Current，Chatgpt can be described as good results，But far from reaching the level close to human mind。Look roughly，Artificial intelligence technology can be said to be a technology that humans use mathematics to simulate themselves and other intelligent technology，Originally the machine intelligence established using a regular mathematical model (such as an expert system)，Then use statistical probability data to process machine learning and classification，The next step is to try to learn with/unsupervised learning、Sample pre -training、fine -tuning alignment、Human -machine calibration and other methods to realize context perception intelligent system。The development trend of these three types of artificial intelligence technology，Continuing from human to machine to human -computer、Research path of human -machine environmental system。From the perspective of the intelligent perspective of human -machine environmental system，The explicit description part of ChatGPT is still in the shallow stage of human interaction between humans and big data，The part of its silent part is not reflected。Dark knowledge、analogy、Metaphor and other seemingly not rigorous、Lack of mathematical bet365 best casino games logic is hidden part，It is difficult to get through rational thinking，But these things are exactly an important part of human intelligence。In the field of intelligent research，Analysis and understanding of basic concepts is very important，The analysis of ChatGPT should also be like this。From the data below、Inference、Indications、In terms of consciousness。

From the perspective of data，ChatGPT does not have the essential characteristics of intelligence。Small samples Small data to solve big problems，is the expression of intelligence。In many scenes，The intention of both parties in the interaction is often in the uncertain non -complete dynamic small data，Small probability appears and gradually evolved。Human intelligence can use these small data，From different dimensions、Different angles and different particle size guess the opponent's intention，To realize confidante (see the signs of the signs)、Fun Shi (seizing the opportunity)、Change (random strain)。This is completely different from the big data that the machine intelligence is good at repeatable、Extraction of verified laws。Human intelligence is also good at using strange data other than statistics probability，and can fully extract the possible intentional intention from valuable small data，Especially can break the conventional，Realize the facts or anti -facts of the cross -domain connection、Guess of value or counter -value。ChatGPT is a program mode of big data+machine learning+fine -tuning transformation+human bet365 Play online games machine alignment，Action based on the factualization of data，But the value of action formed by generalization is completely unknown。Sometimes the result of this generalized behavior is error or even dangerous，For example, various "nonsense" phenomena appearing in the conversation。

From the perspective of reasoning logic，ChatGPT does not have the essential characteristics of intelligence。Taking intelligence as calculation or logic is a misunderstanding。In fact，Real intelligence includes not only the rational logic part，also includes non -/ultra -logic perceptual parts。Mathematics tools as an artificial intelligence basis，It's just part of the logic system based on axiom。The core of ChatGPT is calculating intelligence、Data Smart，Its perception、Cognitive features are on -demand matching combinations of pre -training texts，There are problems such as the risk liability of the source of knowledge and the result of content generation results。Although the ethical and moral threshold constraints are set in the ChatGPT algorithm，But the misleading hazards it may bring still cannot be underestimated (especially the interpretation of the multi -source causality of unknown knowledge、Non -causal correlation description)。and，The "autonomy" of the ChatGPT system is very different from the "autonomy" of human beings。Chatgpt's "autonomy" is the rule or statistical reasoning process of big data in the text symbol space and space，It is the individual order process bet365 best casino games based on the mathematical computing algorithm; and the "autonomy" of human beings is the cause and effect derivation or inference process of small data or unruly data in physical/cognitive/information (symbols)/social hybrid time and space，This is a massive process。Further,，Chatgpt calculation is a causal restoration theory，The knowledge is equivalent to explicit facts。But the key to intelligence is not only computing power，More lies in the calculation ability with reflection。The calculation is a common system theory，The knowledge is the hidden value of the equivalent。Calculation is stronger than calculations than anti -facts、Anti -value ability。For example，Human "autonomy" often contains reflection (fact feedback+value feedback) ability。

From the perspective of referrals，ChatGPT does not have the essential characteristics of intelligence。ChatGPT and other generated artificial intelligence，Different from most artificial intelligence in the past, it can only analyze existing data，can create a new content (such as text、Picture、Video、Music, etc.)。But compared with human intelligence，The limitations of ChatGPT include: limited common sense and causal reasoning (biased knowledge rather than intelligence)、Limited natural language and logical reasoning、Lack of foundation in the real world (no visual input or physical interaction)、The performance is unreliable and unpredictable。One of the important disadvantages，It is not possible to realize the "referral"bet365 live casino games 。Wittgenstein proposed in the "Research on Philosophy" that the "indication" question outside the logic，That is, "Unable to say" and "Keep silence"。Wittergenstein is aware of the lack of human intelligence，But there is no proper proposal solution。His student Turing thought of a way: if the human rational logic and sensuality are referred to，Then you can simulate human intelligence through the formal system of mathematics。This is here，In the rules、Logical fields (such as Go play、shallow treatment of text)，Artificial intelligence and machines can reach or surpass human levels。Of course，This simulation will lose a lot of things (such as perceptual、Intuition, etc.)。"Indication" is not only a problem of natural language and mathematical language，It involves thinking (such as intuition、Cognition) problems other than "language" and groups。

From the perspective of consciousness，ChatGPT does not have the essential characteristics of intelligence。​​ChatGPT on the basis of marking of words，Realized the context of the chapter、Many logical marks，and the common marking of human -machine。Its core is to rely on the selfFormer conversion model of input and output for calculating the self -attention mechanism，Realized it faster、stronger calculation。But it is far more core question in intelligence: How to generate consciousness? As Marx pointed out，Language and consciousness，Just because it is needed，Due to the urgent needs of others to be generated。Perhaps，In a sense,，Consciousness is a kind of "interaction"bet365 live casino games 。Regardless of the inner interaction，Still external interaction，All are the calculation of the situation and value of the human -machine environmental system.，instead of simply a simple fact label calculation。Knowledge has its scope and background，Leaving these range and background，The connotation of knowledge will change greatly (even the opposite)。We not only need to find unknown，Sometimes it is necessary to find known in the unknown。In these respects，ChatGPT is basically incompetent。Further talk，Artificial intelligence can perceive the objective fact environment environment through various sensors、Identification、response, etc.，but generating subjective value is a problem that is far from being solved at present。Smart and Language、Existing rules，Existing rules、A rational side，There are also right rules、irrational side。Whether it can be achieved with limited reflection unlimited、To reflect inevitable、Objectively reflect subjective，Perhaps it is an important sign that tests a system intelligent level。Intelligence can not be achieved by relying on calculations，Instead, you need to add calculations，It is possible to form an intelligent system of human -computer environment systems with deep situation perception。

　　bet365 best casino games