Banca de DEFESA: VICTOR PEREIRA GOMES

Uma banca de DEFESA de DOUTORADO foi cadastrada pelo programa.
STUDENT : VICTOR PEREIRA GOMES
DATE: 15/12/2023
TIME: 14:30
LOCAL: https://www.youtube.com/c/PPGFILUFRN
TITLE:

REVISITING THE TURING TEST: analysis and consequences

KEY WORDS:

The Turing Test; The Chinese Room Argument; Turing-intelligence; The Ideal Turing Test; Chat-GPT.

PAGES: 106
BIG AREA: Ciências Humanas
AREA: Filosofia
SUMMARY:

The Turing Test, developed by Alan Turing as “the imitation game” in his paper Computing Machinery and Intelligence (1950), brought to light the discussion about the (im)possibility of thinking and intelligent digital machines. After its publication, the Turing Test has been the subject of countless criticisms and objections. One of the most relevant objections to this test is The Chinese Room Argument, which was developed by the North-American philosopher John Searle in his paper Minds, Brains, and Programs (1980), aiming to refute what he calls Strong Artificial Intelligence, as well as the Turing Test itself. This text intends to revisit the Turing Test, as well as the Chinese Room Argument, and to provide an analysis of the concept of intelligence in the context of the Turing Test, from the following issue: “How does Turing understand the concept of intelligence within his test? Is it human intelligence (called here genuine intelligence), or is it some other type of intelligence (whatever that intelligence may be)?” As research results, it is argued that 1) It is possible to interpret the Turing Test as not being developed to assess whether the digital computer involved possesses genuine (human) intelligence or not, but rather to evaluate if it can be considered intelligent in the sense of what is here called Turing-intelligence; 2) the Turing Test is not necessarily committed to Strong Artificial Intelligence; 3) the Chinese Room Argument does not always challenge the Turing Test; 4) the Turing Test is possible, i.e., achievable in practice, as long as it undergoes some modifications, resulting in a new version of the test, which is here denominated the Ideal Turing Test; 5) Chat-GPT is an example of an AI system capable of passing the Ideal Turing Test and, consquentely, being classified as Turing-intelligent. With such results, as a conclusion of the research and as a response to the previously stated question, the TT can be interpreted in such a way as to support the hypothesis that passing such a test is a sufficient condition for an AI system to possess not genuine intelligence, but rather another type of intelligence, namely Turing-intelligence.

COMMITTEE MEMBERS:
Interno - 1883189 - BRUNO RAFAELO LOPES VAZ
Interno - 1451225 - DANIEL DURANTE PEREIRA ALVES
Externo à Instituição - FRODE BJØRDAL
Externo à Instituição - KÊNIO ANGELO DANTAS FREITAS ESTRELA
Presidente - 1048318 - SAMIR BEZERRA GORSKY