Causal AI
{{Short description|Development of artificial intelligence}}
{{Use dmy dates|date=May 2023}}
{{Use British English|date=May 2023}}
Causal AI is a technique in artificial intelligence that builds a causal model and can thereby make inferences using causality rather than just correlation. One practical use for causal AI is for organisations to explain decision-making and the causes for a decision.{{Cite web |last=Blogger |first=SwissCognitive Guest |date=2022-01-18 |title=Causal AI |url=https://swisscognitive.ch/2022/01/18/casual-ai/ |access-date=2022-10-11 |website=SwissCognitive, World-Leading AI Network |language=en-US}}{{Cite journal |last1=Sgaier |first1=Sema K |last2=Huang |first2=Vincent |last3=Grace |first3=Charles |date=2020 |title=The Case for Causal AI |journal=Stanford Social Innovation Review |volume=18 |issue=3 |pages=50–55 |issn=1542-7099 |id={{ProQuest|2406979616}}}}
Systems based on causal AI, by identifying the underlying web of causality for a behaviour or event, provide insights that solely predictive AI models might fail to extract from historical data.{{Cite web |date=2024-06-29 |title=Beyond the Limits of Historical Data {{!}} causa |url=https://www.causa.tech/post/beyond-the-limits-of-historical-data |access-date=2024-06-29 |website=causa.tech |language=en-US}} An analysis of causality may be used to supplement human decisions in situations where understanding the causes behind an outcome is necessary, such as quantifying the impact of different interventions, policy decisions or performing scenario planning.{{Cite web |date=2023-02-28 |title=How to Understand the World of Causality {{!}} causaLens |url=https://causalens.com/resources/blogs/how-to-understand-the-world-of-causality/ |access-date=2023-10-07 |website=causalens.com |language=en-US}} A 2024 paper from Google DeepMind demonstrated mathematically that "Any agent capable of adapting to a sufficiently large set of distributional shifts must have learned a causal model".{{cite web|title=Robust agents learn causal world models|s2cid=267740124 }} The paper offers the interpretation that learning to generalise beyond the original training set requires learning a causal model, concluding that causal AI is necessary for artificial general intelligence.
History
The concept of causal AI and the limits of machine learning were raised by Judea Pearl, the Turing Award-winning computer scientist and philosopher, in 2018's The Book of Why: The New Science of Cause and Effect. Pearl asserted: “Machines' lack of understanding of causal relations is perhaps the biggest roadblock to giving them human-level intelligence.”{{Cite book |last=Pearl |first=Judea |publisher=Penguin Books| url=https://www.worldcat.org/oclc/1047822662 |title=The book of why : the new science of cause and effect |date=2019 |others=Dana Mackenzie |isbn=978-0-14-198241-0 |location=London, UK |oclc=1047822662}}{{Cite web |last=Hartnett |first=Kevin |date=15 May 2018 |title=To Build Truly Intelligent Machines, Teach Them Cause and Effect |url=http://www.quantamagazine.org/to-build-truly-intelligent-machines-teach-them-cause-and-effect-20180515/ |access-date=11 October 2022 |website=Quanta Magazine}}
In 2020, Columbia University established a Causal AI Lab under Director Elias Bareinboim. Professor Bareinboim’s research focuses on causal and counterfactual inference and their applications to data-driven fields in the health and social sciences as well as artificial intelligence and machine learning.{{Cite web |title=What AI still can't do |url=https://www.technologyreview.com/2020/02/19/868178/what-ai-still-cant-do/ |access-date=2022-10-18 |website=MIT Technology Review |language=en}} Technological research and consulting firm Gartner for the first time included causal AI in its 2022 Hype Cycle report, citing it as one of five critical technologies in accelerated AI automation.{{Cite web |title=What is New in the 2022 Gartner Hype Cycle for Emerging Technologies |url=https://www.gartner.co.uk/en/articles/what-s-new-in-the-2022-gartner-hype-cycle-for-emerging-technologies |access-date=2022-10-11 |website=Gartner |language=en-GB}}{{Cite web |last=Sharma |first=Shubham |date=2022-08-10 |title=Gartner picks emerging technologies that can drive differentiation for enterprises |url=https://venturebeat.com/ai/gartner-picks-emerging-technologies-that-can-drive-differentiation-for-enterprises/ |access-date=2022-10-11 |website=VentureBeat |language=en-US}}
One significant advance in the field is the concept of Algorithmic Information Dynamics:{{cite journal | last=Zenil | first=Hector | title=Algorithmic Information Dynamics | journal=Scholarpedia | date=25 July 2020 | volume=15 | issue=7 | doi=10.4249/scholarpedia.53143 | doi-access=free | bibcode=2020SchpJ..1553143Z | hdl=10754/666314 | hdl-access=free }} {{cite book | last1=Zenil | first1=Hector | last2=Kiani | first2=Narsis A. | last3=Tegner | first3=Jesper | title=Algorithmic Information Dynamics: A Computational Approach to Causality with Applications to Living Systems | publisher=Cambridge University Press | year=2023 | doi=10.1017/9781108596619 | isbn=978-1-108-59661-9 | url=https://doi.org/10.1017/9781108596619}} a model-driven approach for causal discovery using Algorithmic Information Theory and perturbation analysis. It solves inverse causal problems by studying dynamical systems computationally. A key application is causal deconvolution, which separates generative mechanisms in data with algorithmic models rather than traditional statistics. {{cite journal | last1=Zenil | first1=Hector | last2=Kiani | first2=Narsis A. | last3=Zea | first3=Allan A. | last4=Tegner | first4=Jesper | title=Causal deconvolution by algorithmic generative models | journal=Nature Machine Intelligence | volume=1 | issue=1 | pages=58–66 | year=2019 | doi=10.1038/s42256-018-0005-0 | hdl=10754/630919 | url=https://doi.org/10.1038/s42256-018-0005-0| hdl-access=free }} This method identifies causal structures in networks and sequences, moving away from probabilistic and regression-based techniques, marking one of the first practical Causal AI approaches using algorithmic complexity and algorithmic probability in Machine Learning. {{cite journal | last1=Hernández-Orozco | first1=Santiago | last2=Zenil | first2=Hector | last3=Riedel | first3=Jürgen | last4=Uccello | first4=Adam | last5=Kiani | first5=Narsis A. | last6=Tegnér | first6=Jesper | title=Algorithmic Probability-Guided Machine Learning on Non-Differentiable Spaces | journal=Frontiers in Artificial Intelligence | volume=3 | pages=567356 | year=2021 | doi=10.3389/frai.2020.567356 | doi-access=free | pmid=33733213 | pmc=7944352 }}