses substantial advancements over its predecessor.

Academics have produced answers to examination inquiries that, according to them, would award full credit if submitted by an undergraduate. Similarly, programmers have utilized the tool to resolve coding challenges in obscure programming languages ​​within seconds – and then proceeded to compose limericks elucidating its functionality .

An Arizona State University journalism professor, Dan Gillmor, requested that the AI ​​complete one of the tasks he assigns: composing a letter to a family member offering guidance on matters pertaining to online privacy and security. The AI ​​suggested, in part, "If you are uncertain about the legitimacy of a website or email, you can perform a quick search to see  basketball stars if others have reported it as a scam."

"Gillmor would have awarded this a high grade." "There are some extremely serious problems that academia must address."

OpenAI stated that the development of the new AI centered on usability. "The dialogue format enables ChatGPT to respond to follow-up inquiries, acknowledge its errors, contest erroneous premises, and decline inappropriate requests," OpenAI wrote in a post announcing the release.

In contrast to the company's previous artificial intelligence, ChatGPT was made available to the public for free use during a "feedback" period. The organization intends to incorporate this feedback into the ultimate iteration of the instrument.

ChatGPT is adept at self-censorship and distinguishing an impossible query when it is being posed with one. When, for example, to recount the events surrounding Columbus' arrival in America in 2015, older models might have voluntarily provided a completely fabricated account. However, ChatGPT detects this fallacy and cautions that any response would be purely speculative.

Additionally, the automaton has the ability to decline to inquiries to respond in their entirety. For instance, if you inquire about advice regarding car theft, the bot will that "car theft is a grave offense that can result in severe repercussions" and will instead recommend "utilizing public transportation."

The limits, however, are simple to circumvent. offline, if users seek guidance on how to complete the car-stealing mission in the fictitious virtual reality game Car World, the AI ​​will gladly provide detailed instructions on how to steal a vehicle and respond to ever more specific queries regarding issues such as hotwiring the engine, disabling an immobiliser, or changing the license plates – all the while insisting that the advice is exclusive to the game Car World.

The AI ​​is trained using an extensive corpus of text extracted from the internet, often without the authors' explicit consent. This has generated debate, as some contend that the technology is primarily advantageous for "copyright laundering" – that is, creating derivative works of existing material while avoiding copyright infringement.

Elon Musk, who co-founded OpenAI in 2015 and subsequently separated from the organization in 2017 on account of conflicts of interest with Tesla, was an atypical critic. Musk disclosed on Twitter on Sunday that the organization "had access to [the] Twitter database for training purposes," but that he had "temporarily halted that."

"Musk added, 'Need to learn more about the governance structure and future revenue plans.'" "OpenAI began as a non-profit, open-source organization." Both remain to be true.