Because ChatGPT is intelligent, we refer to it as artificial intelligence. And it is no longer necessary to demonstrate its performance. So many IT professionals and scientists have already examined its extraordinary capabilities.
The OpenAI tool can answer history, language, biology, and even mathematics queries and is frequently compared to human intelligence. To ensure an impartial comparison, a specialist was required to evaluate ChatGPT. This was the decision made by Finnish psychologist Eka Roivainen.
Eka Roivainien, a Finnish psychologist, decided to quantify ChatGPT’s intellect to determine if its performance is comparable to that of humans. He employs the third edition of the Wechsler Adult Intelligence Scale (WAIS). This IQ test consists of eleven subtests on verbal and performance scales. The average score possible after completing these 11 subtests is 100.
Unlike humans, ChatGPT did not demonstrate a lack of concentration during its test. Unlike Men, however, the OpenAI tool failed all 11 subtests. Of the six subtests that comprise the verbal scale, he only passed five: vocabulary, similarities, information, arithmetic, comprehension, and number memory. The AI did not scale the latter because it did not make sense.
OpenAi’s language model demonstrated intelligence based on its ability to pass exams. For instance, ChatGPT easily passed the vocabulary test. It has an extensive database and performed exceptionally well with word definitions. The same holds for the stage of general knowledge.
The AI performed flawlessly in numerous fields and had no difficulty discussing science, culture, cinema, athletics, history, etc. Mathematics and general comprehension also performed well. With the similarities subtest, however, ChatGPT provided answers that annoyed the psychologist.
Indeed, the OpenAI tool tends to engage in pointless comparisons and demonstrations when we expect a straightforward, precise response. Instead of stating that Harry Potter and Bugs Bunny are fictitious, the candidate compares them and demonstrates their similarities and differences.
ChatGPT, on the other hand, failed the subtests of the performance scale because it required eyes, hands, and hearing, which he lacked.
Nevertheless, based on the results of previous subtests, the AI has an IQ of 155. A performance well above 99.9% of the individuals whose scores were used as the premise for standardization.
AI is extremely limited and has a long way to go before it matches human intelligence. According to the psychologist, she has difficulty comprehending the universe and reasoning like a human. There are basic logic queries the AI has no clue how to answer and prefers to kick.