AI is the simulation of human intelligence processes by machines, especially computer systems. These processes include learning (the acquisition of information and rules for the use of information), reasoning (using the rules to reach approximate or final conclusions) and self-correction. Particular applications of the AI include expert system speech recognition and artificial vision.
2. Artificial Intelligence AI :
AI is the simulation of human intelligence processes by machines, especially computer
systems. These processes include learning (the acquisition of information and rules for
the use of information), reasoning (using the rules to reach approximate or final
conclusions) and self-correction. Particular applications of the AI include expert system
speech recognition and artificial vision.
AI was born in a meeting held in the summer of 1956 in Dartmouth
(United States) in which participated who later have been the principal
investigators of the area. For the preparation of the meeting, J. McCarthy,
M. Minsky, N. Rochester and CE Shannon drafted a proposal in which
the term "artificial intelligence" appears for the first time. It seems that
this name was given at the behest of J. McCarthy.
AI is now a name that includes many processes and various
applications. Artificial Intelligence is essentially divided into two
areas: Machine Learning and Deep Learning.
3. History of Artificial Intelligence
Here we look back at five moments in history that have influenced the development of AI:
1. Alan Turning, universally acclaimed as the father of modern computers , published a paper in
1950 describing the Turing Test - also known as the "imitation game" - testing whether a
machine could influence a person to believe it she is human.
2. The computer scientist John McCarthy coined the term 'artificial intelligence' in 1956 at a
conference at Dartmouth University. As a result, the US government gave McCarthy and his
fellow scientist Marvin Minsky the financial resources to develop AI to strengthen their position
in the Cold War with Russia. There were efforts to use artificial intelligence to understand the
patterns of the Russian language in the hope that it would enable them to translate Russian
documents to a greater extent faster.
3. The 1970s saw the beginning of the winter season . Government funding for AI was cut when
not enough progress was visible. In 1973, Professor Sir James Lighthill argued that machines
would never be able to achieve more than the level of an "experienced amateur" in chess.
4. As a result of a rise in funding for AI and its economic success in the 1980s, IBM's
supercomputer Deep Blue hit world chess champion Garry Kasparov in 1997. Deep Blue was
able to analyze up to 200 million potential positions in a second.
5. In 2016, a team at Google taught their computer to keep secrets by creating neutral networks
that could encrypt and hide information from each other. They taught the networks "Alice" and
"Bob" to encrypt and share information while preventing the third network "Eve" from
decrypting it.
4. What is Machine Learning?
Machine Learning "combines algorithms that learn from examples, data" . This makes it possible
to "predict values from datasets that serve as an example". Consequence: the quality of the
result depends on the quality of the data provided to the learning system.
•Supervised learning , which provides examples to learn the system and provides the right
answer. This type of AI is used to identify video content, predict the selling price of a home based
on a history, or predict medical risks.
•Learning unsupervised , is to provide many examples to the system, but this time without
giving good answers. What machine learning to group customers by affinities, detect anomalies
(banking fraud), or detect correlations (to put two products side by side in a store rack for
example).
•The semi-supervised learning is to provide many examples to the system, and the right
answer for some of them. Enough to allow Google or Facebook to "recognize" people in the
photos you have hosted by these services.
•Reinforcement learning allows a system to evolve in a physical (i.e., a robot) or virtual
environment. The system is improving with punishments and rewards. This is how a robot learns
to walk alone, or a bot in a video game improves.
5. What is Deep Learning?
Deep learning is based on the ability of a technology to learn from raw data. Word processing,
voice or facial recognition; the applications are numerous.
Yann LeCun and Geoffrey Hinton are the two leading specialists in this field, which exploits neural
networks, inspired by biological neurons. If these digital neurons can be connected in various
ways, the neural networks are in most cases composed of superimposed layers. The neural
network is trained to recognize the contents of an image. Depending on the result, the
"connection strength" between each neuron is corrected. Thus, the neural network is perfected
until no more recognition error.
"The Deep Learning is very powerful but also very expensive" says the Cigref guide. "It should be
kept for the most important classes and use other machine learning algorithms that can result in a
sufficient result at a lower cost.“
AI does not boil down to machine learning and deep learning however warns Olivier Ezratty in a
recent blog post . "The vast field of AI includes other techniques, including symbolic AI, logic
programming and rule engines, and the news has muted them because of the tinkle around deep
learning. This one has limitations: the best AI solutions often integrate and assemble several
different techniques. “
6. Types of artificial
intelligenceAI can be categorized in any number of ways, but here are two examples.
The first classifies AI systems as weak AI or strong AI. Weak AI, also known as narrow AI, is an
AI system that is designed and trained for a particular task. Virtual personal assistants, such as
Apple's Siri, are a weak form of AI.
Strong AI, also known as artificial general intelligence, is an AI system with generalized human
cognitive abilities, so that when presented with an unknown task, it has enough intelligence to
find a solution. The Turing test, developed by the mathematician Alan Turing in 1950, is a
method used to determine if a computer can really think like a human, although the method is
controversial.
The second example is from Arend Hintze, an assistant professor of integrative biology and
engineering and computer science at Michigan State University. It categorizes the AI into four
types, from the type of AI systems that exist today to the sensitive systems, which do not yet
exist. Its categories are the following:
7. Types of artificial
intelligence• Type 1: Reactive machines. An example is Deep Blue, the IBM chess program that beat Garry
Kasparov in the 1990s. Deep Blue can identify pieces on the chessboard and make predictions,
but has no memory and can not use past experiences to inform future ones. Analyze possible
movements - your own and those of your opponent - and choose the most strategic
movement. Deep Blue and Google's Alpha GO were designed for narrow purposes and can not
be easily applied to another situation.
• Type 2: Limited memory. These AI systems can use past experiences to inform future
decisions. Some of the decision-making functions in autonomous vehicles have been designed
in this way. The observations are used to inform the actions that occur in the not so distant
future, such as a car that has changed lanes. These observations are not stored permanently.
• Type 3: Theory of the mind. This is a psychological term. It refers to the understanding that
others have their own beliefs, desires and intentions that affect the decisions they make. This
type of AI does not exist yet.
• Type 4: Self-knowledge. In this category, AI systems have a sense of themselves, they have
consciousness. Self-aware machines understand their current state and can use the information
to infer what others are feeling. This type of AI does not exist yet.
8. Examples of AI technology
• Automation is the process of automatically creating a system or a process function. Robotic
process automation (RPA), for example, can be programmed to perform high volume
repeatable tasks normally performed by human beings. RPA is different from IT automation in
that it can be adapted to changing circumstances.
• Machine learning is the science of getting a computer to act without programming. The deep
learning is a subset of machine learning, in very simple terms, it can be considered as the
automation of predictive analytics. There are three types of machine learning algorithms:
supervised learning, in which data sets are labeled so that patterns can be detected and used
to label new data sets; unsupervised learning, in which the data sets are not labeled and
classified according to similarities or differences; and reinforcement learning, in which the data
sets are not labeled, but after performing an action or several actions, the AI system receives
feedback.
• The vision of the machine is the science of making computers see. The vision of the machine
captures and analyzes the visual information using a camera, the analog-to-digital conversion
and the digital signal processing. It is often compared to human sight, but artificial vision is not
linked to biology and can be programmed to see through the walls, for example. It is used in a
wide range of applications, from the identification of the signature to the analysis of medical
images. Computer vision, which focuses on machine image processing, is often combined with
artificial vision.
9. Examples of AI technology
• Natural language processing (NLP) is the processing of human and non-computer
language by a computer program. One of the oldest and best-known examples of NLP is
the detection of spam, which looks at the subject line and the text of an email and
decides if it is garbage. Current approaches to NLP are based on machine
learning . NLP tasks include text translation, feelings analysis and speech recognition.
• Pattern recognition is a branch of machine learning that focuses on the identification of
patterns in the data. The term, today, is outdated.
• Robotics is an engineering field focused on the design and manufacture of
robots. Robots are often used to perform tasks that are difficult for humans to perform or
it is difficult for them to perform consistently. They are used in assembly lines for the
production of cars or by NASA to move large objects in space. More recently,
researchers are using machine learning to build robots that can interact in social
settings.
10. AI applications
AI in healthcare. The biggest bets are on improving patient outcomes and reducing
costs. Companies are applying machine learning to make diagnoses better and faster than
humans. One of the best-known healthcare technologies is IBM Watson. He understands
natural language and is able to answer the questions that are asked. The system extracts
patient data and other available data sources to form a hypothesis, which it then presents with a
confidence score scheme. Other AI applications include chatbots, a computer program used
online to answer questions and help clients, to help schedule follow-up appointments or to help
patients through the billing process, as well as in virtual health assistants who provide basic
medical feedback.
AI in business. The automation of robotic processes is being applied to highly repetitive tasks
that humans normally perform. The machine learning algorithms are being integrated into the
analysis and CRM platforms to discover information on how to better serve
customers. The chatbots have been incorporated into websites to provide immediate service to
customers. The automation of jobs has also become a point of conversation between
academics and IT consultants, such as Gartner and Forrester.
11. AI applications
AI in education. The AI can automate the rating, giving educators more time. AI can evaluate
students and adapt to their needs, helping them to work at their own pace. AI tutors can
provide additional support to students, ensuring that they stay on track. AI could change
where and how students learn, perhaps even replacing some teachers.
AI in finance. The AI applied to personal finance applications, such as Mint or Turbo Tax, is
transforming financial institutions. Applications like these could collect personal data and
provide financial advice. Other programs,IBM Watson being one, have been applied to the
process of buying a house. Today, software performs much of the operations on Wall Street.
12. AI applications
AI in the law. The process of discovery, through the review of documents, in the law is
often overwhelming for human beings. Automating this process is a better use of time and
a more efficient process. The startups are also building computerized assistants with
questions and answers that can sift questions programmed to answer by examining the
taxonomy and ontology associated with a database.
AI in manufacturing. This is an area that has been at the forefront of incorporating
robots into the workflow. Industrial robots used to perform unique tasks and were
separated from human workers, but as technology advances that has changed.
13. Conclusions
Computing has advanced enormously since its inception 70 years ago. The computing
power has been multiplied by two every 18 months, following Moore's law. It is believed that,
if Moore's law continues to be met, by 2030 the computing capacity of a processor will
correspond to that of a person.
In turn, the amount of information stored digitally today is huge. Search engines like Google
store millions of copies of existing web pages, and mail services companies accumulate
millions of our messages. Social networks register our interests and our
friendships. Companies keep any information, however insignificant, in case it may be of
some use in the future.
Naturally, an increase in the speed of computing and a greater storage capacity will mean
that the systems have more resources to make decisions and that these decisions are made
in a more informed way and, in turn, in a more personalized way.