Science11 - Artificial Intelligence

Artificial intelligence (AI) is rapidly becoming critical to the modern world.  This article will explore the science of artificial intelligence, where it came from, what it is doing for us today, and what it might do for us in the future.

 

I will start with a definition of AI and how it works, then talk about the components of today’s AI systems, followed by the history of AI, its applications to specific problems so far, and what we can expect in the future.  I will conclude with a discussion of the potential of ultimate AI systems.

My principal sources include “Artificial Intelligence,” “History of Artificial Intelligence,” and “Artificial General Intelligence,” Wikipedia; “How Does Artificial Intelligence Work?”  csuglobal.edu; “The History of Artificial Intelligence,” sitn.hms.harvard.edu; “A Complete History of Artificial Intelligence,” g2.com; “What is Artificial Intelligence:  Types, History, and Future,” simplilearn.com; “The Brief History of Artificial Intelligence,” ourworldindata.org; “Why AI will never replace humans,” antino.io/blog; “AI Won’t Replace Human Intuition,” and “The Future of AI:  5 Things to Expect in the Next 10 years,” forbes.com; “What is Artificial General Intelligence,” techtarget.com; “The Future of AI:  How Artificial Intelligence Will Change the World,” builtin.com; “The four biggest challenges in brain simulation,” nature.com; and numerous other online sources.

Introduction to AI and How it Works

AI is the science that allows machines and computer applications to mimic human intelligence by modeling human behavior so that it can use human-like thinking processes to solve complex problems.

AI is accomplished by studying the patterns of the human brain and by analyzing the cognitive process.  The outcome of these studies enables intelligent software and AI systems.

Born in the 1950s, the science of AI has progressed irregularly, due to both technology limitations and periodic funding restraints. 

Two basic goals of AI have emerged: “narrow” AI and “full” AI.  The goal of narrow AI is to solve specific tasks or problems, often repetitive, time-consuming jobs that normally require human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages.  The goal of full AI, called Artificial General Intelligence (AGI), is to achieve generalized human cognitive abilities in software so that, faced with an unfamiliar task, the AGI system could find a solution.  The intention of an AGI system is to be able to perform any task that a human being can.  Some researchers extend this goal to computer programs that experience sentience or consciousness. 

The first generation of AI researchers were convinced that AGI was possible and that it would exist in just a few decades.  However, by the 1970s and 1980s, it became obvious that researchers had grossly underestimated the difficulty of achieving AGI.  (See Artificial General Intelligence below.)

However, in the 1990s and early 21^st century, researchers achieved real progress by focusing on narrow AI, specific problems where they could produce verifiable results and commercial applications.  These "applied AI" systems are now used extensively throughout industry, with applications found in E-commerce, Education, Internet Operations, Road and Air Vehicles, Healthcare, Marketing, Finance, Entertainment, and more.  (See Today’s Applications of Narrow AI below.)

Today’s AI systems work by repeatedly analyzing large sets of data to learn from patterns and features in the data that they analyze.   Each time an AI system runs a round of data processing, it tests and measures its own performance and develops additional expertise.

Because AI never needs a break, it can run through hundreds, thousands, or even millions of tasks extremely quickly, learning a great deal in very little time, and becoming extremely capable at whatever it’s being trained to accomplish.

To solve these problems, AI researchers have adapted and integrated a wide range of problem-solving techniques - including search and mathematical optimization, formal logic, artificial neural networks, and methods based on statistics, probability, and economics. AI also draws upon computer science, psychology, linguistics, philosophy, and many other fields.  AI isn’t just a single computer program or application, but an entire discipline, or science.

Components of Today’s AI Systems

There are many different sub-fields of the overarching science of today’s artificial intelligence.  Each of the following components is commonly utilized by AI technology today:

Machine Learning.  Allows AI systems to learn automatically and develop better results based on experience, all without being programmed to do so.  Machine Learning allows AI to find patterns in data, uncover insights, and improve the results of whatever task the system has been set out to achieve.

Deep Learning.  A specific type of machine learning that allows AI to learn and improve by processing data.  Deep Learning uses artificial neural networks which mimic biological neural networks in the human brain to process information, find connections between the data, and come up with inferences, or results based on positive and negative reinforcement.

Neural Networks.  Operate like networks of neurons in the human brain, allowing AI systems to take in large data sets, uncover patterns amongst the data, and answer questions about it.

Cognitive Computing.  Imitates the interactions between humans and machines, allowing computer models to mimic the way that a human brain works when performing a complex task, like analyzing text, speech, or images.

Natural Language Processing.  Allows computers to recognize, analyze, interpret, and truly understand human language, either written or spoken.  Natural Language Processing is critical for any AI-driven system that interacts with humans in some way, either via text or spoken inputs.

Computer Vision.  Interprets the content of an image via pattern recognition and deep learning, and lets AI systems identify specific objects in visual data.

History of AI

Precursors.  From ancient times, various mathematicians, theologians, philosophers, professors, and authors mused about mechanical techniques, calculating machines, and numeral systems that eventually led to the concept of mechanized “human” thought in non-human beings.

Depictions of all-knowing machines akin to computers were more widely discussed in popular literature starting in the early 1700s.  Jonathan Swift’s novel Gulliver’s Travels mentioned a device called the engine, which is one of the earliest references to modern-day technology, specifically a computer.  This device’s intended purpose was to improve knowledge and mechanical operations to a point where even the least talented person would seem to be skilled - all with the assistance and knowledge of a non-human mind (mimicking artificial intelligence). 

In 1921, Karel Čapek, a Czech playwright, released his science fiction play “Rossum’s Universal Robots.” His play explored the concept of factory-made artificial people who he called robots - the first known reference to the word.  From this point onward, people took the “robot” idea and implemented it into their research, art, and discoveries.

In 1927, the sci-fi film Metropolis featured a robotic girl who was physically indistinguishable from the human counterpart from which it took its likeness.  This film is significant because it is the first on-screen depiction of a robot, and thus lent inspiration to other famous non-human characters such as C-P30 in the movie Star Wars.

Robotic girl from the 1927 film Metropolis.

Note:  The principle of the modern computer was proposed by the British polymath Alan Turing in his seminal 1936 paper, “On Computable Numbers.”  The first digital electronic calculating machines were developed during World War II, and used to break German wartime codes.  After World War II, computers rapidly improved.

By the 1950s, we had a generation of scientists, mathematicians, and philosophers with the concept of computers, AI, and intelligent robots culturally assimilated in their minds.  Early on, Alan Turing explored the mathematical possibility of artificial intelligence. Turing suggested that humans use available information as well as reason in order to solve problems and make decisions, so why can’t machines do the same thing? This was the logical framework of his 1950 paper, “Computing Machinery and Intelligence” in which he discussed how to build intelligent machines and how to test their intelligence.

Alan Turing was instrumental in the development of digital computers and artificial intelligence.

 

Birth of AI.  Five years later, the proof of concept was initialized through Allen Newell, Cliff Shaw, and Herbert Simon’s, Logic Theorist.  The Logic Theorist was a computer program designed to mimic the problem-solving skills of a human and was funded by Research and Development Corporation.  It’s considered by many to be the first artificial intelligence program, and was presented at the Dartmouth Summer Research Project on Artificial Intelligence, hosted by John McCarthy and Marvin Minsky in 1956. 

In this historic conference, McCarthy brought together top researchers from various fields for an open-ended discussion on artificial intelligence, the term which he coined at the very event.  Everyone whole-heartedly aligned with the sentiment that AI was achievable. This event catalyzed the next twenty years of AI research.

Founding Fathers of AI.

 

Progression of Narrow AI.  From 1957 to 1974, narrow AI flourished.  Computers could store more information and became faster, cheaper, and more accessible.  Machine learning algorithms (a set of rules to be followed) also improved, and people got better at knowing which algorithm to apply to their problem.  

Early demonstrations such as Newell and Simon’s General Problem Solver and Joseph Weizenbaum’s ELIZA showed promise toward the goals of problem solving and the interpretation of spoken language respectively.  These successes, as well as the advocacy of leading researchers, convinced government agencies such as the Defense Advanced Research Projects Agency to fund AI research at several institutions.  The government was particularly interested in a machine that could transcribe and translate spoken language as well as high throughput data processing.

By the middle of the 1960s, research in the U.S. was heavily funded by the Department of Defense, and laboratories had been established around the world.

The biggest obstacle was the lack of computational power to do anything substantial: computers simply couldn’t store enough information or process it fast enough.  In order to communicate, for example, one needs to know the meanings of many words and understand them in many combinations.  Hans Moravec, a doctoral student of McCarthy at the time, stated that “computers were still millions of times too weak to exhibit intelligence.”  In 1974, as patience dwindled, so did the funding, and research slowed for ten years.

In the 1980’s, AI was reignited by two sources: an expansion of the algorithmic toolkit, and a boost of funds.  John Hopfield and David Rumelhart popularized “deep learning” techniques which allowed computers to learn using experience.  Edward Feigenbaum introduced expert systems which mimicked the decision-making process of a human expert.  The program would ask an expert in a field how to respond in a given situation, and once this was learned for virtually every situation, non-experts could receive advice from that program.  Expert systems were widely used in industries. 

The Japanese government heavily funded expert systems and other AI related endeavors as part of their Fifth Generation Computer Project (FGCP).  From 1982-1990, they invested $400 million dollars with the goals of revolutionizing computer processing, implementing logic programming, and improving artificial intelligence. Unfortunately, most of the ambitious goals were not met.  However, it could be argued that the indirect effects of the FGCP inspired a talented young generation of engineers and scientists.  Regardless, funding of the FGCP ceased, and AI fell out of the limelight again.

Ironically, in the absence of government funding and public hype, AI thrived.  During the 1990s and 2000s, many of the landmark goals of narrow artificial intelligence were achieved.  In 1997, reigning world chess champion and grand master Gary Kasparov was defeated by IBM’s Deep Blue, a chess playing computer program.  This highly publicized match was the first time a reigning world chess champion lost to a computer and served as a huge step towards an artificially intelligent decision-making program.  In the same year, speech recognition software, developed by Dragon Systems, was implemented on Windows.  This was a great step forward in the direction of the spoken language interpretation. 

Faster computers, algorithmic improvements, and access to large amounts of data enabled advances in machine learning and perception.  In a 2017 survey, one in five companies reported they had "incorporated AI in some offerings or processes.” The amount of research into AI (measured by total publications) increased by 50% in the years 2015-2019.

The language and image recognition capabilities of AI systems developed very rapidly.  The chart below shows how we got here by zooming into the last two decades of AI development.  The plotted data stems from a number of tests in which human and AI performance were evaluated in five different domains, from handwriting recognition to language understanding. 

 

Within each of the five domains, the initial performance of the AI system is set to -100, and human performance in these tests is used as a baseline that is set to zero.  This means that when the model’s performance crosses the zero line is when the AI system scored more points in the relevant test than the humans who did the same test.  AI systems have become steadily more capable and are now beating humans in tests in all these domains. 

Today’s Applications of Narrow AI

Artificial intelligence is no longer a technology of the future; narrow AI is here, and much of what is reality now would have looked like sci-fi just recently.  There’s virtually no major industry that modern narrow AI hasn’t already affected.  Some sectors are at the start of their AI journey, others are veteran travelers.  Both have a long way to go.  Regardless, the impact AI is having on our present day lives is hard to ignore.

Here is a partial list of AI applications that impact all of us today.

1. E-Commerce

Personalized Shopping:  AI creates recommendations to engage better with customers.  These recommendations are made in accordance with the customer’s browsing history, preference, and interests.  It helps in improving relationships with customers and their loyalty towards the seller.

AI-powered Assistants:  Virtual shopping assistants and chatbots, designed to simulate conversation with human users, help improve the user experience while shopping online.  Natural Language Processing makes the conversation sound as human and personal as possible.  Moreover, these assistants can have real-time engagement with customers.

AI-powered virtual shopping assistants improve the user experience.

 

Fraud Prevention:  Credit card frauds and fake reviews are two of the most significant issues that E-Commerce companies deal with.  By considering usage patterns, AI can help reduce the possibility of credit card frauds taking place.  Many customers prefer to buy a product or service based on customer reviews.  AI can help identify and handle fake reviews. 

Manufacturing:  AI solutions help forecast load and demand for factories, improving their efficiency, and allow factory managers to make better decisions about ordering materials, completion timetables, and other logistics issues.

Human Resources:  AI helps with blind hiring.  Machine learning software can scan job candidates' profiles and resumes to provide recruiters an understanding of the talent pool they must choose from.  

Retail:  AI systems are being consulted to design more effective store layouts and handle stock management.

2. Education

Administrative Tasks:  AI helps educators with tasks like facilitating and automating personalized messages to students, back-office tasks like grading paperwork, arranging and facilitating parent and guardian interactions, routine issue feedback, managing enrollment, courses, and HR-related topics. 

Smart Content:  AI helps digitize content like video lectures, conferences, and text book guides.  We can apply different interfaces like animations and learning content through customization for students from different grades.  AI helps create a rich learning experience by generating and providing audio and video summaries and integral lesson plans.  Without even the direct involvement of the lecturer or the teacher, a student can access extra learning material or assistance through Voice Assistants and also provide answers to very common questions easily.

Personalized Learning:  AI can monitor students’ data thoroughly, and habits, lesson plans, reminders, study guides, flash notes, frequency or revision, etc., can be easily generated.

3. Internet Operations

Spam Filters:  The email that we use in our day-to-day lives has AI that filters out spam emails sending them to spam or trash folders, letting us see the filtered content only.

Facial Recognition:  Our favorite devices like our phones, laptops, and PCs use facial recognition techniques by using face filters to detect and identify in order to provide secure access.  Apart from personal usage, facial recognition is a widely used AI application even in high security-related areas in several industries.

Internet Searches:  Without the help of AI, search engines like Google would not be able to deliver relevant and timely information to drive countless daily decisions.  AI figures out what search results you will see and what related topics may be relevant to help you get the “right” answers.

Chatbots:  AI chatbots respond to people online who use the "live chat" feature that many organizations provide for customer service.  AI chatbots are effective with the use of machine learning, and can be integrated in an array of websites and applications.

Voice Assistants:  Virtual assistants like Siri and Google Assistant use voice queries, gesture-based control (human body language), focus-tracking, and a natural-language user interface to answer questions, make recommendations, and perform actions by delegating requests to a set of Internet services.  With continued use, they adapt to users' individual language usages, searches, and preferences, returning individualized results.

 

The Siri voice assistant operates from today’s smart phones.

Social Media:  On Instagram, AI considers your likes and the accounts you follow to determine what posts you are shown on your explore tab.  AI helps Facebook understand conversations better.  It can be used to translate posts from different languages automatically.  AI is used by Twitter for fraud detection, removing propaganda, and hateful content.  Twitter also uses AI to recommend tweets that users might enjoy, based on what type of tweets they engage with.

Recommendation Systems:  Various platforms that we use in our daily lives like E-commerce, entertainment websites, social media, video sharing platforms, like YouTube, etc., all use a recommendation system to get user data and provide customized recommendations to users to increase engagement.  This is a very widely used AI application in almost all industries.

4. Road Vehicle Operation

Safety:  GPS technology provides users with accurate, timely, and detailed information to improve safety.  AI neural networks automatically detect the number of lanes and road types behind obstructions on the roads. 

Driving Efficiency:  AI is heavily used by Uber and many logistics companies to improve operational efficiency, analyze road traffic, and optimize routes.  Through mapping applications, AI has streamlined the way we plan for and think about car travel.   AI enables smart traffic lights to improve traffic control.

In-Vehicle AI:  AI improves the in-vehicle experience and provides additional systems like emergency braking, blind-spot monitoring, and driver-assist steering.

Autonomous Vehicles:  Automobile manufacturing companies like Toyota, Audi, Volvo, and Tesla use machine learning to train computers to think and evolve like humans when it comes to driving in any environment and object detection to avoid accidents.

AI-enabled autonomous vehicles are in testing today.

 

5.  Air Travel

Safety:  AI brings valuable data and real-time information to pilots so they can use their skills to make the best decisions possible, particularly in critical and potentially life-saving situations.  Since the initial implementation of these sensors - and newer technologies like wind shear and microburst detection - air travel has never been safer. The availability of data via AI enables pilots to be better prepared and significantly reduces weather-related issues.

Efficiency:  When you book a flight, it is often an AI system, and no longer a human, that decides what you pay.  When you get to the airport, it is an AI system that monitors what you do at the airport.  And once you are on the plane, an AI system assists the pilot in flying you to your destination. 

6. Robotics

AI-assisted robots are already a mainstay in automobile production.  Robotics can also be used for carrying goods in hospitals, factories, and warehouses, cleaning offices and large equipment, inventory management, and cooking food.

Today’s automobile assembly lines employ AI-assisted robotic machines.

 

7. Healthcare

Diagnosis:  AI quickly accesses and examines thousands of medical records, pulling relevant information like preexisting conditions, drug interactions, or Covid status, for example, to guide important diagnoses and treatment plans that will keep patients safe.

Smart Machines:  AI helps build sophisticated machines that can detect diseases and identify cancer cells - and do it faster with no less accuracy.  AI can help analyze chronic conditions with lab and other medical data to ensure early diagnosis. 

Personalized Medicine:  AI systems are trained to provide personalized medicine, including giving reminders about when patients need to take their medicine and suggestions for specific exercises patients should perform to improve their recovery from injuries. 

New Drugs:  AI also uses the combination of historical data and medical intelligence for the discovery of new drugs.

8. Agriculture

AI identifies defects and nutrient deficiencies in the soil, using computer vision, robotics, and machine learning applications.  AI can analyze where weeds are growing. AI bots can help to harvest crops at a higher volume and faster pace than human laborers.

9. Marketing

Using AI, marketers can deliver highly targeted and personalized ads with the help of behavioral analysis and pattern recognition.  It also helps with retargeting audiences at the right time to ensure better results and reduced feelings of distrust and annoyance.  AI can be used to edit and optimize marketing campaigns to fit a local market's needs.  AI can also be used to handle routine tasks like performance, and campaign reports.  

10. Finance 

Tools:  AI tools detect and prevent fraudulent financial transactions, provide more accurate assessments than traditional credit scores can, and automate all sorts of data-related tasks that were handled manually.  AI can also better predict and assess loan risks.

Online Banking:  Some people no longer use brick-and-mortar banks at all, conducting all their business online or via an app.  AI completes mobile check deposits, checks account balances, and enables bill pay.

 

AI-enabled online banking adds great flexibility to personal finance management.

11.  Weapon Systems

Several governments are developing AI-enabled autonomous weapons systems that can search out targets, decide to engage, and attack and destroy the target - completely without human involvement.   Not only will these killer robots become more intelligent, more precise, faster, and cheaper; they will also learn new capabilities, such as how to form swarms with teamwork and redundancy, making their missions virtually unstoppable.

Other military applications for AI include wargaming and battle strategy development, reconnaissance, and defense suppression.

12.  Entertainment

Movies:  AI helps with scriptwriting, pre-production (planning and scheduling), formulating release strategies, predicting success at the box office, casting, promotion, and creating spectacular visual effects.  It’s also become increasingly popular for video editing, coloring, and music creation.

Streaming in Real-Time:  AI aids in the personalization, packaging, and transmission of content in real-time, enhancing the viewer’s experience. It also helps to increase ad sales by allowing for tailored ad insertions. Live sports event ad earnings are maximized with digital billboard replacement options.

Gaming:  AI can be used to create smart, human-like NPCs (nonplayer characters) to interact with the players.  It can also be used to predict human behavior to improve game design and testing.

AI-enabled online gaming is increasing in popularity.

 

In addition to the applications listed above, AI systems also increasingly determine whether you get a loan, are eligible for welfare, or get hired for a particular job.  Increasingly they even help determine who gets released from jail.

Other applications predict the result of judicial decisions, create art (such as poetry or painting), and prove mathematical theorems.

Future of Narrow AI

Artificial intelligence is shaping the future of nearly every industry and it will continue to act as a technological innovator for the foreseeable future.

With companies spending billions of dollars on narrow AI products and services annually, tech giants like Google, Apple, Microsoft and Amazon spending billions to create those products and services, universities making AI a more prominent part of their curricula, and the U.S. Department of Defense upping its AI game, big things are bound to happen.  Some of those developments are well on their way to being fully realized; some are merely theoretical and might remain so. 

Note:  Some of the optimism regarding future narrow AI development is associated with Moore’s Law which predicts that the speed and memory capacity of computers doubles every two years as a result of the number of transistor components doubling every two years.  The fundamental problem of "raw computer power" is slowly being overcome.  (The observation is named after Gordon Moore, the co-founder of Fairchild Semiconductor and Intel.)

Predictions.  In addition to enabling continued vast efficiency and capability improvements in the industry applications discussed earlier and others, AI is poised to fundamentally restructure broader swaths of our economy and society over the next decade.  Here are four predictions from Gaurav Tewari, Founder and Managing Partner of Omega Venture Partners technology investment firm.

1. AI will transform the scientific method.

Important science - think large-scale clinical trials or building particle colliders - is expensive and time-consuming. In recent decades there has been considerable, well-deserved concern about scientific progress slowing down.  Scientists may no longer be experiencing the golden age of discovery.

With AI, we can expect to see orders of magnitude of improvement in what can be accomplished.  AI enables an unprecedented ability to analyze enormous data sets and computationally discover complex relationships and patterns.  AI, augmenting human intelligence, is primed to transform the scientific research process, unleashing a new golden age of scientific discovery in the coming years.

2. AI will become a pillar of foreign policy.

We are likely to see serious government investment in AI.  U.S. Secretary of Defense Lloyd J. Austin III has publicly embraced the importance of partnering with innovative AI technology companies to maintain and strengthen global U.S. competitiveness.

The National Security Commission on Artificial Intelligence has created detailed recommendations, concluding that the U.S. government needs to greatly accelerate AI innovation.  There’s little doubt that AI will be imperative to the continuing economic resilience and geopolitical leadership of the United States.

3. Addressing climate will require AI.

We are currently working to mitigate the socioeconomic threats posed by climate change.  Many promising emerging ideas require AI to be feasible.  One potential new approach involves analyzing the relationship of environmental policy to impacts.  This would likely be powered by digital Earth simulations that would require staggering amounts of real-time data and computation to detect nuanced trends imperceptible to human senses.  Other new technologies such as carbon dioxide sequestration (capturing and storing atmospheric carbon dioxide) cannot succeed without AI-powered risk modeling, downstream effect prediction, and the ability to anticipate unintended consequences

AI is poised to have a major effect on climate change and environmental issues.  Ideally, and partly through the use of sophisticated sensors, cities will become less congested, less polluted, and generally more livable. 

4. AI will enable truly personalized medicine.

One compelling emerging application of AI involves synthesizing individualized therapies for patients. Moreover, AI has the potential to one day synthesize and predict personalized treatment options in near real-time - no clinical trials required.

AI is uniquely suited to construct and analyze individual biology’s and is able to do so in the context of the communities an individual lives in.  The human body is mind-boggling in its complexity, and it is shocking how little we know about how drugs work.  Without AI, it is impossible to make sense of the massive datasets from an individual’s physiology, let alone the effects on individual health outcomes from environment, lifestyle, and diet.   

Issues.  While narrow AI is expected to produce great benefits for mankind in the future, there are several issues that need to be considered.

1.       Job Displacement.

Many people believe that AI will supplant humans in various ways.  Oxford University’s Future of Humanity Institute published the results of a 2017 AI survey, “When Will AI Exceed Human Performance? Evidence from AI Experts.”  It contains estimates from 352 machine learning researchers about AI’s evolution in years to come.  By 2026, a median number of respondents said, machines will be capable of writing school essays; by 2027, self-driving trucks will render drivers unnecessary; by 2031, AI will outperform humans in the retail sector; by 2049, AI could author a best-selling book, and by 2053, AI could be the next neurosurgeon.  The researchers believed that there is a 50% chance of AI outperforming humans in all tasks in 45 years, and of automating all human jobs in 120 years.

Note:  One recent development is far ahead of its predicted availability.  ChatGPT (Generative Pre-trained Transformer) was launched as a chatbot prototype by OpenAI on November 30, 2022, and quickly garnered attention for its detailed responses and articulate answers across many domains of knowledge.  The chatbot - which cannot think for itself, but is trained to generate conversational text - can be used for a wide array of applications, from writing college-level essays and poetry in a matter of seconds to composing computer code and legal contracts, or for more playful uses such as writing wedding speeches, hip-hop lyrics, or comedy routines.  It’s already abundantly clear that ChatGPT has far-ranging implications and potential uses for education, entertainment, research and especially our workforce.

Others argue that while narrow artificial intelligence is designed to replace manual labor with a more effective and quicker way of doing work, it cannot override the need for human input in the workspace.  Narrow AI systems lack sensory perception, natural language understanding, social and emotional engagement, and untrained problem-solving skills.  Good businesses recognize that these capabilities and skills are needed in the customer-relation market place of today and the future.

The World Economic Forum suggests that while machines with AI will replace about 85 million jobs by 2025, about 97 million jobs will be made available by the same year thanks to AI. So, the big question is: How can humans work with AI, instead of being replaced by it?

These results should inform discussion amongst researchers and policymakers about anticipating and managing trends in AI.  One of the absolute prerequisites for AI to be successful in many areas, is that we invest tremendously in education to retrain people for new jobs.

2.       Weaponized AI.

AI provides a number of tools that are particularly useful for authoritarian governments to employ cybercrime and terrorism: smart spyware, facial recognition, and voice recognition allow widespread surveillance; such surveillance allows machine learning to classify potential enemies of the state and can prevent them from hiding; recommendation systems can precisely target propaganda and disinformation        for maximum effect.  Applications such as the recently-introduced ChatGPT could even create disinformation, and become a tool for hackers and phishing schemes.

AI-assisted autonomous weapons are already a clear and present danger, and will become more intelligent, nimble, lethal, and accessible at an alarming speed.  The deployment of autonomous weapons will be accelerated by an inevitable arms race that will lack the natural deterrence of nuclear weapons.  By 2015, over 50 countries were reported to be researching battlefield robots.  Cybersecurity systems and elections are also potentially vulnerable to bad actors employing AI.   It is not at all clear how we can control this lethal threat to humanity.

Turkish autonomous attack drones.

 

3.       Privacy. 

AI’s reliance on huge data bases (called Big Data today) is already impacting privacy in a major way.  Look no further than Amazon’s Alexa eavesdropping, just one example of tech gone wild.  Search engines and social media platforms have been accused of greed-driven data mining.  Without proper regulations and self-imposed limitations, critics argue, the situation will get even worse. 

Artificial General Intelligence

Very little has been accomplished so far to meet the AGI goal of general human intelligence, the ability of an intelligent agent to understand or learn any intellectual task that a human being can.   The same can be said about the ultimate goal of achieving computer programs that experience sentience or consciousness.  AGI’s future is highly speculative.

History of AGI.  The term "artificial general intelligence" was used as early as 1997.   By 2010, AGI research had been founded as a separate sub-field, and there were academic conferences, laboratories, and university courses dedicated to AGI research, as well as private consortiums and new companies.

Today, most AI researchers have devoted little attention to AGI, with some claiming that intelligence is too complex to be completely replicated.  However, a small number of computer scientists are still active in AGI research.  For those who believe that AGI goals can be achieved, estimates of the time required to achieve success range widely from ten years to over a century.

Here are a few achievements to date that at least inspire some optimism in AGI researchers:

In 2005, the Human Brain Project was started by a European research group hoping to recreate a complete human brain inside a computer, with electronic circuits in the computer emulating neural networks in the brain - a digital mind, composed of computer code, complete with a sense of self consciousness and memory.  The researchers thought that within a few decades, we could have an AGI system that could talk and behave very much as a human does.  But little real progress was made toward this goal, and in 2013, the project was rebranded with a new less ambitious goal of “putting in place a cutting-edge research infrastructure that will allow scientific and industrial researchers to advance our knowledge in the fields of neuroscience, computing, and brain-related medicine.”

In 2016, a humanoid robot named Sophia was created by Hanson Robotics.  She is known as the first “robot citizen.”  What distinguishes Sophia from previous humanoids is her likeness to an actual human being, with her ability to see (image recognition), make facial expressions, and communicate through AI.

In 2016, the humanized robot Sophia was introduced by Hanson Robotics.

 

In 2022, DeepMind developed Gato, a "general-purpose" system trained on 604 tasks including playing Atari games, accurately captioning images, chatting naturally with a human, and stacking colored blocks with a robot arm.  According to DeepMind, Gato would be better than human experts in 450 of the 604 tasks it has been trained for.

And, as mentioned earlier, OpenAI introduced ChatGPT in late 2022.  There is consensus that ChatGPT is not an example of AGI, but it is considered by some to be too advanced to classify as a narrow AI system.

Future of AGI.  Optimists still predict that AGI will improve at an exponential rate, leading to breakthroughs that enable AGI systems to operate at levels beyond human comprehension and control.

But human brain simulation efforts over the last almost two decades do not look promising.   

The human brain contains one hundred billion neurons (basic working unit of the brain, a specialized cell designed to transmit information to other nerve cells, muscle, or gland cells) and one thousand trillion synapses (junctions that transmits signals between neurons), all working in parallel.  Producing a biologically faithful simulation of the brain would require an almost limitless set of parameters, including the brain’s extracellular interactions, and molecular-scale processes.  There are no known solutions to these problems of scale and complexity.  Some aspects of mind, such as understanding, agency (control over voluntary actions and the outcomes of those actions), and consciousness, might never be captured by digital brain simulations.

Simulation of the human brain is an extremely complex task.

 

Simply stated, AGI is a very complicated problem and there are no clear paths to a solution.  The hoped for “breakthroughs” are unknown to researchers today.  Thus, many experts are skeptical that AGI will ever be possible.

Others question whether achieving full AGI is even desirable.  More than a few leading AI figures subscribe to a nightmare scenario, whereby superintelligent machines take over and permanently alter human existence through enslavement or eradication.

English theoretical physicist, cosmologist, and author Stephen Hawking warned of the dangers in a 2014 interview with the British Broadcasting Corp.  "The development of full artificial intelligence could spell the end of the human race," he said. "It would take off on its own and redesign itself at an ever-increasing rate.  Humans, who are limited by slow biological evolution, couldn't compete and would be superseded."

The slow pace of AGI development may actually be a blessing.  One expert opines, “Time to understand what we’re creating and how we’re going to incorporate it into society, might be exactly what we need.”

 

In the meantime, we’ll have to get along with our favorite fictional intelligent robots, C3PO and R2D2.