The Reality of AI in Social Media

I am perusing a travel website, looking for some cruises that may interest me. A pop-up window appears and Frank asks me if he can be of assistance. Cool; what a nice guy. After a some quick Q and A Frank points out some cruises that may match what I am looking for. He also makes sure I am aware of the special offers available that are associated with those cruises. Can He be of any further assistance? No thank you; that was great. Frank logs off but I can see his little window in case I have further need. I ponder how tedious a job that could be for a person. It is a person right. Right?

The intersection of Artificial Intelligence (AI) and social media is considered a Web 3.0 phenomenon (see my blog There are versions of Web? Really?) and the ability for machines to interpret Web data has been referred to as a facet of the Symantic Web (W3C 2013). Poole, Mackworth & Goebel (1998), define AI as “any device that perceives its environment and takes actions that maximize its chance of successfully achieving its goals” and is commonly interpreted as the actions of computers mimicking human cognitive functions (Russel and Norig 2009).

robot artificial intelligence by geralt (CC0 1.0)

Of interest is the argument that the ‘intelligence’ component no longer applies if the technology becomes routine (Schank 1991), creating a grey area between what could and would not be considered AI. This perception places AI above algorithmic data interpretation and subsequent reaction. For me, this would place AI in the Web 4.0 domain, involving functions beyond the data activity attributed to it in Web 3.0. If then, AI is not yet routine or is still conceptual or under development, I would perceive an AI attribution occurring if an interaction between ourselves and a device occurs and it is not immediately perceivable as including machine cognition. Our use of social media would intersect with AI if our activity included manipulation by a device as a non-obvious participant in that activity.

Our use of social media would intersect with AI if our activity included manipulation by a device as a non-obvious participant in that activity

"Our use of social media would intersect with Artificial Intelligence (AI) if our activity included manipulation by a device as a non-obvious participant in that activity." What are the ethics of explicit vs non-explicit participation of AI in social media? #ALC701 pic.twitter.com/wDYOnkgyUt

— Peter (@Digital_P3T3R) July 26, 2018

Woah; hang on a sec. I just had a Terminator moment. So the concept here is that social activity can involve mimicked cognitive action by a machine and that participation is not explicit. That has to be sci-fi narrative; right?

Kaplan and Haenlein (2010, in Zafarani et.al. 2014) define social media as the “group of internet-based applications.. that allow the creation and exchange of user-generated content”. It is generally accepted that the “Internet and mobile technologies have been the primary force behind the rise of social media” (Zeng et.al 2010) with user/consumer-generated media and associated Web-based applications, characteristic of Web 2.0, producing an explosion of user data and meta-data (Zafarani et.al. 2014). These applications, also referred to as platforms, include categories of networking, blogging, media sharing, live-casting, virtual worlds, social gaming, and social search.

How does AI fit into this definition of social media?

 artificial intelligence technology by sujins (CC0 1.0)

Contemporary social media research and (AI) application has focused primarily on data analytics, “developing and evaluating informatics tools and frameworks to collect, monitor, analyze, summarize, and visualize social media data” (Zeng et.al 2010). The focus on data mining is a dominant informatics research and development domain (Zeng et.al 2010) with a large investment in social media mining (Zafarani et.al. 2014) analysing and representing actionable patterns extracted from social media generated data. A key element of this is the ‘mining’ of human behavioural patterns (Zafarani et.al. 2014).

Traditionally then, AI is the application of technology to perform mining tasks. Not really a component of social media, but rather a third party application. This mining was able to generate insertions into Web-based applications, such as advertising, responding to the data analysis. Google’s search prompting would be another example of this.

Zafarani et.al. (2014) promote the term social media intelligence (SMI) to describe the development of AI within social media elements. As well as “extracting useful patterns and intelligence to serve entities” (Zafarani et.al. 2014), social media intelligence would primarily facilitate interaction, a key element of Web 2.0, and a facet of the symbiotic concepts of Web 4.0. Zafarani et.al. (2014) describe the aims of social media intelligence as AI deriving actionable information, in context-rich application settings, from social media; a definition similar to the definition of AI by Poole et.al. (1998) presented at the commencement of this blog. As Web 4.0 develops and integrates into our connectivity, I predict AI will become referred to more as a symbiotic intelligence (SI).

I predict AI will become referred to more as a symbiotic intelligence (SI).

artificial intelligence brain by geralt (CC0 1.0)

Coming back to the question of explicit AI interaction, to what degree is SMI obvious to the social media participant? As discussed, Frank analysed my patterns of search about cruises and facilitated further interaction with the organisations promoting consumerism of their cruise products. Does it matter if Frank is human or AI? I assume Frank was actually human in this case, but I am not 100% sure.

What is your understanding of the role of Artificial Intelligence (AI) in your social media participation? #ALC701 pic.twitter.com/n0jYFodggC

— Peter (@Digital_P3T3R) July 26, 2018

Podcast

(Music track on this podcast is Life Is by Scott Buckley (CC By 4.0))

Of Interest

An interesting background video The Rise of AI, posted to YouTube by Hello World on 5 June 2018.

CBC News posted this story on 26 July 2018, Calgary malls use facial recognition to track shoppers’ age, gender without consent.

Curious how facial recognition works? A look into AI in action by Dave Smith‘s blog cutting edge face recognition is complicated.

Matteo Gianpietro Zago writes an interesting article with a section on the Web 3.0 revolution and thoughts on a pro-privacy, anti-monopoly web.

ColdFusion posted a YouTube video titled AI Codes its Own ‘AI Child’ – Artificial Intelligence breakthrough! on 15 January 2018 with content discussing neural networks building their own neural networks – AI learning!

I followed this up with a tweet. Feel free to contribute your own thoughts.

The thought of AI not even needing us to learn and grow is a bit sci-fi scary but is reality! #ALC701 https://t.co/hKmZUXlZJ4

— Peter (@Digital_P3T3R) July 27, 2018

 

 

References

 Poole, D, Mackworth, A and Goebel, R 1998, Computational Intelligence: A Logical Approach, Oxford University Press, New York.

Russell, S J  and Norvig, P 2009, Artificial Intelligence: A Modern Approach, 3rd ed., Prentice Hall, New Jersey.

Schank, R C 1991, Where’s the AI, AI Magazine, vol. 12, no. 4, p.38, retrieved 26 July 2018, https://doi.org/10.1609/aimag.v12i4.917.

W3C 2013, Semantic Web Activity, retrieved 26 July 2018, https://www.w3.org/2001/sw/

Zafarani, R, Abbasi, M A and Liu, H 2014, Social Media Mining: An Introduction,  Cambridge University Press, New York.

Zeng, D, Chen, H, Lusch, R and Li, S 2010, Social Media Analytics and Intelligence. IEEE Intelligent Systems, vol. 25, no. 6, pp.13-16.