What is your idea about next generation of AI?
Artificial intelligence (AI), sometimes called machine intelligence, is intelligence demonstrated by machines, in contrast to the natural intelligence displayed by humans and other animals. In computer science AI research is defined as the study of "intelligent agents": any device that perceives its environment and takes actions that maximize its chance of successfully achieving its goals. Colloquially, the term "artificial intelligence" is applied when a machine mimics "cognitive" functions that humans associate with other human minds, such as "learning" and "problem solving".
The scope of AI is disputed: as machines become increasingly capable, tasks considered as requiring "intelligence" are often removed from the definition, a phenomenon known as the AI effect, leading to the quip in Tesler's Theorem, "AI is whatever hasn't been done yet." For instance, optical character recognition is frequently excluded from "artificial intelligence", having become a routine technology. Modern machine capabilities generally classified as AI include successfully understanding human speech, competing at the highest level in strategic game systems (such as chess and Go), autonomously operating cars, and intelligent routing in content delivery networks and military simulations.
Ref: https://en.wikipedia.org/wiki/Artificial_intelligence
Dear H. Naderpour,
Artificial Intelligence (A.I.) will soon be at the heart of every major technological system in the world including: cyber and homeland security, payments, financial markets, biotech, healthcare, marketing, natural language processing, computer vision, electrical grids, nuclear power plants, air traffic control, and Internet of Things.
While A.I. seems to have only recently captured the attention of humanity, the reality is that A.I. has been around for over 60 years as a technological discipline. In the late 1950’s, Arthur Samuel wrote a checkers playing program that could learn from its mistakes and thus, over time, became better at playing the game. MYCIN, the first rule-based expert system, was developed in the early 1970’s and was capable of diagnosing blood infections based on the results of various medical tests. The MYCIN system was able to perform better than non-specialist doctors.
https://www.anexas.net/Blog/Artificial-Intelligence%20
Regards, Shafagat
Dear H. Naderpour,
Artificial Intelligence (A.I.) will soon be at the heart of every major technological system in the world including: cyber and homeland security, payments, financial markets, biotech, healthcare, marketing, natural language processing, computer vision, electrical grids, nuclear power plants, air traffic control, and Internet of Things.
While A.I. seems to have only recently captured the attention of humanity, the reality is that A.I. has been around for over 60 years as a technological discipline. In the late 1950’s, Arthur Samuel wrote a checkers playing program that could learn from its mistakes and thus, over time, became better at playing the game. MYCIN, the first rule-based expert system, was developed in the early 1970’s and was capable of diagnosing blood infections based on the results of various medical tests. The MYCIN system was able to perform better than non-specialist doctors.
https://www.anexas.net/Blog/Artificial-Intelligence%20
Regards, Shafagat
Artificial Intelligence has changed the present era of technology.In the next generation the Blue Brain Project may be at the target, which is still undergoing. We have not completely designed the artificial brain which can feel the emotions....
Hello,
The next generation will be Real AI mimicking natural intelligence.
Regards,
Perhaps the next achievement in this field is to simulate human emotions through deep learning methods. Deep learning is a new name for an approach to artificial intelligence called neural networks, a means of doing machine learning in which a computer learns to perform some tasks by analyzing training examples. Today, popular applications of deep learning are everywhere.The technique drives image recognition, self-driving cars, medical image analysis, surveillance and transportation systems, and language translation, for instance. The current state of AI can do many things, but there are still challenges to workout before we reach human-like brains. Some of the current challenges with AI include learning without labels and generalization of learning.
Reading your: "AI is whatever hasn't been done yet." we can look for missing parts: One (or Few) Shot Learning, Experience Transfer (experience gained in one area can be applied to solve problems (classify, plan, etc...) in other, Continuous Online Learning, Reasoning, Memory.
Last two points are what is missing from current Deep Learning Methods, IMHO DeepMind with their Differentiable Neural Computer goes in a wrong direction by adding memory next to artificial neural nets. I'm kind of fan of Hierarchical Temporal Memory as their view combines Memory and Computation into Computational Memory (pardon for pun), this seems to be most plausible/realistic approach (on how to model brain). The problem though is that their systems have not yet solved any of the existing challenges (like image classification, recognition, sequence prediction etc.), which might be Ok, but limits adoption and therefore lacks focus of scientific community. At the same time they have won competition of object recognition based on touch ( don't have link at hand - you can find their reports on their web-site).
So, to sum-up: 1) Computational Memory + 2) Mix of unsupervised and semi-supervised algorithms (why - I think 2nd topic is important - is another page of text :) )
Hope this helps a bit.
From my viewpoint, the answer depends on how we want to define AI.
If we want to see AI as a agent taking "rational human like (or better) decisions" in a complex environment, then the next wave of AI common people will see if an aggressive optimization of the systems around them. So far planning and optimization of the systems around us - like agriculture, logistics, transportation networks, power generation. These are very complex system even individually that we cannot hope humans can optimize them taking all the zillion constraints into account. But for AI, with the huge amount of historical data we have, it can learn to optimize the systems individually and may be combined in a way which is not in human realms. If our definition for AI is taking these rational decisions, then next generation of AI is about applying/developing AI techniques to solve real world problems with domain specific knowledge. This will be something that raise the curiosity of public and benefit the world in general.
From another viewpoint, if we want to see AI as the ability of machines to develop cognitive capabilities like humans, then Andrey Bondarenko 's answer discusses some of the most pressing issues and promising avenues of opportunity. Especially, I think pure unsupervised learning (based on curiosity?) will be one the greatest breakthroughs we can look forward for. This will be something that can raise the curiosity of a researcher with some long term benefits to public.
Or if we want to see AI as machines developing humans like emotional capabilities, the answer to me is very vague. To develop such techniques, may be we have to first define objectively what does the words "emotions" and "consciousness" means to humans. This may as well be a philosophical discussion before we can scientifically "touch" it. One useful discussion along this line can be found at https://www.researchgate.net/post/Can_we_mathematically_model_consciousness.
Coming closer to human dialogue capability for AI is still a big issue.
I believe it would be a system able to build vocabulary of terms describing environment it is happen to be placed in, using both - supervised and unsupervised learning methods. The vocabulary will grow from most basic terms describing physical constraints of the agent’s embodiment, to the most complex ones describing abstract notions.
In my opinion it’s critical that learning agent has intrinsic understanding of it’s physical limits (body) as a starting point from which it will build the vocabulary of environmental terms. The agent should implement intrinsic motivation to induce explorative behavior. But at the same time, it will get active positive or negative feedback from the environment, because learning agent has clearly defined physical enclosure which is limited by laws of physics, social interactions, etc.
In my opinion attempts to build AI agents as abstract entities, without providing to the agent at least very basic notion of it’s limits (physical) will fail to produce natural intelligence.
From the engineering point of view, it seems like good direction to move in, is to start with creation with modular systems where each module has highly specialized neural network (Term NN) incorporating knowledge about specific term in the vocabulary of environment. Such modules can be generated by learning agent during exploration and upon maturation will be incorporated into it’s main substrate. At the same time it’s important to introduce some kind of plasticity (e.g. in form of Hebbian learning) to guarantee further module adaptation to comply with changes in the environment.
In order to generate highly optimized and energy efficient NN modules, we need a bit different approach than Deep NN, where usually a lot of neural units mostly in saturated or undersaturated state in which they just consume energy without any meaningful contribution into inference or learning. Going even further, with power efficient Term NN modules, it will be possible to incorporate them into specialized chips in form of firmware, i.e. each module will be unikernel OS without any intermediates between it and bare-bones hardware.
To summarize, in my opinion, the natural intelligence is possible when we have hardware-software dualism with positive/negative feedback from the environment, which is possible due to the fact of the physical nature of hardware component, which has clearly defined physical limits and sensorimotoric constraints. All known natural intelligence has some kind of embodiment in form of individual organism or a swarm of organisms, which allows it to build vocabulary of the world starting from the locus point of it’s physical being and it’s inherent abilities defined by it’s body structure.
In a very broader perspective , we can expect AI to leap forward asymptotically with natural intelligence in achieving the following ;
*AI agents capable of taking a spontaneous decision(s) in an unexpected situation.
*AI agents that can show humanistic traits like empathy, compassion, emotion and love.
*AI agents that can help animals to learn quickly.
Next generation A.I will control human beings by reading human minds.The A.I robots will surely dominate this world.So the future is alarming for human beings.SOPHIA robot has already received world attention by mimicking some human behaviors.
Comparing Sophia to AI is a bit of an insult to AI researchers - so thinks one of the father figures of Modern AI: https://www.gearbrain.com/sopha-the-robot-artificial-intelligence-2526789206.html
@Muhammad Ali Sophia is a exactly what you told - "mimicking human behavior", not intelligent in most sense.
Artificial intelligence is one of the key advanced information processing technologies of the fourth technological revolution, known as Industry 4.0.
The current technological revolution, known as Industry 4.0, is determined by the development of the following technologies of advanced information processing:
Big Data database technologies, cloud computing, machine learning, Internet of Things, artificial intelligence, Business Intelligence and other advanced data mining technologies.
Best wishes
The above question inspired me to the following considerations:
Will the artificial intelligence for Business Intelligence application be implemented as part of the ongoing computerization of analytical processes?
Business entity management processes are more and more often supported by computerized Business Intelligence platforms that facilitate multi-criteria analysis and reporting.
Complex, multi-criteria analyzes regarding the verification of large companies' operations require aggregation and analytical processing of large data sets in Big Data database systems.
Specialized IT companies produce applications that help in conducting economic analyzes, i.e. the Business Intelligence platform.
More and more often, large and medium-sized companies use these platforms to adapt them to the specifics of their business.
In addition, I note the interesting discussion inspired me to the following considerations:
Will future Big Data database systems supported by artificial intelligence be used in precise forecasting in order to verify futurological projections?
Currently, it is difficult to define this type of analytic problem. The key issue is forecasting future global problems. It is necessary to collect additional analytical data over the next years, and perhaps in the 21st century, in huge Big Data database systems supported by another generation of artificial intelligence, it will be possible to predict what may happen to the planet Earth in the future.
In the context of the above considerations, the following question is also current:
Will artificial intelligence soon replace some of the services provided by people?
Some researchers in the field of artificial intelligence indicate that in the future many professions, including work done by people, will be replaced with artificial intelligence.
For example, advisors in call center departments and commentators who write comments on internet forums, etc., are pointed out that they will be replaced by artificial intelligence.
Apparently, internet technology companies that run social media portals are working on artificial intelligence, which will take over from the operator of the social media portal.
But some researchers in the field of artificial intelligence suggest that artificial intelligence can also replace advisors in the field of high qualifications and direct contact with people, which is highly debatable.
In such theories one can find suggestions that, for example, doctors in the field of certain specialties and GPs will be replaced by artificial intelligence.
Sometimes these theories also indicate, for example, engineers, some of which specialties are to be replaced by artificial intelligence, but for engineers it depends on specific technical specialty.
Do you agree with these types of far-reaching theories?
In my opinion, it is rather excessive, exaggerated, idealized, too far-reaching theories regarding the prospects of artificial intelligence in the future.
In view of the above, it is also important to include in the above discussion also the following issue:
Will artificial intelligence combined with the analysis of data collected in Big Data database systems be used in automated forecasting of valuations of securities and other assets on capital markets?
In the background of the plot of the film "Transcendence" from 2014, directed by Wally Pfister, the topic of analysis is analyzed by a computerized system combining artificial intelligence with human consciousness loaded into a computerized system of artificial intelligence connected to Internet resources and using the resources on its own.
This independence also means access to various databases, including internal Big Data database databases belonging to certain corporations, including listed companies that are issuers of securities.
The artificial intelligence connected to the Internet, according to a specific strategy, also independently conducts transactions on the market of valuable securities.
In addition, it establishes statrtsets offering innovative technologies, which in a short time become large capital companies characterized by a high capitalization of exchange quotations of shares of these companies listed on stock exchanges.
The above-presented film "Transcendence" presented above picture of integration of artificial intelligence, human consciousness and Internet resources is a picture of typical science fiction.
However, on the other hand, the function of this image is also to inspire to discuss the potential possibilities of integration of the above-mentioned elements into one system of autonomous and working in the Internet resources artificial intelligence, in addition also containing some human feelings.
The above image from the movie "Transcendence" became an inspiration to formulate the following question below.
The mentioned motif from the movie "Transcendence" looks impressive and convincing in this film mainly because it has a full spectrum of innovation.
The use of new, innovative instruments for forecasting specific economic categories usually works until these innovative instruments continue to be innovative, ie until they are disseminated to the majority of participants in specific markets.
In addition, I note the interesting discussion inspired me to the following considerations:
Are there any ongoing work to implement artificial intelligence for the analysis processes of large collections of information collected in Big Data database systems?
Business entity management processes are more and more often supported by computerized Business Intelligence platforms that facilitate multi-criteria analysis and reporting.
Probably in the future, the business analyst will be supported by artificial intelligence.
It would be a great advance in the field of automation and objectification of multi-criteria economic analyzes of business entities.
Complex, multi-criteria analyzes regarding the verification of large companies' operations require aggregation and analytical processing of large data sets in Big Data database systems.
However, in what direction will technological progress be realized in this field?
In the future, as part of the progressing computerization of analytical processes, it will be possible to implement artificial intelligence to the processes of analyzing large collections of information collected in Big Data database systems.
Therefore, in the context of the above considerations, the following important question appears:
Will robotics and artificial intelligence once attain the level of complexity of biological organisms?
Currently, a significant part of the human genome and the genomes of many species of flora, fauna and microorganisms are already researched, known and deciphered.
On the other hand, the development of information technology also concerns the successive improvement of IT systems, operational systems and computer applications.
The computing power of processors is increasing year by year and increasingly complex operating systems, integrated IT systems and increasingly complex computer programs are being created.
Gradually, artificial intelligence is also improved, hence the complexity of multi-aspect algorithms used to develop systems based on a specific generation of artificial intelligence is increasing.
Autonomous, increasingly intelligent robots are already being created, which in Japan are already available for sale in versions of home help for the elderly.
Progress in this area probably knows no limits, at least now.
In view of the above, it is also important to include in the above discussion also the following issue:
Will artificial intelligence in the future reach the state of self-perfecting intelligence?
Different types of intelligence are studied and described, eg emotional intelligence is significantly different from classical intelligence. Intelligence is determined by various factors, both genetic and behavioral, developmental. In addition, the development of artificial intelligence, which in my opinion never matches human is done under the control of a human being, which means that the probability of realizing a negative development scenario of this technology presented in literature and film from the science fiction genre is small. It is unlikely that self-perfecting machines, which would be autonomous robots at the same time, would achieve a high level of intelligence, matching human intelligence and would rebel against people as demonstrated, for example, in science fiction movies from the "Terminator" series. Does anyone of you in this matter have a different opinion? If so, why?
I suggest the following questions and research problems in the above topic:
What are the effects of artificial intelligence implemented in autonomous cars?
Do you think that artificial intelligence will be implemented in the control systems of driving and orientation in the field in autonomous cars?
What are the effects of artificial intelligence implemented in the field of driving control systems and orientation in the field of autonomous cars?
Will autonomous cars be safe?
Will autonomous cars be mostly electric cars at the same time?
Another proposal for a research topic in this field:
Can the development of robotization and artificial intelligence cause a significant rise in unemployment in the future?
These types of questions have appeared many times in every era of the technological and industrial revolution, the period of accelerating technological development. These types of questions have already appeared in the periods of increasing the scale of objectification, arming technical human labor, from when the processes of manufacturing goods in manufactories transformed into mass production. This was the case during the Industrial Revolution of the XIX century, when the invention of a steam engine significantly accelerated the development of industry and mass production. Then, the introduction of tape production in various branches of mass production in the early twentieth century. In the second half of the twentieth century, ie in the era of subsequent stages of mechanization, automation, then the computerization of the production of many mass goods, this question appears again. Through these periods of technological progress, national economies have been transforming structurally from agricultural, industrial to modern-day domination of services. At the same time, the importance of new generation factors, which include information, technology, entrepreneurship and innovation, was gradually growing. Some branches of industry were shrinking, others were growing in the whole production of goods in the economy. At the same time, new professions, professions and specializations of human work were created, related to information, IT, analytical and technological services related to the development of new fields of knowledge and technology. So the earlier fears about the lack of work for people in connection with the technical progress that took place over the last several hundred years turned out to be essentially exaggerated. However, currently the same questions reappear: Can the development of robotization and computerization cause a significant rise in unemployment in the future? If such questions arise, then we are dealing with another era of technical progress or another technological revolution. The attributes of this revolution are also increasingly added to the development of new online media, computerized computing techniques, artificial intelligence, machine learning, Big Data, etc. in the applications of such areas of knowledge and science as biotechnology, metrology, ecology, energy, communication, medicine and many other fields of life science and new tech.
Artificial intelligence (AI) is still to imitate, extend, simulate and augment some aspects of human intelligence. For example, one of the contemporary successes is artificial foot (driverless car or other advanced car or airplane like Boeing 737 Max 8), which imitates, extends, simulates and arguments our foot so that we can use artificial foot to drive and fly from one place to another, from one country to another.
The next generation of artificial intelligence is to realize intelligent artificiality, different from imitating, extending, simulating and augmenting some aspects of human intelligence.
Intelligent artificiality consists of structured intelligence, hierarchical intelligence, process intelligence, business intelligence, marketing intelligence and other intelligences in industry, service and governance. These are more advanced intelligence than human intelligence. What we study now is the human intelligence evolved from the industry revolution and agricultural revolution. What I mentioned above for next AI is the intelligences evolved from scientific revolution since 1492.
Prof Zhaohao Sun
2019-3-14
The current level of artificial intelligence is training machine to learn raw data and associated characteristics between data points，so that the well-trained machine can predict or analyze upcoming data which is similar to training data. This is what we all say: training, validating and testing as well as predicting.
What does the next generation of artificial intelligence look like？ To my best knowledge, the next generation of artificial intelligence should focus on letting machine or artificial neural networks models understand human logic so as to process more abstract informations or signals.
The results of neurophysiological research show that human intelligence mainly depends on cerebral cortex, which is a biological deep neural network with large scale interconnection. When processing different information, biological deep neural network will enable different neuronal connection structure, that is to say, it adopts a variable structure information processing mechanism.
With respect to human intelligence, human can understand or consider a series of states as a event, which is what machine can not do currently. Let us image that if machines or artificial neural networks models can understand events and self-organize their structure sizes and learning algorithm according event-based optimization idea, we not only promote artificial intelligence move forward, but also realize real dimensionality reduction of large-scale data.
Gongming,
Thank you for your comments.
what you mentioned belong to machine learning, which is a part of AI, ML AI.
Traditional AI focused what I mentioned here. However, this is not enough. What is intelligence of intelligence, intelligence of intelligence of intelligence, this is a hierarchical intelligence. Superintelligence of intelligence. If we consider AI = A, what square of A, A2 cube of A, A3....
Mathematically, this is a topic of infinite limit of intelligence. When n trends to infinite, then what is An =?.
all these deserve our attention.
Dear All,
Though I have not had enough back ground knowledge and appreciation of the discipline, I would like to say that the next generation of AI will contain and can express human feelings like love, hate, angry etc.
Next Generation Artificial Intelligence endeavors to rise above the structure and restrictions of current AI innovation to build a novel plan of science and innovation in regards to human and computerized reasoning. It will be really advantageous for people, driving future society, industry, economy, culture and research. It, will likewise propose and understand a dream or structure of things to come society and add to next Generation training of next pioneers.
H. Naderpour According to Naci and Ioannidis, "Wellness refers to diverse and interconnected dimensions of physical, mental, and social well-being that extend beyond the traditional definition of health. It includes choices and activities aimed at achieving physical vitality, mental alacrity, social satisfaction, a sense of accomplishment, and personal fulfillment"….
Technology revolves around well being of individual and society …..these..Reserach and innovations are done to improve these. As far as I hope ...next generation of artificial intelligence ….will work for giving directions to these…...quite Broad and simple categorisation ..
1.take real time input from human body and predict …different food need of body ,exercise etc,
2. take input from brain signals ...
3.social satisfaction prediction tools..
4.May be like a mentor ….guiding you ..giving choices ..
5.Specific to a personal fulfilment ...
These are domains of work which I think …will catch race …in coming decade Various constraints and limitations are there ..with AI ,Machine learning ,Deep learning ,Adaptive learning …….will continue
- Badges
- Science topic