Aligning advanced AI with human interests

Applied Mathematical Modelling

How can we use mathematical language so that we can describe the behaviours of artificially intelligent system?

Simple Linear Regression


Problems of self-reference in self-improvingspace-time embedded intelligence
Regression analysis is a statistical procedure for assessing the connections among variables. It incorporates numerous methods for demonstrating and investigating several variables, when the emphasis is on the relationship between a dependent variable and one or more independent variables. Usually, more than one independent variable impacts the dependent variable. Simple linear regression is one of the methods used in regression analysis where there is only one independent variable. In simple linear regression we predict one variable from the second variable. In this process when prediction of unknown variable are plotted with the help of known variable a straight line is formed.

Logical uncertainty

This section is under development, We are working to get online soon.

Error Tolerance

This section is under development, We are working to get online soon.

Value specification

This section is under development, We are working to get online soon.

Artificial Intelligence Software

How should bounded agents reason with uncertainty about the necessary consequences of their decision criteria and beliefs?

Abstract agents are often assumed to be logically omniscient. The largest obstacle to a formal understanding of highly reliable general-purpose agents may be our poor understanding of how agent behavior changes when that assumption ceases to hold.

Fields that will need to be extended to accommodate logical uncertainty range from probability theory and provability theory, which model how agents should ideally revise their beliefs, to decision theory and game theory, which model how agents should ideally calculate the expected utility of their available actions (and the actions of others). Together with realistic world-modeling, we consider these to be general problems for

Artificial Intelligence


Problems of self-reference in self-improvingspace-time embedded intelligence
There are two types of people in scientific community at the time, one who think that one day they would be slaves to beings with artificial intelligence and the other who think that the world will be a great place because of the joint work of humans and artificial intelligence. John McCarthy an American computer scientist and cognitive scientist was the one who coined the term Artificial Intelligence in the 20th Century. He was of the opinion that artificial intelligence can have beliefs. A philosopher John Searle disagreed with McCarthy on this topic. Other than these two impressive minds a lot of other people were also been involved in discussions favoring both sides. The argument is not just limited to beliefs but also to feelings, thinking capability and emotions.

Facial Recognition and Image Processing


Problems of self-reference in self-improvingspace-time embedded intelligence
In this era of information, security for your telecommunication devices and other private information has become more and more important however keeping that private information secure has become a great risk. One example is the leak of personal information of celebrities and other people from different websites and forums .There have been a lot of inventions for improving the security to your personal information, but then there are hackers who can hack into your passwords and pattern locks. Facial recognition is a tool that is one of the strongest piece of security for your device. Facial recognition system is a software that matches the shape and facial features of your face to a picture or video that you had already stored in the system. It is mostly used for securing important information or getting access to a certain area but now a days media companies are also using facial recognition and image processing to improve their video quality and to add CGI content to their movies.

Image processing plays a great role in facial recognition system. Government agencies use image processing in order to improve the quality of photos that they possess. Facial recognition and image processing complement each other one without the other is almost useless. If facial recognition software doesn’t have a good image processing system then it would face a lot of problems while identifying a person. Better image processing results in better facial recognition. Same way if your image processing system is good but the facial recognition system is outdated then still it might face a lot of problems.

Natural Language Speech processing


Problems of self-reference in self-improvingspace-time embedded intelligence
Living beings whether animals or human beings tend to communicate with one another in order to fulfill their social needs. With time human beings have come up with technologies that help us communicate with machines. Even a simple calculation on a calculator is a form of communication between man and machine, as we put in the data and the calculator responds with a solution. With the advent of 21st century, the information technology sector all over the world saw a massive boom. With the advancement in Information technology these communications between man and machine have become more modified and the communications have become easier.

Deep Learning & Thinking


Problems of self-reference in self-improvingspace-time embedded intelligence
The theory of evolution suggests the homo-sapiens or human beings are getting intelligent with time. This means that with the passage of time, the critical thinking skills in human being are developing but still people have to be groomed to be deep learners and deep thinkers from the start. It should be the duty of teachers and parents to develop and groom critical thinking skills in their students and children respectively. People like Isaac Newton and Albert Einstein were both deep learners and deep thinkers. This was one of the biggest reasons that they succeeded in their respective fields. It is a leader’s quality to be a deep learner and deep thinker, this is why most of the high rank officials in the army are deep thinkers and learners as they have to make very important real time decisions.

Deep learning relates to analyzing data on real time basis and then remembering it as an experience to use in any similar future matters. Learning something and then forgetting it is not deep learning, it is more like surface learning so when you learn something you should try to keep it in your mind. This would ease up your decision making process. Revisiting what you have learnt from past experiences while analyzing real time data is what makes one a good deep learner.

Emotional Factor Recognition


Problems of self-reference in self-improvingspace-time embedded intelligence
While trying to build a robot similar to that of a human, scientists have come up with a lot of different inventions that are not as efficient as their natural counterparts but they are surely very useful inventions. Human beings feel emotions, our emotions are triggered by external stimulus that hit our brain e.g. a failure in life may make you sad. Most commonly our facial expression is what shows our emotions, if we are happy we smile, if we are sad we have a sad frown on our face etc. This is a natural emotion recognition system. Emotional factor recognition is one of the most important and most user friendly artificial intelligence application to date. It is a tool that actually takes care of you as it adjusts the environment according to your emotions and likening. As it is a form of artificial intelligence, it is also called emotional intelligence.

Smart Navigation


Problems of self-reference in self-improvingspace-time embedded intelligence
Gone are the days when you had to ask people for directions. Technology has proved overtime that it has the solution to all the problems that mankind faces. Navigation system for cars was first introduced in 1980s by HONDA. The concept of navigation system for cars was developed after the Global Positioning System model used by lost travelers to find their way in unfamiliar lands. It was a simple navigation tool which showed a map of your surroundings.

With time navigation has turned into smart navigation. In the end of 20th century people started to rely more and more on navigation system to reach their destinations, so by the start of 21st century navigation companies started manufacturing more modified navigation systems for cars. Information technology plays a very important t role in our lives. It is involved in almost everything that we come in contact with on daily basis, same is the case with navigation system. With the boom in information technology navigation systems started to become more smart and user friendly. One other advancement that we have seen in the twenty first century is that a lot of new companies have laid their foot in the market and they are coming up with different types of navigation systems. Some companies have even introduced navigation systems for motor bikes as well as bicycles.

Neural Networks


Problems of self-reference in self-improvingspace-time embedded intelligence
A neural network is a term coined for the interconnected neuron structure in the brain. Work started on neural networks in early 1940s.The neural network is the basic and most important part of the brain. It is what helps the nervous system to process all the messages sent from the body to the brain. It is considered to be the most complex part of a human brain. Neural networks trigger emotions in a person. Extensive studies on the human brain have helped scientists to replicate the neural network to some extent, which is called the artificial neural network or ANN. Artificial neural network is information processing tool which makes it easier for machines to respond to any external stimulus .Artificial neural network is actually the brain of an artificially intelligent machine or robot. An artificial neural network or ANN can be used to solve complex mathematical problems, data processing and as discussed earlier it is also used in robotics. It is one of the most important scientific inventions to date. A lot of continuous improvements and modifications are being performed on neural networks since mid 2000s.

Self Programming and Correcting Languages


Problems of self-reference in self-improvingspace-time embedded intelligence
We call self-programming the worldwide procedure that animates computational operationally autonomous systems, i.e. the application of mutually operational and semantic conclusions. Therefore, a self-programming machine (the self) is instituted by three categories of code:

C1: The programs that perform on the word and the self
These are the programs that calculate the assembly and implementation of code. They function in any of the three categories subsequently.
C2: These are the models that designate the program in C1, comprising the self in the world and programs in the self.

Unsupervised Learning


Problems of self-reference in self-improvingspace-time embedded intelligence
One of the most basic problems of language acquisition is accounting for how children learn the inventory of word forms from speech—phonological lexicon discovery. In learning a language, children face a number of challenging, mutually interdependent inference problems. Words are represented in terms of phonemes, the basic phonological units of a language. However, phoneme inventories vary from language to language, and the underlying phonemes
which make up individual words often have variable acoustic realizations due to systematic phonetic and
phonological variation, dialect differences, speech style, environmental noise, and other factors. To
learn the phonological form of words in their language children must determine the phoneme inventory
of their language, identify which parts of the acoustic signal correspond to which phonemes— while discounting surface variation in the realization of individual units—and infer which sequences of phonemes correspond to which words


Artificial Intelligence HardWare

How can we program an advanced artificial intelligence machine to assist with its own online debugging and modification of its goal orientation?

Rational agents pursuing some goal have an incentive to protect their goal-content. No matter what their current goal is, it will very likely be better served if the agent continues to promote it than if the agent changes goals. This suggests that it may be difficult to improve an AI’s alignment with human interests over time, particularly when the AI is smart enough to model and adapt to its programmers’ goals.

Material Science


Problems of self-reference in self-improvingspace-time embedded intelligence
With the name “material sciences” we can easily understand that it’s a study something about the materials emphasizing the solids. It includes the detailed study how materials are constructed, how they are applicable in our routine life, the manufacturing and properties of materials. This study makes it possible to understand the behavior of materials in a very systematic manner. It’s quite interesting to study how a material is obtained, why we obtained that kind of structure, and the facts that how and why the material in behaving or reacting in a peculiar behavior. Material sciences make you aware of how to establish a fabric by correlating the properties of materials with their structures. It is the field of science that explains the application of diverse materials in science and technology for example the plastics, metals and ceramics. The use of this field of science is quite expanded. It covers almost everything that we use in our daily routines. From the construction of huge buildings to tiny silicon chips a material sciences expert is needed to analyze and determine the optimal usage of each and every material. Manufacturing of giant girders, even the Nano machines comes under the branch of material sciences.

Upgradable

This Section under developement.

Arduino Compatible

This Section under developement.


Artificial Intelligence Mechatronics

How can we sequence an advanced artificial intelligence machine to assist with its own online debugging and change its own objectives?

Smart Components


Problems of self-reference in self-improvingspace-time embedded intelligence
If you believe “nothing is impossible”, then you are absolutely right. There is nothing a human cannot do. A human brain has the ability to discover and innovate things as per the situation. In the times we are living, we observe thousands of new innovations and technologies which are introduced in the market on a daily basis.

Like many other technologies, scientists have made some components which can resist lateral force. The simplest answer to the question that what are smart components is that they are the systems used to resist lateral force. They are usually used in construction of any infrastructure in order to avoid any damage. They help the building resist any shocks or any other strong waves. In any infrastructure the everlasting roof and the strong floor are the most important things. These smart components can be used in designing a building to make it safe from any major catastrophe.

Integrated Technology


Problems of self-reference in self-improvingspace-time embedded intelligence
As we all know this is the age of technology and lots of new innovations are seen daily in every domain of life. To complete a task, we require the help of technology. Every single task in our life is done with a piece of technology, most commonly known as devices. Researchers are now working towards new innovations which includes qualities of different technological devices in one. Through this, not only various technologies will be accessible with the use of a single device, but it will also be easier to handle one single unit rather than carrying three or four devices. To be quite honest we you, we simply can’t imagine life without technology.

Alternative Energy


Problems of self-reference in self-improvingspace-time embedded intelligence
For many years humans have relied on fossil fuels as a source of energy. With the passage of time more and more resources have been discovered and new ways of generating energy have been proposed. The term alternative energy refers to any source of energy other than fossil fuel. Some ecologists believe that the term alternative energy is somewhat controversial as other sources are also as effective and important in generating energy as fossil fuels.
Commonly alternative energy refers to solar energy, energy generated from wind turbines, biomass, geo-thermal energy etc. With each passing year these sources are becoming as much important in generating energy as fossil fuels. One very important alternative to fuel is electricity because now a days we have electrically powered vehicles and stoves.


Maximising User Experience

How can we maximise the user experience of artificial intelligence machines?

Design Thinking


Problems of self-reference in self-improvingspace-time embedded intelligence
Design thinking is a process which helps in solving complex problems in the most creative way. It is a cyclic process in which the prototypes of the solutions are built and feedbacks are asked regarding the prototypes until a final solution is designed. It is like going on a trip without a map or without even knowing the destination, but with the confidence that you will end up somewhere great.
Empathize is the foundation of design thinking. It involves asking questions based on the empathy map. An empathy map is a set of questions which helps in understanding the customers better. It involves questions such as what do the customers think about the services provided by organization, what customers hear from the organization about them and from others about the organization, what are their hopes from the organization, what are their fears related to the organization etc.

Limits to self organizing


Problems of self-reference in self-improvingspace-time embedded intelligence
What and how much can children learn without subject teachers? In an attempt to find a limit to self organized learning, we explored the capacity of 10–14 year old Tamil-speaking children in a remote Indian village to learn basic molecular biology, initially on their own with a Hole-in-the-Wall public computer facility, and later with the help of a mediator without knowledge of this subject. We then compared these learning outcomes with those of similarly-aged children at a nearby average-below average performing state government school who were not fluent in English but were taught this subject and another group of children at a high-performing private school in New Delhi who were fluent in English and had been taught this subject by qualified teachers.We found that the village children who only had access to computers and Internet-based resources in the Hole-in-the-Wall learning stations achieved test scores comparable with those at the local state school and, with the support of the mediator, equal to their peers in the privileged private urban school.

User Friendly

This Section under developement.

Intuitive

This Section under developement.

Customizable

This Section under developement.