Weak artificial intelligence (weak AI) is artificial intelligence that implements a limited part of the mind, or, as narrow AI,[1][2][3] is focused on one narrow task.
Weak AI is contrasted with strong AI, which has been defined variously as:
- Artificial general intelligence (AGI): a machine with the ability to apply intelligence to any problem, rather than just one specific problem (as with narrow AI).
- Artificial superintelligence (ASI): a machine with a vastly superior intelligence to the average human being.
- Artificial consciousness: a machine that has consciousness, sentience and mind (John Searle uses "strong AI" in this sense).
Scholars such as Antonio Lieto have argued that the current research on both AI and cognitive modelling are perfectly aligned with the weak-AI hypothesis (that should not be confused with the "general" vs "narrow" AI distinction) and that the popular assumption that cognitively inspired AI systems espouse the strong AI hypothesis is ill-posed and problematic since "artificial models of brain and mind can be used to understand mental phenomena without pretending that that they are the real phenomena that they are modelling"[4] (as, on the other hand, implied by the strong AI assumption).
Narrow AI can be classified as being “limited to a single, narrowly defined task. Most modern AI systems would be classified in this category.”[5] Narrow means the robot or computer is strictly limited to only being able to solve one problem at a time. Strong AI is conversely the opposite. Strong AI is closer to the human brain. This is all believed to be the case by philosopher John Searle. This idea of strong AI is also controversial. Searle believes that the Turing test (created by Alan Turing during WW2, originally called the Imitation Game, used to test if a machine is as intelligent as a human) is not accurate or appropriate for testing strong AI.[6]
Weak AI versus strong AI
editThe differences between weak AI vsersus strong AI are not widely cataloged at the moment. Weak AI is often associated with basic technology like voice-recognition software such as Siri or Alexa. Whereas strong AI is not fully implemented or testable yet, it is only really fantasized about in movies or popular culture media.[7]
It seems that one approach to AI moving forward is one of an assisting or aiding role to humans. There are some sets of data or numbers that even we humans cannot fully process or understand as quickly as computers can, so this is where AI will play a helping role for us.[8][relevant?]
Impact
editSome commentators[who?] think narrow AI could be dangerous because of this "brittleness" and fail in unpredictable ways. Narrow AI could cause disruptions in the electric grid, damage nuclear power plants, cause global economic problems, and misdirect autonomous vehicles.[1]
Examples
editSome examples of narrow AI are AlphaGo,[9] self-driving cars, robot systems used in the medical field, and diagnostic doctors. Narrow AI systems are sometimes dangerous if unreliable. Medicines could be incorrectly sorted and distributed. Also, medical diagnoses can ultimately have serious and sometimes deadly consequences if the AI is faulty or biased.[10] Another issue with narrow AI, currently, is that behavior that it follows can become inconsistent.[11] It could be difficult for the AI to grasp complex patterns and get to a solution that works reliably in various environments.
Simple AI programs have already worked their way into our society unnoticed. Autocorrection for typing, speech recognition for speech-to-text programs, and vast expansions in the data science fields are examples.[12] As much as narrow and relatively general AI is slowly starting to help out societies, they are also starting to hurt them as well. AI had already unfairly put people in jail, discriminated against women in the workplace for hiring, taught some problematic ideas to millions, and even killed people with automatic cars.[13] AI might be a powerful tool that can be used for improving our lives, but it could also be a dangerous technology with the potential for things to get out of hand.
Social media
editFacebook, and other similar social media platforms, have been able to figure out how to use AI and machine learning, or more specifically narrow AI, to predict how people will react to being shown certain images. Narrow AI systems have been able to identify what users will engage with, based on what they post, following the patterns or trends.[14]
Twitter has started to have more advanced AI systems to figure out how to identify narrower AI forms and detect if bots may have been used for biased propaganda, or even potentially malicious intentions. These AI systems do this through filtering words and creating different layers of conditions based on what AI has had implications for in the past, and then detecting if that account may be a bot or not.[15]
TikTok uses its "For You" algorithm to determine a user's interests very quickly through analyzing patterns in what videos the user initially chooses to watch. This narrow AI system uses patterns found between videos to determine what video should be shown next including the duration, who has shared or commented on it already, and music played in the videos. The "For You" algorithm on TikTok is so accurate, that it can figure out exactly what a user has an interest in or even really loves, in less than an hour.[16]
See also
edit- Artificial intelligence – Intelligence of machines
- A.I. Rising – 2018 film directed by Lazar Bodroža
- Deep learning – Branch of machine learning
- Expert system – Computer system emulating the decision-making ability of a human expert
- History of artificial intelligence
- Virtual assistant – Software agent
- Machine learning – Study of algorithms that improve automatically through experience
- Philosophy of artificial intelligence
- Artificial general intelligence – Human-level or stronger AI for a wide range of tasks
- Hardware for artificial intelligence – Hardware specially designed and optimized for artificial intelligence
- Synthetic intelligence – Alternate term for or form of artificial intelligence
References
edit- ^ a b Dvorsky, George (1 April 2013). "How Much Longer Before Our First AI Catastrophe?". Gizmodo. Retrieved 27 November 2021.
- ^ Muehlhauser, Luke (18 October 2013). "Ben Goertzel on AGI as a Field". Machine Intelligence Research Institute. Retrieved 27 November 2021.
- ^ Chalfen, Mike (15 October 2015). "The Challenges Of Building AI Apps". TechCrunch. Retrieved 27 November 2021.
- ^ Lieto, Antonio (2021). Cognitive Design for Artificial Minds. London, UK: Routledge, Taylor & Francis. p. 85. ISBN 9781138207929.
- ^ Bartneck, Christoph; Lütge, Christoph; Wagner, Alan; Welsh, Sean (2021). An Introduction to Ethics in Robotics and AI. SpringerBriefs in Ethics. Cham: Springer International Publishing. doi:10.1007/978-3-030-51110-4. ISBN 978-3-030-51109-8. S2CID 224869294.
- ^ Liu, Bin (28 March 2021). ""Weak AI" is Likely to Never Become "Strong AI", So What is its Greatest Value for us?". arXiv:2103.15294 [cs.AI].
- ^ Kerns, Jeff (15 February 2017). "What's the Difference Between Weak and Strong AI?". ProQuest. ProQuest 1876870051.
- ^ LaPlante, Alice; Maliha, Balala (2018). Solving Quality and Maintenance Problems with AI. O'Reilly Media, Inc. ISBN 9781491999561.
- ^ Edelman, Gary Grossman (3 September 2020). "We're entering the AI twilight zone between narrow and general AI". VentureBeat. Retrieved 16 March 2024.
- ^ Szocik, Konrad; Jurkowska-Gomułka, Agata (16 December 2021). "Ethical, Legal and Political Challenges of Artificial Intelligence: Law as a Response to AI-Related Threats and Hopes". World Futures: 1–17. doi:10.1080/02604027.2021.2012876. ISSN 0260-4027. S2CID 245287612.
- ^ Kuleshov, Andrey; Prokhorov, Sergei (September 2019). "Domain Dependence of Definitions Required to Standardize and Compare Performance Characteristics of Weak AI Systems". 2019 International Conference on Artificial Intelligence: Applications and Innovations (IC-AIAI). Belgrade, Serbia: IEEE. pp. 62–623. doi:10.1109/IC-AIAI48757.2019.00020. ISBN 978-1-7281-4326-2. S2CID 211298012.
- ^ Earley, Seth (2017). "The Problem With AI". IT Professional. 19 (4): 63–67. doi:10.1109/MITP.2017.3051331. ISSN 1520-9202. S2CID 9382416.
- ^ Anirudh, Koul; Siddha, Ganju; Meher, Kasam (2019). Practical Deep Learning for Cloud, Mobile, and Edge. O'Reilly Media. ISBN 9781492034865.
- ^ Kaiser, Carolin; Ahuvia, Aaron; Rauschnabel, Philipp A.; Wimble, Matt (1 September 2020). "Social media monitoring: What can marketers learn from Facebook brand photos?". Journal of Business Research. 117: 707–717. doi:10.1016/j.jbusres.2019.09.017. ISSN 0148-2963. S2CID 203444643.
- ^ Shukla, Rachit; Sinha, Adwitiya; Chaudhary, Ankit (28 February 2022). "TweezBot: An AI-Driven Online Media Bot Identification Algorithm for Twitter Social Networks". Electronics. 11 (5): 743. doi:10.3390/electronics11050743. ISSN 2079-9292.
- ^ Hyunjin, Kang (September 2022). "AI agency vs. human agency: understanding human-AI interactions on TikTok and their implications for user engagement". academic.oup.com. Retrieved 8 November 2022.