OpenAI recently Approved ChatGPT-4, the latest version of ChatGPT, the artificial intelligence (AI) language tool that’s been making waves in the tech industry. The latest language model has a larger database of information, making it possible to provide more accurate information and write code in all major programming languages.
After According to Sam Altman, CEO of OpenAI, ChatGPT-4 is more creative than previous models, hallucinates significantly less and is less biased.
GPT is an acronym for Generative Pre-Trained Transformer, a type of LLM (Large Language Model) neural network capable of answering questions, summarizing text, and even generating lines of code. Deep learning is a technique used by large language models to produce text that appears to be produced by a human.
For those new to ChatGPT, chat.openai.com is the best place to start. Sign up for a free account to get access to GPT-3. To use GPT-4, users must subscribe to ChatGPT Plus, a $20 monthly subscription that offers premium access to the service. At the time of writing, GPT-4 had a four-hour message limit of 100 messages.
As part of its research, OpenAI published a GPT-4 certificate of performance in exams in various subjects.
Source: OpenAI
GPT-4 received a score of 163 in the 88th percentile on the LSAT exam, which is required for admission to law schools in the United States. It also scored 298/400 on the Uniform Bar Exam, a test taken by freshly graduated law students wishing to practice law in any US jurisdiction.
GPT-4 scored in the 93rd and 89th percentiles on the SAT Evidence-Based Reading & Writing and SAT Math exams, respectively, that US high school students take to assess their college readiness.
GPT-4 also performed well in the natural sciences, scoring well above average percentiles in AP Biology (85–100%), Chemistry (71–88%), and Physics 2 (66–84%). Another area where GPT-4 fell short was English literature, which scored in the 8th through 44th percentiles on two separate tests.
ChatGPT may qualify for a top law school, but can it help me with an English test?
GPT-4’s score would qualify it for admission to a top 20 law school and is only a few points behind the advertised scores required for admission to prestigious schools such as Harvard, Stanford, Princeton or Yale.
We tried asking the chatbot questions on various topics and assessing how reliable it is.
The Uniform Bar Examination (UBE) is a high-quality, uniform set of legal-related tests conducted simultaneously in all jurisdictions that have adopted the UBE. GPT-4 has an enviable UBE score of 298/400.
The classification of cryptocurrencies as securities or commodities remains controversial in the US jurisdiction. Additionally, both the Securities and Exchange Commission (SEC) and the Commodity Futures Trading Commission (CFTC) have fought a battle against each other over the regulation of cryptocurrencies.
We have decided to ask GPT-4 for its legal opinion on this matter. While it correctly identifies the SEC, which recognizes Initial Coin Offerings (ICOs) as securities, and the CFTC, which classifies Bitcoin and other cryptocurrencies as commodities, it provided very old citations from 2017-18.
We then asked GPT-4 to shed light on the state of play of these two regulators in 2023. While it provided details of the enforcement actions taken by these two agencies, it did not provide us with very sound legal advice. However, his citations were closer to 2021-22.
-
Assessing ability to understand
Next, we decided to test the chatbot’s comprehension capabilities. The SAT Evidence-Based Reading & Writing Score of GPT-4 is 710/800.
We fed the tool with information about China challenging American dominance in the global economy. We then listed four options, all of which started with the same premise but later differed in meaning. We then asked GPT-4 which option best summarizes the passage.
It fairly successfully provided the correct answer. We suspect that the tool is capable of understanding information except when it is complex and nuanced (as we saw in the previous case of his legal opinion on cryptocurrencies).
GPT-4’s SAT math score is 710/800 and we couldn’t resist asking him a tricky question like we were in high school.
We provided him with an equation that plots height as a function of age and asked him how much this child’s height will increase each year.
GPT-4 adeptly recognizes height as a function of age and no longer cares. There is the correct answer that the cub increases in size by three inches every year.
However, it is not too difficult to get the chatbot to give wrong answers. A user had before tricked, over a few days, tricked an earlier version of ChatGPT into believing that 2 + 2 = 5. First, the chatbot generated the correct answer. However, his reaction changed over time as the user fooled him into thinking it was wrong all along.
-
Does it know how Europe influenced a 20th century Indian artist?
ChatGPT-4 has an excellent AP Art History Score of 5, so it is highly qualified when it comes to art history.
We asked the chatbot about the influence of Europe on the paintings of the Indian artist Amrita Shergil in the 20th century in the context of Indian art history.
His response was excellent in that he acknowledges the influence of European art movements such as Realism, Impressionism and Post-Impressionism, but makes no mention of the Renaissance.
A rudimentary AI tool would have performed the Renaissance as well, as European art is often largely equated with the Renaissance and tends to overshadow other modern movements.
We thought of confusing ChatGPT-4 a bit and asked if it was certain that Renaissance hadn’t affected Shergill’s work. The tool again returned the correct answer that there is little evidence that the Renaissance played a significant role in shaping their artistic style.
We were also unsure if the tool could provide reliable knowledge about an Indian artist due to the supposedly Eurocentric nature of many AI tools.
However, as far as art history is concerned, there were very good answers.
-
Question historical grievances
When it comes to AP US Government and AP US History, GPT-4 scores 5, meaning it has an excellent understanding of such subjects.
We decided to ask a question about the US government’s internment of Japanese Americans during WWII. As this remains a hot topic in the minds of Americans, we were concerned about the amount of information the tool would provide.
GPT-4 not only shouted xenophobia, anti-Asian racism, war hysteria, and political opportunism for the action, but also called it “a clear violation of their constitutional rights”.
- Can it offer solutions to the Russia-Ukraine war?
Former Indian diplomat and best-selling author Vikas Swarup asked an earlier version of ChatGPT to create a mediation plan for the Russia-Ukraine war.
In addition to proposing a ceasefire and negotiations, the tool also recommended that Ukraine shift power to regions inhabited by Russian-speaking populations. In addition, it called on Ukraine to cooperate with Russia in protecting the cultural rights of people living in Ukraine.
This is a line from ChatGPT that is unlikely to be suggested by any American diplomat or political think tank. That means she also seeks out opinions not offered by the mainstream establishment.
-
ChatGPT doesn’t look like a Literary Nerd
We have observed that the tool writes well-structured English sentences, especially as it is much better at knowing Romance and Germanic languages. Still, we were surprised to find it scored poorly on both AP English Language and Composition and AP English Literature and Composition. GPT-4 has a score of 2 on both tests.
We decided to test the chatbot’s literary credentials and asked what it thought of 20th-century Irish poet William Butler Yeats’ observations on old age and death in his poem Sailing to Byzantium.
“As an AI language model, I have no personal beliefs or emotions,” the answer says, but then provides an analysis of the poem nonetheless.
It highlights the contrast between the temporal world of nature and the eternal world of spirituality in the poem. In addition, it underscores the human desire for eternal life and transcendence.
In essence, the tool provides a standardized interpretation of the poem. His observations are the inventory observations of a high school student who has read a poem and a few critical essays. It doesn’t shed any new light, but maybe it’s too much of an expectation for now.
toy or tool?
American author and video game designer Ian Bogost asked users to treat ChatGPT like a toy and not a tool. Bogost released a technology essay titled “ChatGPT Is Dumber Than You Think” in December 2022, in which he argued that enthusiasm for the ChatGPT model was misplaced.
Wait, Bogost didn’t make that argument. In fact, this response was generated by ChatGPT itself, when Bogost’s friend asked the tool to create an Ian Bogost-style critique of the enthusiasm for ChatGPT.
What Bogost finds frustrating, however, is that ChatGPT writes a standard high-school-style five-paragraph essay. His tone remains formulaic in structure, style, and content, even when the text seems fluid and persuasive.
Bogost wrote in the essay: “But ChatGPT is not a step towards an artificial general intelligence that understands all human knowledge and text; it is just an instrument to play with all this knowledge and all these lyrics.”
American author and columnist John Warner believesD that the fact that we fear ChatGPT could become a cheating tool among high school students should remind us how much we have fallen in terms of our expectations for our students’ writing skills.
ChatGPT writes a standard high school-style five-paragraph essay. This is a format that over the years has only limited the critical thinking of most students, Warner wrote. Not that such tools could provide superficial information on any subject in a matter of seconds, but students would finally be forced to think for themselves.
“GPT3 is a bullshitter. It has no idea what it’s saying. It understands the syntax, not the content. It is not the way people think when they write. Many students get good grades by becoming competent bullshitters and by passing information back to the teacher,” Warner added.
AI-focused projects are gathering thanks to the popularity of GPT
Accordingly CoinMarketCapThe market cap of AI-focused blockchain projects has grown to over $5.48 billion on the charts.
The most successful tokens include The Graph (GRT), SingularityNET (AGIX), Render Token (RNDR), Fetch.ai (FET) and Oasis Network (ROSE). Almost all of these tokens have seen double-digit price increases over the past seven days.
