It is not unreasonable to say that the autoregressive language model, also known as GPT-3, is an unprecedented advancement in content creation. In fact, the results it can produce are so advanced that many academics are chomping at the bit for AI text generators. are already able to write as well as a college student (or better), which is leading students to cheat on exams without anyone being able to detect it.
In other sectors, such as business, companies are using GPT-3 to design products and applications that include: automated customer service, reporting, sentiment analysis, knowledge management, and machine translation.
But on Monday, OpenAI changed everything forever. Announced a new model in the GPT-3 family, text-davinci-003, which is supposedly improves on its predecessors by handling more complex instructions and produce longer-form content. This made people rush to try it and they discovered that it is capable of generate songs, poems or rhymes at a level never seen before. This case, for example, manages to raise the eyebrows of more than one: a user told the AI to write a short poem that explained Albert Einstein’s theory of relativity. Another asked him to rewrite the poem in the style of John Keats, and he agreed, dropping lines like “Mass affects the shape of time.”
Ask: “Write a short rhyming poem that explains Einstein’s theory of general relativity in simple but precise terms”
Answer (original in English with rhyme):
If you want to understand Einstein’s thought
It’s not that hard if you give it a shot
General Relativity is the name of the game
Where space and time cannot remain the same
Mass affects the curvature of space
Which affects the flow of time’s race
An object’s motion will be affected
By the distortion that is detected
The closer you are to a large mass
The slower time will seem to pass
The farther away you may be
At Magnet, we have counted as another user has trained the model to write like Rajoy. And he managed to predict what the former president’s third column would be like after the Spain-Japan match. The engine offers a column full of obviousness, like the first two by Rajoy, and of an simple language that precisely imitates the expolitic, ticks included. This same user already explained how he trained the engine with the first column and advancement What would the second one be like?
I have used GPT-3 to generate the following soccer column by M. Rajoy (Spain – Germany). pic.twitter.com/7ISXM2Ofw2
– Juan Alonso (@kokuma) November 26, 2022
Now OpenAI has just introduced ChatGPT, a conversational AI based on the same advanced GPT-3 model. In this case, the new AI is trained to provide detailed answers, and to do it with a more “natural” language. While GPT-3 in its original form simply predicts what text follows a given string of words, ChatGPT attempts to answer user queries in a more humane way.
So human that the bot is capable of imitating the salt shaker of an Andalusian explaining what it is Bolzano’s Theorem in Sevillian. Simply incredible:
As you can see in the examples below, the results are surprisingly fluid and natural. ChatGPT is capable of addressing a wide variety of topics, demonstrating vast improvements over chatbots seen even a few years ago. In the following example, a student asks the bot to write an essay on a historical topic. In 10 seconds she had written an outstanding text:
I guess GPT-3 is old news, but playing with OpenAI’s new chatbot is mindblowing. https://t.co/so1TuXMQB0
We’re witnessing the death of the college essay in real time. Here’s the response to a prompt from one of my 200-level history classes at Amherst
Solid A- work in 10 seconds pic.twitter.com/z1KPxiAc1O
—Corry Wang (@corry_wang) December 1, 2022
Apparently, ChatGPT You can also write some TV series scriptseven combining actors from different sitcoms such as friends either how I Met Your Mother.
So i asked ChatGPT to write dialogues for a romcom starring a few actors from B99, HIMYM, and Friends… pic.twitter.com/tHzxthtbsp
— Sammed Sagare (@sammedsagare_) December 1, 2022
some even GPT-3 chat has been asked to continue the timeless classic When love sets sail, of Camel. You judge:
I have devised the ultimate test for artificial intelligence
I have asked the GPT-3 chat to continue Camela’s timeless classic “When Love Sets Sail”
you judge pic.twitter.com/3BMJcr52Rm
—Antonio Ortiz (@antonello) December 2, 2022
Or have you been asked to explain various scientific conceptssomething that he is especially good at:
I asked ChatGPT a series of technical questions relating to thermodynamics – very impressed by how lucid it’s explanations are. Like a very friendly tutor that doesn’t mind stupid questions pic.twitter.com/VhWmefCbim
—Eric Jang (@ericjang11) December 1, 2022
Or answer questions about current debates, as if paris should ban e-scooters shared:
Should Paris ban shared e-scooters? AI has spoken 😉 #GPT3 pic.twitter.com/eACNkVfJQx
— Prabin Joel Jones (@prabinjoel) December 1, 2022
It is even capable of writing in a few seconds a tutorial blog post on how to plot with Pandas/Matplotlibsection by section, with comments added.
Using OpenAI’s new ChatGPT to write a tutorial blog post on plotting with Pandas/Matplotlib, section-by-section, with conversational feedback. (1/3) pic.twitter.com/RxL61OI3H2
— Riley Goodside (@goodside) December 1, 2022