Artificial Intelligence - a question will be asked for each answer.
Artificial intelligence in speech expert systems is developing in the direction of chat bots. Undoubtedly, this is clear for all the direction of development. Certainly, there are a lot of problems with its formation, it can be seen from the presentations of developers . They talk quite well about their achievements, about the problems that they managed to solve, but I would also like to see the problems that they could not solve.
Let's look at all these problems with information theory. Information - information perceived by a person and (or) special devices as a reflection of the facts of the material or spiritual world in the process of communication (GOST 7.0-99).
The main functions when working with information:
According to the first function. Can be divided into several levels. The local level where, in principle, tasks are performed fairly effectively. An example would be the user's question: “Where is the nearest gas station?”, The systems provide the answer to it quite reliably. Here, in my opinion, the term “local level of searching for new information” should be defined as the possibility of a user receiving information from a source that exists in competition with other sources that are not related to AI. The presented example is good because it is naturally clear that the user can receive information about the nearest gas station from others, maps, that is, from other sources. And the choice of source is determined only by the costs of the user.
')
Tasks that cannot be solved by the user on the basis of local information can be attributed to the external (upper) level, that is, to a level that has a dimension higher than that available to the user. An example is the task of building an optimal route from A to B, taking into account the external situation (traffic jams, etc.) and user performance criteria (time spent, etc.). We see that the solution of these problems, now, is being implemented satisfactorily.
Tasks that require going beyond the knowledge of society, we will designate this level as global, and ignore it. Since, one edge of these problems is prediction, and on the other hand, the class of these problems, there is Gödel's incompleteness theorem.
On the function of saving information. I assume that these systems do it at a level that is achievable at the present time.
The third function is the transfer of information. At the moment, theoretically, any system can transfer the necessary information in various forms. We assume that the level of development for this aspect is satisfactory.
Let's see from the point of view expressed here on the development of such expert systems. Let us turn to the principles used by the developer : “Fortunately, we have all the power of the Yandex search engine, which every day encounters millions of queries, searches for millions of answers and learns to understand which ones are good and which are not.” That is, in essence, the physical scheme for finding a solution is reduced to a simple scheme:
As you can see, the answer is presented as one of the recorded, that is, already existing results, or transformed and / or agreed to one from several recorded results. Or from editorial bases : “We want the personal assistant to be a complete personality with a well-defined set of qualities. Here our editorial texts come to the rescue. Their peculiarity is that they were originally written on behalf of the person that we want to recreate in Alice. It turns out that you can continue to teach Alice on millions of lines of random texts, but she will respond with an eye to the pattern of behavior laid down in the editorial responses. And this is what we are already working on. ”
In fact, it turns out that a chat bot is an intermediary between a user and an indefinite circle of persons (legal language). And I think that with such a look at these systems, it is possible to structure tasks and problems, detail them and define new goals.
What is the conversation. This is the process of transferring information from one subject to another in one session. It is used in various circumstances. With equal subjects, to coordinate actions, with the unequal rights of subjects, to determine the actions of a subordinate subject.
Then the kinds of conversation tasks are shaped by their type.
By the number of participants:
With such clustering, taking into account the attitude to the “local level of searching for new information”, one can see that the search for tasks for AI is mainly carried out when using AI for the fourth type of task “Communication with an indefinite circle of persons”.
Let's look at the other types.
Monologue - here we will consider, as a monologue, an attempt to be the subject to systematize the results of the analysis performed, to record the results obtained. Since M.Tvenu: "Any mention in the press, even the most negative, except for the obituary, is an advertisement." In what way the AI ​​can help a person if he addressed him with a monologue. The class of answers is fairly obvious - to find similar monologues from its base. Immediately there is an aspect that is inherently, and any scientific research, before publication, is a monologue of one or several authors. Is it possible for the AI ​​to find similar studies to date? The answer is obvious - yes.
Dialogue. Take an example from social networks, based on the famous song:
- Do you love me?
- Yeah.
- Will you be with me?
- yeah
- So we will be together, so we will be next, with you always!
Analyzing this example. Can the AI ​​do something for the user at the local level? I suppose that it is insignificant, since at the local level the interlocutor, for leading the dialogue, has priority weight. But, from the point of view of the top-level tasks, there is already AI help, and therefore there are tasks for it. The first is the definition of whether the subject is talking to the bot. The second is the validity of the answers. And further, there is still a huge number of tasks: trolling, degree of interest, hidden goals, etc.
Increasing the importance of dialogue. You can see how to use the power of AI in various fields:
A) hiring;
B) negotiating;
C) Lawsuits.
And the task list appears huge.
Imagine a clear model on the example of recruitment.
Party A (employer). Submission of a vacancy. The AI ​​analyzes the submitted resumes and presents a list of suitable candidates. EychAr analyzed and refused to all, the AI ​​gives advice that it is necessary to change, add, etc. in the submission of a vacancy. If he (HR) did not manage to find an employee, then let him think about his compliance and, on the proportionality of demands and rewards. (This is from my cuff all HR).
Party B (job seeker). Writing a resume. Compilation of AI, tips, ratings, comparative results. Well, the applicant the opportunity to critically look at their capabilities.
The conference. For this type of communication, making an analogy for dialogue, you can also see a lot of tasks in which the AI ​​can be useful. From current tips to the alarm about the level of involvement in the topic of conversation of all its participants. The influence of each participant on the result to which the conference came. Search and connection of external specialists in the current mode. For unequal subjects, these are instructions for using the equipment. And much more.
Communication with an indefinite circle of persons. In this type of communication, examples of the use of AI are already a huge amount, from the creation of advertising texts to the creation of speech at the UN General Assembly.
Here an attempt to systematize the tasks before the AI ​​is presented to the public. The idea of ​​the comparison criterion of various AI is also presented. Basically, this material was cooked, according to the results of Alice’s skills tests that various developers tried to implement.
Let's look at all these problems with information theory. Information - information perceived by a person and (or) special devices as a reflection of the facts of the material or spiritual world in the process of communication (GOST 7.0-99).
The main functions when working with information:
- Search for new information;
- Saving information;
- Transfer of information.
According to the first function. Can be divided into several levels. The local level where, in principle, tasks are performed fairly effectively. An example would be the user's question: “Where is the nearest gas station?”, The systems provide the answer to it quite reliably. Here, in my opinion, the term “local level of searching for new information” should be defined as the possibility of a user receiving information from a source that exists in competition with other sources that are not related to AI. The presented example is good because it is naturally clear that the user can receive information about the nearest gas station from others, maps, that is, from other sources. And the choice of source is determined only by the costs of the user.
')
Tasks that cannot be solved by the user on the basis of local information can be attributed to the external (upper) level, that is, to a level that has a dimension higher than that available to the user. An example is the task of building an optimal route from A to B, taking into account the external situation (traffic jams, etc.) and user performance criteria (time spent, etc.). We see that the solution of these problems, now, is being implemented satisfactorily.
Tasks that require going beyond the knowledge of society, we will designate this level as global, and ignore it. Since, one edge of these problems is prediction, and on the other hand, the class of these problems, there is Gödel's incompleteness theorem.
On the function of saving information. I assume that these systems do it at a level that is achievable at the present time.
The third function is the transfer of information. At the moment, theoretically, any system can transfer the necessary information in various forms. We assume that the level of development for this aspect is satisfactory.
Let's see from the point of view expressed here on the development of such expert systems. Let us turn to the principles used by the developer : “Fortunately, we have all the power of the Yandex search engine, which every day encounters millions of queries, searches for millions of answers and learns to understand which ones are good and which are not.” That is, in essence, the physical scheme for finding a solution is reduced to a simple scheme:
-> -> As you can see, the answer is presented as one of the recorded, that is, already existing results, or transformed and / or agreed to one from several recorded results. Or from editorial bases : “We want the personal assistant to be a complete personality with a well-defined set of qualities. Here our editorial texts come to the rescue. Their peculiarity is that they were originally written on behalf of the person that we want to recreate in Alice. It turns out that you can continue to teach Alice on millions of lines of random texts, but she will respond with an eye to the pattern of behavior laid down in the editorial responses. And this is what we are already working on. ”
In fact, it turns out that a chat bot is an intermediary between a user and an indefinite circle of persons (legal language). And I think that with such a look at these systems, it is possible to structure tasks and problems, detail them and define new goals.
What is the conversation. This is the process of transferring information from one subject to another in one session. It is used in various circumstances. With equal subjects, to coordinate actions, with the unequal rights of subjects, to determine the actions of a subordinate subject.
Then the kinds of conversation tasks are shaped by their type.
By the number of participants:
- Monologue;
- Dialogue;
- Collective conversation (conference);
- Communication with an indefinite circle of persons.
With such clustering, taking into account the attitude to the “local level of searching for new information”, one can see that the search for tasks for AI is mainly carried out when using AI for the fourth type of task “Communication with an indefinite circle of persons”.
Let's look at the other types.
Monologue - here we will consider, as a monologue, an attempt to be the subject to systematize the results of the analysis performed, to record the results obtained. Since M.Tvenu: "Any mention in the press, even the most negative, except for the obituary, is an advertisement." In what way the AI ​​can help a person if he addressed him with a monologue. The class of answers is fairly obvious - to find similar monologues from its base. Immediately there is an aspect that is inherently, and any scientific research, before publication, is a monologue of one or several authors. Is it possible for the AI ​​to find similar studies to date? The answer is obvious - yes.
Dialogue. Take an example from social networks, based on the famous song:
- Do you love me?
- Yeah.
- Will you be with me?
- yeah
- So we will be together, so we will be next, with you always!
Analyzing this example. Can the AI ​​do something for the user at the local level? I suppose that it is insignificant, since at the local level the interlocutor, for leading the dialogue, has priority weight. But, from the point of view of the top-level tasks, there is already AI help, and therefore there are tasks for it. The first is the definition of whether the subject is talking to the bot. The second is the validity of the answers. And further, there is still a huge number of tasks: trolling, degree of interest, hidden goals, etc.
Increasing the importance of dialogue. You can see how to use the power of AI in various fields:
A) hiring;
B) negotiating;
C) Lawsuits.
And the task list appears huge.
Imagine a clear model on the example of recruitment.
Party A (employer). Submission of a vacancy. The AI ​​analyzes the submitted resumes and presents a list of suitable candidates. EychAr analyzed and refused to all, the AI ​​gives advice that it is necessary to change, add, etc. in the submission of a vacancy. If he (HR) did not manage to find an employee, then let him think about his compliance and, on the proportionality of demands and rewards. (This is from my cuff all HR).
Party B (job seeker). Writing a resume. Compilation of AI, tips, ratings, comparative results. Well, the applicant the opportunity to critically look at their capabilities.
The conference. For this type of communication, making an analogy for dialogue, you can also see a lot of tasks in which the AI ​​can be useful. From current tips to the alarm about the level of involvement in the topic of conversation of all its participants. The influence of each participant on the result to which the conference came. Search and connection of external specialists in the current mode. For unequal subjects, these are instructions for using the equipment. And much more.
Communication with an indefinite circle of persons. In this type of communication, examples of the use of AI are already a huge amount, from the creation of advertising texts to the creation of speech at the UN General Assembly.
Here an attempt to systematize the tasks before the AI ​​is presented to the public. The idea of ​​the comparison criterion of various AI is also presented. Basically, this material was cooked, according to the results of Alice’s skills tests that various developers tried to implement.
Source: https://habr.com/ru/post/458164/
All Articles