Intellectual Production #2: Artificial Intelligence
1. Who were these people, and how did/does each contribute to the development of artificial intelligence? How did/does each think “intelligence” could be identified?
Alan Matheson Turing (1912-1954):
British Mathematician, cryptanalyst and computer scientist, seen as the father of computing, introduced the concept of a universal machine, the basis of modern computers. Developed the Imitation Game to evaluate if a computer is capable of accurately mimicking human responses to stimuli (Biography, 2020; Oppy & Dowe, 2021; St. George & Gilles, 2021).
John McCarthy (1927-2011):
American computer and cognitive scientist who created the term “Artificial Intelligence,” inspired work on logic programming, time-sharing and pioneered programming languages like LISP, still used by the AI field. Involved in founding several prominent AI laboratories. Believed AI should develop itself like a human (Chakraborty, 2021; “John McCarthy, 2022).
Herb Simon (1916-2001):
American human behaviourist recognized for contributions to the field of AI in the areas of understanding human cognition regarding decision making and list processing and how it could be simulated by computers. Involved with the development of Information Processing Language (IPL) (Britannica, 2022b; “Herbert A. Simon”, 2022; NobelPrize.org, n.d.; UBS, n.d.).
Marvin Minsky (1927-2016):
American mathematician and computer scientist who pioneered neural-network learning machines. Believed the brain is a machine whose functions could be replicated by computers. Cofounder of MITs AI Lab with McCarthy. Believed intelligence emerges from the interactions of numerous non-intelligent simple components or “agents” (BBC.com, 2016; Knight, 2016; “Marvin Minsky”, 2022).
Timnit Gebru (1982):
American leader in AI ethics research, Google whistleblower and advocate for diversity in the technology industry. Highlighted AIs flaws with regards to women and people of colour and the risks of AIs trained on enormous amounts of text data culturally, financially and ecologically (Hao, 2020; “Timnit Gebru”, 2022).
2. How do “machine (programming) languages” differ from human (natural) ones?
The binary nature of Machine Language (ML) means it will not function with a flaw, whereas Human Languages (HL) have the plasticity to overcome errors or improvisations. The biggest difference between the two is context, in ML grammar does not change due to context in HL context imparts different meanings for many words. ML are precisely predefined at their creation to be unambiguous, and don’t evolve, either replaced, or become the basis of new languages unlike, HLs that constantly evolve in both their semantics and syntax. ML are designed to be logical, like HL, but lack the ability to also express emotion like HLs (Harris, 2018).
3. How does “machine (artificial) intelligence” differ from the human version?
Traditionally AI or machine intelligence has been described or defined, by Chollet (2019) and others, as the ability to perform specific, pre-set tasks, in other words is skills-based – heavily influenced by experience and prior knowledge. Many AIs in current operation are narrow AIs, designed and trained how to preform tasks like pattern recognition to superhuman abilities, but unable to accomplish any other task. Chollet (2019) defines human intelligence as the ability to learn new and unrelated skills without prior knowledge, experience, or exposure which has become the goal of general or robust AIs, those designed to preform any task.
4. How does “machine learning” differ from human learning?
Machine learning is more accurately training, the designer programs the system, sets parameters and tasks to be accomplished, exposes the system to as much data as possible, the system uses that data to make judgements and predictions about the information based on patterns it detects within the data. Machines lack ethics, they only know what the data they were trained on has shown them without context or empathy; they cannot deal with problems that require subjective answers. A big problem is the data used to train the systems lack the diversity of humanity; the system lacks an outside agent to provide evaluation of actions and decisions (Buolamwini, 2019; Heilweil, 2020).
5. And for your LAST challenge, a version of the Turing Test: how do YOUR answers to these questions differ from what a machine could generate?
The condensed nature of these responses, necessitated by the word limits, requires a level of analysis and synthesis to parse out the most pertinent information, and not merely regurgitating information found on the internet. The analyses they contain are based on my life experiences, beliefs, context and underlying biases all of these factors should set them apart from those that a machine could generate. They are also constructed in accordance with the context of this course, which would not be possible for AI. Additionally, a machine generated response would contain greater syntax errors. Though if what Gebru warns is correct, that AIs trained on unthinkable large data sets, under the correct circumstances, will be produce answers that would seem to pass the Imitation Game and be indistinguishable from a response composed by humans would invalidate many of the above reasons (Hao, 2020; Oppy & Dowe, 2021). To thwart more advanced AIs in the future would require the construction of questions/tasks of greater complexity and contextual richness that AIs would not be able to correctly interpret and respond to (Harris, 2018).
References
BBC News. (2016, January 26). AI pioneer Marvin Minsky dies aged 88. BBC. https://www.bbc.com/news/technology-35409119
Biography. (2020, July 22). Alan Turing. Biography.com. https://www.biography.com/scientist/alan-turing
Britannica, T. Editors of Encyclopaedia (2022a, January 11). John McCarthy. Encyclopedia Britannica. https://www.britannica.com/biography/John-McCarthy
Britannica, T. Editors of Encyclopaedia (2022b, February 5). Herbert A. Simon. Encyclopedia Britannica. https://www.britannica.com/biography/Herbert-A-Simon
Buolamwini, J. (2019, February 7). Artificial Intelligence Has a Problem with Gender and Racial Bias. Here’s How to Solve It. Time. https://time.com/5520558/artificial-intelligence-racial-gender-bias/
Chakraborty, M. (2021, March 1). Knowing John McCarthy: The Father of Artificial Intelligence. Analytics Insight. https://www.analyticsinsight.net/knowing-john-mccarthy-the-father-of-artificial-intelligence/
Chollet, F. (2019, November 5). On the Measure of Intelligence. https://arxiv.org/pdf/1911.01547.pdf
Hao, K. (2020, December 4). We read the paper that forced Timnit Gebru out of Google. Here’s what it says. MIT Technology Review. https://www.technologyreview.com/2020/12/04/1013294/google-ai-ethics-research-paper-forced-out-timnit-gebru
Harris, A. (2018, November 1). Human languages vs. programming languages. Medium. https://medium.com/@anaharris/human-languages-vs-programming-languages-c89410f13252
Heilweil, R. (2020, February 18). Why algorithms can be racist and sexist. A computer can make a decision faster. That doesn’t make it fair. Vox. https://www.vox.com/recode/2020/2/18/21121286/algorithms-bias-discrimination-facial-recognition-transparency
Herbert A. Simon. (2022, June 2). In Wikipedia. https://en.wikipedia.org/w/index.php?title=Herbert_A._Simon&oldid=1091148001
John McCarthy (computer scientist). (2022, May 23). In Wikipedia. https://en.wikipedia.org/w/index.php?title=John_McCarthy_(computer_scientist)&oldid=1089355282
Knight, W. (2016, January 26) What Marvin Minsky Still Means for AI. MIT Technology Review https://www.technologyreview.com/2016/01/26/163622/what-marvin-minsky-still-means-for-ai/
Marvin Minsky. (2022, May 25). In Wikipedia. https://en.wikipedia.org/w/index.php?title=Marvin_Minsky&oldid=1089829244
NobelPrize.org. (n.d.). Herbert Simon – Biographical. https://www.nobelprize.org/prizes/economic-sciences/1978/simon/biographical/
Oppy, G. & Dowe, D., (2021, October 4) The Turing Test. In E. N. Zalta (ed.), The Stanford Encyclopedia of Philosophy. (Winter 2021 Edition). https://plato.stanford.edu/archives/win2021/entries/turing-test/
St. George, B., & Gillis, A. S. (2021, June 14). What is the Turing Test? SearchEnterpriseAI. https://www.techtarget.com/searchenterpriseai/definition/Turing-test#:~:text=The%20Turing%20Test%20is%20a,cryptanalyst%2C%20mathematician%20and%20theoretical%20biologist
Timnit Gebru. (2022, May 4). In Wikipedia. https://en.wikipedia.org/w/index.php?title=Timnit_Gebru&oldid=1086231475
UBS. (n.d.). Meet the Nobel Laureates in Economics: Do we understand human behaviour. https://www.ubs.com/microsites/nobel-perspectives/en/laureates/herbert-simon.html