An artificial general intelligence AGI is a type of hypothetical intelligent agent.
The AGI concept is that it can learn to accomplish any intellectual task that human beings or animals can perform.
Alternatively AGI has been defined as an autonomous system that surpasses human capabilities in the majority of economically valuable tasks.
Creating AGI is a primary goal of some artificial intelligence research and companies such as OpenAI DeepMind and Anthropic.
The timeline for AGI development remains a subject of ongoing debate among researchers and experts.
Some argue that it may be possible in years or decades others maintain it might take a century or longer and a minority believe it may never be achieved.
Additionally there is debate regarding whether modern deep learning systems such as GPT 4 are an early yet incomplete form of AGI or if new approaches are required.
Contention exists over the potential for AGI to pose a threat to humanity for example OpenAI treats it as an existential risk while others find the development of AGI to be too remote to present a risk.
A 2020 survey identified 72 active AGI projects spread across 37 countries.
Terminology AGI is also known as strong AI full AI or general intelligent action.
However some academic sources reserve the term strong AI for computer programs that experience sentience or consciousness.
In contrast weak AI or narrow AI is able to solve one specific problem but lacks general cognitive abilities.
Some academic sources use weak AI to refer more broadly to any programs that neither experience consciousness nor have a mind in the same sense as humans.
Related concepts include human level AI transformative AI and superintelligence.
Characteristics Various criteria for intelligence have been proposed most famously the Turing test but no definition is broadly accepted.
Intelligence traits However researchers generally hold that intelligence is required to do the following reason use strategy solve puzzles and make judgments under uncertainty represent knowledge including common sense knowledge plan learn communicate in natural language and if necessary integrate these skills in completion of any given goal.
Many interdisciplinary approaches e g cognitive science computational intelligence and decision making consider additional traits such as imagination the ability to form novel mental images and concepts and autonomy.
Computer based systems that exhibit many of these capabilities exist e g see computational creativity automated reasoning decision support system robot evolutionary computation intelligent agent.
However no consensus holds that modern AI systems possess them to an adequate degree.
Physical traits Other capabilities are considered desirable in intelligent systems as they may affect intelligence or aid in its expression.
These include the ability to sense e g see hear etc and the ability to act e g move and manipulate objects change location to explore etc This includes the ability to detect and respond to hazard.
Mathematical formalisms A mathematically precise specification of AGI was proposed by Marcus Hutter in 2000.
This type of AGI characterized by the ability to maximise a mathematical definition of intelligence rather than exhibit human like behaviour is also called universal artificial intelligence.
This problem stems from AIXI s use of compression as a proxy for intelligence which is only valid if cognition takes place in isolation from the environment in which goals are pursued.
This formalises a philosophical position known as dualism.
Some find enactivism more notion that cognition takes place within the same environment in which goals are pursued.
Subsequently Michael Timothy Bennett formalised enactive cognition and identified an alternative proxy for intelligence called weakness.
The accompanying experiments comparing weakness and compression and mathematical proofs showed that maximising weakness results in the optimal ability to complete a wide range of tasks or equivalently ability to generalise thus maximising intelligence by either definition.
If enactivism holds and dualism does not then compression is not necessary or sufficient for intelligence calling into question widely held views on intelligence see also Hutter Prize.
Whether an AGI that satisfies one of these formalizations exhibits human like behaviour such as the use of natural language would depend on many factors for example the manner in which the agent is embodied or whether it has a reward function that closely approximates human primitives of cognition like hunger pain and so forth.
Tests for testing human level AGI Several tests meant to confirm human level AGI have been considered including The Turing Test Turing A machine and a human both converse unseen with a second human who must evaluate which of the two is the machine which passes the test if it can fool the evaluator a significant fraction of the time.
The AI Eugene Goostman achieved Turing s estimate of convincing 30 % of judges that it was human in 2014.
The Robot College Student Test Goertzel A machine enrolls in a university taking and passing the same classes that humans would and obtaining a degree.
The Ikea test Marcus Also known as the Flat Pack Furniture Test.
An AI views the parts and instructions of an Ikea flat pack product then controls a robot to assemble the furniture correctly.
A machine is required to enter an average American home and figure out how to make coffee find the coffee machine find the coffee add water find a mug and brew the coffee by pushing the proper buttons.
This has not yet been completed.
For example even specific straightforward tasks like machine translation require that a machine read and write in both languages NLP follow the author s argument reason know what is being talked about knowledge and faithfully reproduce the author s original intent social intelligence.
All of these problems need to be solved simultaneously in order to reach human level machine performance.
A problem is informally called AI complete or AI hard if it is believed that to solve it one would need to implement strong AI because the solution is beyond the capabilities of a purpose specific algorithm.
This limitation could be useful to test for the presence of humans as CAPTCHAs aim to do and for computer security to repel brute force attacks.
History Classical AI Modern AI research began in the mid 1950s.
The first generation of AI researchers were convinced that artificial general intelligence was possible and that it would exist in just a few decades.
Their predictions were the inspiration for Stanley Kubrick and Arthur C Clarke s character HAL 9000 who embodied what AI researchers believed they could create by the year.
He said in 1967 Within a generation.
However in the early 1970s it became obvious that researchers had grossly underestimated the difficulty of the project.
In the early 1980s Japan s Fifth Generation Computer Project revived interest in AGI setting out a ten year timeline that included AGI goals like carry on a casual conversation.
In response to this and the success of expert systems both industry and government pumped money back into the field.
However confidence in AI spectacularly collapsed in the late 1980s and the goals of the Fifth Generation Computer Project were never fulfilled.
For the second time in 20 years AI researchers who predicted the imminent achievement of AGI had been mistaken.
By the 1990s AI researchers had a reputation for making vain promises.
Narrow AI research In the 1990s and early 21st century mainstream AI achieved commercial success and academic respectability by focusing on specific sub problems where AI can produce verifiable results and commercial applications such as artificial neural networks and statistical machine learning.
These applied AI systems are now used extensively throughout the technology industry and research in this vein is heavily funded in both academia and industry.
As of 2018 development on this field was considered an emerging trend and a mature stage was expected to happen in more than 10 years.
Most mainstream AI researchers hope that strong AI can be developed by combining programs that solve various sub problems.
Hans Moravec wrote in 1988 I am confident that this bottom up route to artificial intelligence will one day meet the traditional top down route more than half way ready to provide the real world competence and the commonsense knowledge that has been so frustratingly elusive in reasoning programs.
Fully intelligent machines will result when the metaphorical golden spike is driven uniting the two efforts.
For example Stevan Harnad of Princeton University concluded his 1990 paper on the Symbol Grounding Hypothesis by stating The expectation has often been voiced that top down symbolic approaches to modeling cognition will somehow meet bottom up sensory approaches somewhere in between.
A free floating symbolic level like the software level of a computer will never be reached by this route or vice versa nor is it clear why we should even try to reach such a level since it looks as if getting there would just amount to uprooting our symbols from their intrinsic meanings thereby merely reducing ourselves to the functional equivalent of a programmable computer.
Modern artificial general intelligence research The term artificial general intelligence was used as early as 1997 by Mark Gubrud in a discussion of the implications of fully automated military production and operations.
The term was re introduced and popularized by Shane Legg and Ben Goertzel around.
The first summer school in AGI was organized in Xiamen China in 2009 by the Xiamen university s Artificial Brain Laboratory and OpenCog.
The first university course was given in 2010 and 2011 at Plovdiv University Bulgaria by Todor Arnaudov.
As of 2023 a small number of computer scientists are active in AGI research and many contribute to a series of AGI conferences.
Although most open ended learning works are still done on Minecraft its application can be extended to robotics and the sciences.
Feasibility As of 2022 AGI remains speculative.
No such system has yet been demonstrated.
Opinions vary both on whether and when artificial general intelligence will arrive.
This prediction failed to come true.
Microsoft co founder Paul Allen believed that such intelligence is unlikely in the 21st century because it would require unforeseeable and fundamentally unpredictable breakthroughs and a scientifically deep understanding of cognition.
Writing in The Guardian roboticist Alan Winfield claimed the gulf between modern computing and human level artificial intelligence is as wide as the gulf between current space flight and practical faster than light spaceflight.
John McCarthy is among those who believe human level AI will be accomplished but that the present level of progress is such that a date cannot accurately be predicted.
Four polls conducted in 2012 and 2013 suggested that the median guess among experts for when they would be 50 % confident AGI would arrive was 2040 to 2050 depending on the poll with the mean being 2081 Of the experts 16 5 % answered with never when asked the same question but with a 90 % confidence instead.
A report by Stuart Armstrong and Kaj Sotala of the Machine Intelligence Research Institute found that over a 60 year time frame there is a strong bias towards predicting the arrival of human level AI as between 15 and 25 years from the time the prediction was made.
They analyzed 95 predictions made between 1950 and 2012 on when human level AI will come about.
Timescales In the introduction to his 2006 book Goertzel says that estimates of the time needed before a truly flexible AGI is built vary from 10 years to over a century.
As of 2007 the consensus in the AGI research community seemed to be that the timeline discussed by Ray Kurzweil in The Singularity is Near i e between 2015 and 2045 was plausible.
Mainstream AI researchers have given a wide range of opinions on whether progress will be this rapid.
A 2012 meta analysis of 95 such opinions found a bias towards predicting that the onset of AGI would occur within years for modern and historical predictions alike.
That paper has been criticized for how it categorized opinions as expert or non expert.
In 2012 Alex Krizhevsky Ilya Sutskever and Geoffrey Hinton developed a neural network called AlexNet which won the ImageNet competition with a top 5 test error rate of 15 3 % significantly better than the second best entry s rate of 26 3 % the traditional approach used a weighted sum of scores from different pre defined classifiers.
In 2017 researchers Feng Liu Yong Shi and Ying Liu conducted intelligence tests on publicly available and freely accessible weak AI such as Google AI Apple s Siri and others.
At the maximum these AIs reached an IQ value of about 47 which corresponds approximately to a six year old child in first grade.
An adult comes to about 100 on average.
Similar tests were carried out in 2014 with the IQ score reaching a maximum value of 27 In 2020 OpenAI developed GPT 3 a language model capable of performing many diverse tasks without specific training.
According to Gary Grossman in a VentureBeat article while there is consensus that GPT 3 is not an example of AGI it is considered by some to be too advanced to classify as a narrow AI system.
In the same year Jason Rohrer used his GPT 3 account to develop a chatbot and provided a chatbot developing platform called Project December.
This research sparked a debate on whether GPT 4 could be considered an early incomplete version of artificial general intelligence emphasizing the need for further exploration and evaluation of such systems.
In 2023 the AI researcher Geoffrey Hinton stated that The idea that this stuff could actually get smarter than people a few people believed that.
I thought it was 30 to 50 years or even longer away.
Obviously I no longer think that.
Brain simulation Whole brain emulation One possible approach to achieving AGI is whole brain emulation A brain model is built by scanning and mapping a biological brain in detail and copying its state into a computer system or another computational device.
The computer runs a simulation model sufficiently faithful to the original that it behaves in practically the same way as the original brain.
Whole brain emulation is discussed in computational neuroscience and neuroinformatics in the context of brain simulation for medical research purposes.
Early estimates For low level brain simulation an extremely powerful computer would be required.
The human brain has a huge number of synapses.
Each of the 1011 one hundred billion neurons has on average 7 000 synaptic connections synapses to other neurons.
The brain of a three year old child has about 1015 synapses 1 quadrillion.
This number declines with age stabilizing by adulthood.
Estimates vary for an adult ranging from 1014 to synapses 100 to 500 trillion.
For comparison if a computation was equivalent to one floating point operation a measure used to rate current supercomputers then 1016 computations would be equivalent to 10 petaFLOPS achieved in 2011 while 1018 was achieved in 2022.
He used this figure to predict the necessary hardware would be available sometime between 2015 and 2025 if the exponential growth in computer power at the time of writing continued.
Modelling the neurons in more detail The artificial neuron model assumed by Kurzweil and used in many current artificial neural network implementations is simple compared with biological neurons.
A brain simulation would likely have to capture the detailed cellular behaviour of biological neurons presently understood only in broad outline.
The overhead introduced by full modeling of the biological chemical and physical details of neural behaviour especially on a molecular scale would require computational powers several orders of magnitude larger than Kurzweil s estimate.
In addition the estimates do not account for glial cells which are known to play a role in cognitive processes.
Current research Some research projects are investigating brain simulation using more sophisticated neural models implemented on conventional computing architectures.
It took 50 days on a cluster of 27 processors to simulate 1 second of a model.
A longer term goal is to build a detailed functional simulation of the physiological processes in the human brain It is not impossible to build a human brain and we can do it in 10 years Henry Markram director of the Blue Brain Project said in 2009 at the TED conference in Oxford.
Neuro silicon interfaces have been proposed as an alternative implementation strategy that may scale better.
Hans Moravec addressed the above arguments brains are more complicated neurons have to be modeled in more detail in his 1997 paper When will computer hardware match the human brain.
He measured the ability of existing software to simulate the functionality of neural tissue specifically the retina.
His results do not depend on the number of glial cells nor on what kinds of processing neurons perform where.
The actual complexity of modeling biological neurons has been explored in OpenWorm project that aimed at complete simulation of a worm that has only 302 neurons in its neural network among about 1000 cells in total.
The animal s neural network was well documented before the start of the project.
However although the task seemed simple at the beginning the models based on a generic neural network did not work.
Currently efforts focus on precise emulation of biological neurons partly on the molecular level but the result cannot yet be called a total success.
Criticisms of simulation based approaches A fundamental criticism of the simulated brain approach derives from embodied cognition theory which asserts that human embodiment is an essential aspect of human intelligence and is necessary to ground meaning.
If this theory is correct any fully functional brain model will need to encompass more than just the neurons e g a robotic body.
Goertzel proposes virtual embodiment like in Second Life as an option but it is unknown whether this would be sufficient.
Desktop computers using microprocessors capable of more than 109 cps Kurzweil s non standard unit computations per second see above have been available since.
According to the brain power estimates used by Kurzweil and Moravec such a computer should be capable of supporting a simulation of a bee brain but despite some interest no such simulation exists.
There are several reasons for this.
The neuron model seems to be oversimplified see next section.
There is insufficient understanding of higher cognitive processes to establish accurately what the brain s neural activity observed using techniques such as functional magnetic resonance imaging correlates with.
Even if our understanding of cognition advances sufficiently early simulation programs are likely to be very inefficient and will therefore need considerably more hardware.
The brain of an organism while critical may not be an appropriate boundary for a cognitive model.
To simulate a bee brain it may be necessary to simulate the body and the environment.
The Extended Mind thesis formalises this philosophical concept and research into cephalopods demonstrated clear examples of a decentralized system.
In addition the scale of the human brain is not currently well constrained.
One estimate puts the human brain at about 100 billion neurons and 100 trillion synapses.
Another estimate is 86 billion neurons of which 16 3 billion are in the cerebral cortex and 69 billion in the cerebellum.
Glial cell synapses are currently unquantified but are known to be extremely numerous.
Philosophical perspective Strong AI as defined in philosophy In 1980 philosopher John Searle coined the term strong AI as part of his Chinese room argument.
He wanted to distinguish between two different hypotheses about artificial intelligence.
Strong AI hypothesis An artificial intelligence system can think a mind and consciousness.
Weak AI hypothesis An artificial intelligence system can only act like it thinks and has a mind and consciousness.
The first one he called strong because it makes a stronger statement it assumes something special has happened to the machine that goes beyond those abilities that we can test.
The behaviour of a weak AI machine would be precisely identical to a strong AI machine but the latter would also have subjective conscious experience.
This usage is also common in academic AI research and textbooks.
Mainstream AI is most interested in how a program behaves.
According to Russell and Norvig as long as the program works they don t care if you call it real or a simulation.
If the program can behave as if it has a mind then there is no need to know if it actually has mind indeed there would be no way to tell.
For AI research Searle s weak AI hypothesis is equivalent to the statement artificial general intelligence is possible.
Thus according to Russell and Norvig most AI researchers take the weak AI hypothesis for granted and don t care about the strong AI hypothesis.
Thus for academic AI research Strong AI and AGI are two very different things.
In contrast to Searle and mainstream AI some futurists such as Ray Kurzweil use the term strong AI to mean human level artificial general intelligence.
Academic philosophers such as Searle do not believe that is the case and to most artificial intelligence researchers the question is out of scope.
Consciousness Other aspects of the human mind besides intelligence are relevant to the concept of strong AI and these play a major role in science fiction and the ethics of artificial intelligence.
These traits have a moral dimension because a machine with this form of strong AI may have rights analogous to the rights of non human animals.
Bill Joy among others argues a machine with these traits may be a threat to human life or dignity.
The role of consciousness is not clear and there is no agreed test for its presence.
If a machine is built with a device that simulates the neural correlates of consciousness would it automatically have self awareness.
It is possible that some of these traits naturally emerge from a fully intelligent machine.
It is also possible that people will ascribe these properties to machines once they begin to act in a way that is clearly intelligent.
Artificial consciousness research Although the role of consciousness in strong AI AGI is debatable many AGI researchers regard research that investigates possibilities for implementing consciousness as vital.
In an early effort Igor Aleksander argued that the principles for creating a conscious machine already existed but that it would take forty years to train such a machine to understand language.
Research challenges Progress in artificial intelligence has gone through periods of rapid progress separated by periods when progress appeared to stop.
Ending each hiatus fundamental advances in hardware software or both to create space for further progress.
For example the computer hardware available in the twentieth century was not sufficient to implement deep learning which requires large numbers of GPU enabled CPUs.
The field has also oscillated between approaches to the problem.
At times effort has focused on explicit accumulation of facts and logic as in expert systems.
At other times systems were expected to build their own g via machine learning as in artificial neural networks.
A further challenge is the lack of clarity in defining what intelligence entails.
Must it display the ability to set goals as well as pursue them.
Is it purely a matter of scale such that if model sizes increase sufficiently intelligence will emerge.
Are facilities such as planning reasoning and causal understanding required.
Does intelligence require explicitly replicating the brain and its specific faculties.
Gelernter writes No computer will be creative unless it can simulate all the nuances of human emotion.
Benefits AGI could have a wide variety of applications.
If oriented towards such goal AGI could help mitigate various problems in the world such as hunger poverty and health problems.
For example in public health AGI could accelerate medical research notably against cancer.
It could take care of the elderly and democratize access to rapid high quality medical diagnostics.
It could offer fun cheap and personalized education.
The need to work to subsist could become obsolete if the wealth produced is properly redistributed.
This also raises the question of the place of humans in a radically automated society.
It could also help to reap the benefits of potentially catastrophic technologies such as nanotechnology or climate engineering while avoiding the associated risks.
If an AGI s primary goal is to prevent existential catastrophes such as human extinction which could be difficult if the Vulnerable World Hypothesis turns out to be true it could take measures to drastically reduce the risks while minimizing the impact of these measures on our quality of life.
Risks Potential threat to human existence The thesis that AI poses an existential risk for humans and that this risk needs much more attention than it currently gets is controversial but has been endorsed by many public figures including Elon Musk Bill Gates and Stephen Hawking.
Gates states he does not understand why some people are not concerned and Hawking criticized widespread indifference in his 2014 editorial So facing possible futures of incalculable benefits and risks the experts are surely doing everything possible to ensure the best outcome right.
The fate of humanity has sometimes been compared to the fate of gorillas threatened by human activities.
Additional intelligence caused humanity to dominate gorillas which are now vulnerable in ways that they could not have anticipated.
The gorilla has become an endangered species not out of malice but simply as a collateral damage from human activities.
The skeptic Yann LeCun considers that AGIs will have no desire to dominate humanity and that we should be careful not to anthropomorphize them and interpret their intents as we would for humans.
He said that people won t be smart enough to design super intelligent machines yet ridiculously stupid to the point of giving it moronic objectives with no safeguards.
On the other side the concept of instrumental convergence suggests that almost whatever their goals intelligent agents will have reasons to try to survive and acquire more power as intermediary steps to achieving these goals.
And that this does not require having emotions.
Nick Bostrom gives the thought experiment of the paper clips optimizer Suppose we have an AI whose only goal is to make as many paper clips as possible.
Because if humans do so there would be fewer paper clips.
Also human bodies contain a lot of atoms that could be made into paper clips.
The future that the AI would be trying to gear towards would be one in which there were a lot of paper clips but no humans.
A 2021 systematic review of the risks associated with AGI while noting the paucity of data found the following potential threats AGI removing itself from the control of human owners managers being given or developing unsafe goals development of unsafe AGI AGIs with poor ethics morals and values inadequate management of AGI and existential risks.
Many scholars who are concerned about existential risk advocate possibly massive research into solving the difficult control problem to answer the question what types of safeguards algorithms or architectures can programmers implement to maximise the probability that their recursively improving AI would continue to behave in a friendly rather than destructive manner after it reaches superintelligence.
Solving the control problem is complicated by the AI arms race which will almost certainly see the militarization and weaponization of AGI by more than one nation state resulting in AGI enabled warfare and in the case of AI misalignment AGI directed warfare potentially against all humanity.
The thesis that AI can pose existential risk also has detractors.
Jaron Lanier argued in 2014 that the idea that then current machines were in any way intelligent is an illusion and a stupendous con by the wealthy.
Much criticism argues that AGI is unlikely in the short term.
Computer scientist Gordon Bell argues that the human race will destroy itself before it reaches the technological singularity.
Gordon Moore the original proponent of Moore s Law declares I am a skeptic.
I don t believe a technological singularity is likely to happen at least for a long time.
Mass unemployment Researchers from OpenAI estimated that 80 % of the U S workforce could have at least 10 % of their work tasks affected by the introduction of LLMs while around 19 % of workers may see at least 50 % of their tasks impacted.
They consider office workers to be the most exposed for example mathematicians accountants or web designers.
According to Stephen Hawking the outcome of automation on the quality of life will depend on how the wealth will be redistributed Everyone can enjoy a life of luxurious leisure if the machine produced wealth is shared or most people can end up miserably poor if the machine owners successfully lobby against wealth redistribution.
So far the trend seems to be toward the second option with technology driving ever increasing inequalityElon Musk considers that the automation of society will require governments to adopt a universal basic income.
Despite its high IQ ChatGPT fails at tasks that require real humanlike reasoning or an understanding of the physical and social world.
External links The AGI portal maintained by Pei Wang The Genesis Group at MIT s CSAIL Modern research on the computations that underlay human intelligence OpenCog open source project to develop a human level AI Simulating logical human thought What Do We Know about AI Timelines Literature review.
