Let us focus our projects on improving Productivity to increase our participation in business.
1.What is called Artificial Intelligence
I propose that we define concepts to know what we are talking about when we say "intelligence" and " Artificial Intelligence"
The Royal Spanish Academy defines:
INTELLIGENCE:
- Ability to understand or comprehend Ability to solve problems Knowledge, comprehension, act of understanding
ARTIFICIAL INTELLIGENCE:
1.f. Inform. Scientific discipline that deals with creating computer programs that perform operations comparable to those performed by the human mind, such as learning or logical reasoning.
With these considerations in mind, we can assume that a simple manual calculator that solves elementary mathematical operations meets the conditions described.
The answer is yes
For example multiply 128 x 2 =
The calculator performs a procedure called the Genetic Algorithm that we will see later.
First solve the unit 2 x 8 = 16 16
Then solve the ten 2 x 2 = 4 and the addition to the ten of the previous result
4 + 1 = 5 56
And finally solve the hundred 2 x 1 = 2 and the sum of the result already obtained in the corresponding place of the hundred 2 + 0 = 2 256 final result
And so, complying with this algorithm loaded in the program, it solves all the multiplication operations that are presented to it.
This simple calculator understood the problem before it, solved it, and finally stated the expected answer 128 x 2 = 256
Very simple right? But it helps us to understand what we are talking about, from this simple example to a Chatbot, for example, that understands what is being proposed verbally and responds consistently to the query that is posed, also with an algorithm built into the machine. The range of applications is very wide and we will see later everything that these new tools cover.
Beyond that we present a problem, understand it and respond to it on a screen, the result can also be an order that is assigned to a Robot.
And here appears a new term that we had not described and that the Royal Spanish Academy also defines as:
Robot:
- Programmable mechanical / electronic machine or device capable of manipulating objects and carrying out operations previously reserved only for people. Program that automatically scans the web to find information.
Robots operate with simple mechanical algorithm programs and can also contemplate AI contributions, examples:
- In an automotive terminal assembly line there are Robots that operate when a limit switch indicates that a part of the car reaches a certain point and operates a mechanically programmed weld and, each time the limit switch announces that a new one has arrived. part, performs the welding and the part continues traveling.In the same line there is another robot that places a door on the bodywork and after locating it controls that the light in its environment is the corresponding one using the VISION PRO program that can read the spaces between door and bodywork in all its contour and if they are not the ones that correspond, arrange the door to place it where it should be located.
Conclusion: we define what we are talking about when we mention INTELLIGENCE, when we mention ARTIFICIAL INTELLIGENCE and what ROBOTS mean in business management.
2. Circumstance and date AI was considered an independent science.
Since the origins of life on the planet, humans have carried out work and solved their needs with the effort of their hands, arms and legs, with physical effort.
As events progressed, imagination and creativity devised tools that helped man carry out his tasks.
Later they designed and manufactured machines that man helped to do the job, without the need to strain or perform strenuous tasks.
These stages relieved people from fatigue and exhaustion due to effort and repetition of movements, always acting on the body and physical activities. But gradually, throughout this period, an attempt was made to resolve the mental work as well.
Although it seems incredible, there are traces that indicate that ancient civilizations (Greeks, Chinese and Mayans) were already concerned about incorporating some intelligence into certain machines (1384 BC)
In 1849 George Booler succeeded in establishing principles of Proportional Logic.
In 1874 Frege invented the system of mechanical reasoning which he called "concept writing"
In 1950 Alan Turing wrote the first modern article that addresses the analysis of the possibility of mechanizing intelligence.
In 1956 John Mc.Carthi and Claude Shannon introduced the term Artificial Intelligence in the technical and scientific community.
Herbert Simon, Allen Newell and Marvin Minsky formalized the basic ideas about AI and developed the subject area of specialty in 1980.
However, in 1984 E. Dison was the first to speak out against this trend, leading many to think that AI was dead.
However, the studies and applications of AI continued to advance and the year 1956 was considered the birth of the Quanta Industrial Revolution by separating the treatment of AI as an independent science of Computer Science.
In 1997 IBM uploads the Deep Blue program to a computer and proposes a confrontation with current world chess champion Gary Kasparov.
The result of the confrontation gave Deep Blue the winner by 3 ½ to 2 ½, enthusing scientists and unleashing a career of researchers and developers of AI programs that resulted in a number of useful applications for business and life in general.
This was the best answer on the possibility of life of the AI
+ PRODUCTIVITY + EFFICIENCY
We focus projects on improving Productivity to increase our participation with more efficiency in business
The axes of study on which we worked.
DEFINITIONS OF ARTIFICIAL INTELLIGENCE
Rich & Knight, Stuart, generally define AI as "the ability of machines to perform tasks that are currently performed by human beings"; other authors such as Nebendah, Delgado, provide more complete definitions and define them as the "field of study that focuses on the explanation and emulation of intelligent behavior based on computational processes based on experience and continuous knowledge of the environment."
In other words, it is the ability to understand the problems and situations that arise, carry out situation analysis emulating the human mind, memorize and propose or order actions aimed at solving the stages under consideration.
Basically what is intended of artificial intelligence is to create a machine or a sequential program that repeats indefinitely a set of instructions generated by a human being or by the same machine / program.
RESEARCH AND DEVELOPMENT ON AI ROLLED ON THREE AXES
- NEURONAL NETWORKS DIFFUSE LOGIC SYSTEMS GENETIC ALGORITHMS
Neural networks
Remember that the human brain is made up of billions of neurons interconnected to form circuits or networks that perform specific functions.
A typical neuron picks up signals from other neurons or different inputs through structures called dendrites.
The neuron emits pulses of electrical activity along a thin, thin layer of connectors called axons. The extremities of these branches reach the dendrites of other neurons and establish connections called synapses, which transform the electrical impulse into a neurochemical message by releasing substances called neurotransmitters that excite or inhibit other neurons.
In this way, the information is transmitted from one neuron to another and is being processed through the synaptic connections, concluding in an output information that generates the expected actions.
Neural networks are structured in several layers:
First layer as input buffer, storing the raw information supplied in the network or carrying out a simple pre-processing of it, we call it input layer.
Another layer acts as an interface or output buffer that stores the response of the network so that it can be read, we call it the output layer.
And the intermediate layers, the main ones in charge of extracting, processing and memorizing the information, are called the hidden layer.
Fuzzy logic systems
Fuzzy logic systems, also called fuzzy logic, is the second tool that allows us to emulate human reasoning. This type of logic takes two random values, but contextualized and referred to each other. Thus, for example, a person who measures two meters is clearly a tall person, if previously the value of short person has been taken and has been established in one meter. Both values are contextualized to people and refer to a linear metric measure.
Human beings think and reason through words and in degrees between two states, for example black and white, hot and cold, etc. These fuzzy logic systems differ from traditional expert systems that interpret concrete and absolute values. Fuzzy logic systems allow the use of human language as we reason and express ourselves.
Expert systems are computer applications that make decisions or solve problems in a certain field, such as production systems, finance or medicine, using the knowledge and analytical rules defined by experts in that field. Experts solve problems using a combination of fact and data-based knowledge and reasoning skills
Genetic algorithms
In mathematics as well as in communication sciences and related disciplines, an algorithm is a well-defined, ordered and finite sequence of operations that allow finding the solution to a certain problem.
Starting from an initial state (entry) and through successive steps, successful final results can be reached. Its importance lies in showing how to carry out the process in order to solve mathematical or other problems.
An algorithm to be considered as such must be an ordered, finite and defined sequence of instructions.
Thus, in this way, it is possible to follow and predict the behavior of a particular algorithm from a possible input and, from there, following the ordered and defined sequences of instructions without giving rise to ambiguities, so that only the prescribed path can be followed from beginning to end.
If we make an analogy with the Darwinian genetic algorithms assimilated to the UCCM (brain, body, mind unit) we find: short algorithms, less precise and with lower energy consumption and long algorithms, more precise and with higher energy consumption.
The short path, complex CR + CM algorithms (reptilian path + mammalian path) only use 5% of the incoming information to start acting.
The long-way, complex CR + CM + LPF (low pass filter) algorithms use 100% of the incoming information, they are slower but substantially more accurate.
In our daily lives we have incorporated algorithms of various types and functions, from the algorithm that allows us to perform a multiplication between two numbers, play music, drive a vehicle, etc.
The algorithm gives us the generic resolution to a problem and we can use it every time that problem occurs, for example, the division algorithm can be used whatever the numbers we have to operate with. We don't need to understand how that algorithm works because it follows the preset and initially coded instructions.
4. Test that defines AI programs
"A computer can be called" intelligent "if it manages to fool a person into believing that he is a human" This phrase was pronounced by the English mathematician Alan Turing who was called the father of Artificial Intelligence.
Given the advancement of research and development of computer programs (1950), Alan Turing was in charge of designing a test that would define what behavior a computer had to have in order to consider that it was working with Artificial Intelligence.
The Turing Test was born as a method to determine if a machine can think. Its development is based on the imitation game.
The proposal had the participation of three people: a man, a woman and an interrogator who communicates with them only by writing in a language understandable to all three and does not see or is seen by the other two participants.
The experience consists in that the interrogator had to discover who was the woman and who the man, while the interrogated tried to convince the interrogator that they were both women.
The next step was to replace one of the two anonymous participants with a computer loaded with an AI program, and the interrogator should not notice the presence of the machine, assuming that it continued to communicate with two humans.
Other variants were also implemented replacing the man or the woman in different tests, but the objective of this experience was to define that the machine worked with AI when the interrogator could not recognize who was a woman, who was a man or who was a machine.
Criticisms of the TT immediately appeared with different reasoning but, fundamentally, they came from those who could not accept the idea that a machine behaved like a person to the point of being able to deceive a human interrogator.
One of the objections contemplated the "lack of awareness" of the machine, both of itself and of others, and generating positive or negative feelings about the information it contains or the actions it performs.
This behavior is called Solipsism, which indicates that the only way to know if a machine thinks is to be that same machine. The problem is that, the only way to know if another human thinks is to be that other human, which is known as the problem of other minds.
Later, the TT was perfected, transforming it into the Total Turing Test TTT and later new evaluation tests were incorporated.
Augusta Ada Byron King was born on December 18, 1815 in England and in 1838 became Countess of Lovelace .
In 1833, at only 17 years old, she was presented to Charles Babbage, an English mathematician and scientist who had the first idea of the conception of a computer, since the Analytical Engine that he built worked with the same principles of current computers.
In 1843 Lady Lovelace described and analyzed the Analytical Engine including the demonstrations of how to calculate trigonometric functions with variables, and the first program with instructions that made the calculating machine work and is recognized as the first programmer in history.
In honor of Ada Lovelace and her prestigious contribution to computing, the test that to accept that a machine works with Artificial Intelligence was called Lovelace 2.0 proposes to verify if the machine in question is capable of writing a fictional story, creating a poem or elaborating a painting.
At the moment no machine has been able to pass the Lovelace 2.0 test www.progresa–pga.com.ar
Eduardo Bronzino [email protected]
LET'S TALK ABOUT THE PRACTICAL APPLICATION IN OUR COMPANIES AND OUR BUSINESSES
"Entrepreneurs who have to define an investment must decide between: investing in a production machine or investing in artificial intelligence programs that will ensure their entry into the Fourth Industrial Revolution, the new era of business."
5. Cognitive Artificial Intelligence
The advancement of AI development went through the Analytical, Predictive and Cognitive stage.
Cognitive Intelligence is nourished, to a large extent, by Bio-inspired Algorithms that are those that emerge from the study of the biological behavior of living beings that evolved during the millions of years of existence.
Cognitive Artificial Intelligence (IAC), also called Cognitive Computing, is a specific branch that emerged from Artificial Intelligence (AI), capable of understanding and emulating the functioning of the human mind.
Let us remember that the developments of AI programs are concerned with imitating the Genetic Algorithms that allow computers to run programs that behave substituting the procedures performed by human beings.
For many years, scientists focused efforts on trying to summarize the human brain in a set of Algorithms. However, certain more modern theories, which have revolutionized the state of the art, began to consider it important to include what is known as Bio-inspired Systems, that is, machines and algorithms capable of solving problems and perceiving the environment as it does. the cognitive system of a person.
The cognitive is what belongs to or is related to knowledge, it is the accumulation of information that is available thanks to a learning process.
The development of CI results from the will of people to understand reality, incorporate it and relate it to the information memorized in the brain and produce consequent opinions and actions.
Machines are voraciously feeding on more and more data in their wake .
Cognition is the ability of living beings to process information from perception, knowledge and subjective characteristics. When a question is asked to a machine it generates a hypothesis, it generates an answer and a level of reliability; then she shows the steps that led her to give that answer. In other words, you are reasoning and learning through interaction. With each experience it gets faster and smarter.
Supercomputers understand natural language, that is, the way we humans speak, in addition, they process billions of structured and unstructured data while formulating responses based on predictions made in real time by analyzing the data collected.
AI programs evaluate thousands of pages of action fields and capture scientific insights and offer the best options in response to inquiries.
During the procedure the human load of information is produced and then all the outdated material is discarded, later the program generates an index and method to have the best access to the content (ingestion). The program "partners" with human experts who teach you to find information patterns, creating the "learning machine" that learns linguistic patterns and trains you with questions and answers, thus continuing to learn in interaction with users.
As new information is published, the program incorporates and updates the linguistic knowledge and updating of any field. In this way, the program is able to answer any type of question, providing a wide range of answers and recommendations.
6. Bio-inspired algorithms
Bio-inspired algorithms are those that result from imitating the behavior of living beings, decoding the action procedures and proposing them for the operation of AI programs
Collective Behavior of Ant Colonies
It must be remembered that ants are practically blind, and yet, moving randomly, they end up finding the shortest path from their nest to the food source and back to their anthill. It is important to make some considerations:
- On the one hand, a single ant is not able to carry out the previous work, but ends up being the result of the complete anthill. They do not do it without "instruments", but an ant, when it moves, leaves a chemical signal on the ground, depositing a substance called pheromone, so that others can follow it.
The following steps explain why the behavior of the ants makes paths of minimum distance appear between nests and food sources:
- An ant (scout) moves randomly around the colony. If it finds a food source, it returns to the colony more or less directly, leaving behind a trail of pheromones. These pheromones are attractive, ants more Nearby they will be attracted to them and will follow their trail in a more or less direct way that leads them to the source of food found by the explorer.When they return to the colony with food, these ants deposit more pheromones, thus strengthening the connection routes If there are two routes to reach the same food source, the shorter route will be traveled by more ants than the longer route. Consequently, the shorter route will increase the amount of pheromones deposited more and will be more attractive to the following ants.The longest route will gradually disappear because the pheromones are volatile and evaporate. Eventually all the ants will have determined and chosen the shortest route.
In this way, even if an isolated ant (explorer) moves essentially at random, a group of them that belong to the same anthill will decide their movements considering that they follow the path with the highest amount of pheromones more frequently.
This analysis was deepened and applied to real situations that occur in companies such as the problem of the traveling agent (TSP) that, having to visit clients in several towns and cities, the algorithm resulting from the collective behavior of the ant colonies finds the shortest and cheapest way.
7. Some practical applications of AI
The skills of computers with AI programs can be divided into three large groups: Analytical, Predictive and Cognitive
Fields of application of AI
The following are just examples, and not a complete list of fields:
- AI in medicine, including interpretation of medical images, diagnostics, expert systems to assist physicians, monitoring and control in intensive care units, prosthesis design, drug design, intelligent guardian systems for various aspects of the medicine. AI in robotics, including vision, motor control, learning, planning AI in many aspects of engineering: fault diagnosis, intelligent control systems, intelligent manufacturing systems, intelligent design assistance, design, production, maintenance, configuration tools AI in education: includes various types of intelligent tutor systems and student management systems.AI in information management: this includes web crawling, mail filtering, etc. AI in mathematics: designing tools to help with different kinds of mathematical functions. AI in the entertainment industry: AI is increasingly used in computer games with the generation of interactive cartoon movies in virtual worlds. AI in Law: for example, expert systems to help lawyers, or systems to give legal advice and help to non-lawyers. AI in architecture, urban design, traffic management, help predict the behavior of people in new environments. AI in the literature, art and music: the identification of the authors, the modeling of the processes of generation and recognition, the teaching applications. AI in crime detection and prevention, counterfeit detection, learning to detect signs of police corruption, software to monitor Internet transactions, help plan police operations, search police databases for evidence that crimes are committed by the same person, etc. AI in commerce: The Internet has allowed one of the fastest growing areas in terms of the number of applications developed to be commerce, especially electronic commerce. AI in space: remote control of autonomous space vehicles and robots.
8. Artificial intelligence in the company.
So far we have learned about a series of situations related to AI, we understood what was being said when we mentioned it, we saw how this science spun off from computer science was born, we observed the tests that define whether a machine works with AI, we knew what participation the algorithms bioinspired in the growth of the activity, etc, etc.
But what is it for us entrepreneurs? What role does it play in our jobs? What contribution can they make to our management? Etc.
As we mentioned in previous paragraphs, the range of AI or IAC programs is very wide and offers us, from simple solutions to simple problems that, nevertheless, occupy our minds, commit us to make decisions and tire us, to complex problems, with many variables, with commercial and economic risks that are difficult to solve, which address situations that require a sum of data that we do not usually have fresh in memory and are difficult to obtain.
That range of alternatives have a wide range of costs. Some of a few thousand pesos to others of several hundred thousand dollars.
If we browse the web looking for practical applications, we find a bundle of information on specific cases that, although it expands our knowledge, does not bring us closer to what is ours, our problems, our needs.
The important thing is to know what are the problems of our companies that AI can solve more efficiently and faster?
An industrial company in the market that works by projects, is constantly faced with the problem of budgeting the jobs that are presented as a commercial opportunity, first pre-design the equipment in question and only then prepare the economic budget.
This task used to take between fifteen days and a month, with the risk of losing the opportunity to access the business due to delay in response.
It was not easy to convince the heads of the company and those specifically responsible for this function that this work could be solved by an AI program. Fortunately, they agreed to develop a specific program to solve this function quickly and quickly making them more competitive.
Another well-known company for the distribution of mineral water drums at home has a large fleet of trucks, vans and personnel to make deliveries, in many different directions and times that involve a significant logistics cost.
The distribution program is carried out manually by a team of people who require a significant amount of working hours with certain difficulties, while they could use the "Traveler Agent Problem" program that would efficiently solve the optimization of distribution, faster, safer and cheaper.
One of the problems that companies dealing with retail face is defining the convenience, or not, of incorporating a new product into the supply line. This decision contemplates the application of a particular formula for each company, which includes: the physical space that the required stock would occupy, the rotation of products, the financial cost of maintaining the stock, the economic contribution of each unit and the maintenance stock, etc. The result of the application of this formula avoids doubts in decision making. Surely the development of this formula takes time and the participation of some professionals, but only once, then the response is immediate.
It is important for SMEs to familiarize themselves with AI and IAC because this is a management model that does not go back, on the contrary, it progresses faster and more efficiently.
The fundamental objective is to " Improve Productivity" and this is a very efficient model to achieve it.
It is advisable for SMEs to progressively make contact with these management models, estimating, first, whether it is convenient to use formulas that solve repetitive problems that require analyzing them each time they arise and make decisions.
Then, later on, study the possibility of implementing an ERP (Enterprise Resource Planning) involving a computer professional.
When they feel trained to run the company relying on computer systems, we suggest that they evaluate the alternative of using AI and if it is justified by IAC
IBM's Watson program results from very broad participation in the field of Cognitive Intelligence, exploring all kinds of information about the field of application, analyzing the contents, discarding obsolete information and incorporating the new one in its replacement.
| ENTRY LAYER | HIDDEN LAYER | OUTLET LAYER |
| MARKETING | GENETIC ALGORITHMS | PRODUCT OR SERVICE |
| SALES | ANALYSIS OF MANAGEMENT STAGES | BILLING / COLLECTION |
| PURCHASES | AI SOLUTIONS | DISTRIBUTION |
| FINANCE | POS-SALES |
9. Management Algorithms in Companies
The management algorithms result from meticulously analyzing all the stages of adding value of the company, from the commercial opportunity to the delivery, billing and collection of the product or service, going through all the procedures that enable the operation of the business and the interconnections that they imitate. the neural connections with their axons, dendrites and synapses.
From the graphing of the procedures, we can analyze, in practice, the possibilities of incorporating the proposed models, to optimize the time and costs of managing the company.
Let's see some examples of programs that are on the market:
Crystal
Crystal is a platform that allows you to see the personality profiles of customers and gives advice on how we should communicate with them. Thanks to its artificial intelligence, it helps us build healthier and more productive relationships. Specifically, Crystal offers clues about how and through which channel to speak, details about the behavior of our interlocutors, etc. All this working on the old principle of empathy.
Tamr
Tamr is a tool, used by large companies such as Cisco, HP or Huawei as well as by SMEs, that allows to automatically integrate and analyze endless company data to, through machine learning, allow to detect areas of improvement in the organization, foci where reduce the expense or potential risks of the company. All this through a very simple interface in which questions can be made to the system and it will be in charge of processing and returning the corresponding data interpreted so that we can simply make the best business decisions.
Recorded Future
Recorded Future is an intelligent tool that monitors, in real time and proactively, any possible cyber threat against our company. Its artificial intelligence engine is capable of analyzing billions of data continuously and anticipating a possible attack. Not surprisingly, this firm claims that they manage to launch alerts on data leaks up to 36 hours earlier than other platforms, save half a day when vulnerabilities warnings occur and obtain cybersecurity investigations up to 10 times faster than their rivals.
Gluru
Who wants secretaries when we have Gluru? It is a personal assistant that, through its artificial intelligence, is capable of managing our calendar, notifying us of meetings or events, making us reports on pending tasks, controlling our email and taking care of our electronic files.
X.ai
Along the same lines as the previous one, we find X.ai, a personal assistant who, in this case, is specialized in managing meetings in an automated way, even answering emails with our guests as if they were a real-life secretary. Thus, when we receive a request to meet with someone, we just have to copy the helpful Amy (an intelligent bot) and the virtual assistant will be in charge of “writing” a personalized email with the available dates and the proposed place for the meeting. meeting. Once the other party confirms the appointment, Amy will take care of quickly adding that appointment to our calendar.
Siri
Apple's personal assistant, Siri, can also be used in the professional environment, although obviously far from the possibilities mentioned above. It can help us, through its voice recognition, to write emails while we do other tasks and it can also remind us of appointments in our calendar that we have forgotten. And, best of all, it's free for users of devices from this company.
Talk
Conversica is an automated sales assistant that is able to autonomously engage with potential customers through two-way email conversations. The tool sends emails to leads that it has captured or that we have incorporated as sources; after which he is able to interpret the contact's response, follow up on it and, finally, notify a salesperson when there is a clear sales opportunity.
Artificial Intelligence (AI) and Cognitive Intelligence (IAC) are no longer a threat, they are a strong opportunity to take advantage of in our businesses.
Download the original file
