Arriving Late to the AI Party
Depending on who you ask, you are either late to the AI party or early to it.
The fact that artificial intelligence (AI) is developing at an anxiety-inducing, accelerated speed is continuously reiterated, turbo-charged by the media’s need to inform, but also encash on keyword search trends — words and phrases that you are punching into your search (‘Will AI take over my job?’) to get answers that may help you sleep at night, which in the hands of a few are data points to optimally target content and advertisements at you (You are the product, as you know).
If you do a basic search on Google Trends for the keywords ‘AI’ and ‘ChatGPT’ — with the filters set at worldwide, past 5 years, and web search — you will see that the graphs for both rise around 2022 when OpenAI launched ChatGPT — an AI chatbot and ‘virtual assistant’. Interestingly, at the time of writing this (2024), the value for ‘AI’ has already hit 100 on the graph, indicating the peak popularity of the term.
This “dizzying pace of change” is captured by Thomas L Friedman in Thank You For Being Late: An Optimist’s Guide to Thriving in the Age of Accelerations [1]. Friedman writes,
“[W]e are living through one of the greatest inflection points in history… The three largest forces on the planet — technology, globalization, and climate change — are all accelerating at once… it is easy to get overwhelmed by it all.”
And,
“So many people today seem to be looking for someone to put on the brakes, or take a hammer to the forces of change — or just give them a simple answer to make their anxiety go away.”
For tech researchers, journalists, and content creators, this anxiety is manyfold as they try to keep up with the latest developments in AI; engage with the legal, political, ethical, environmental, and social concerns around it; and gauge its impact with all its myriad contexts and complexities.
Every day, it seems like we are already late to the AI party — and already late to start this project.
Increasingly, however, there is a shared understanding of the limits of AI and the gap between tech and its tailored-to-task implementation — an acknowledgement that we may, in fact, be early to the party.
AI Productivity Paradox
According to Irene Ek, leader of the AI portfolio at the Swedish Agency for Growth Policy Analysis, AI is perceived to be a ‘growth engine’ and you see it almost everywhere, but its impact hasn’t been seen in many countries’ productivity statistics. One major reason for this ‘productivity paradox’, Ek says, is that there is a time lag between the development of technology and its implementation [2].
In ‘Artificial Intelligence and the Modern Productivity Paradox’, researchers Erik Brynjolfsson, Daniel Rock, and Chad Syverson suggest four major reasons for this paradox: false hopes, mismeasurement, redistribution, and implementation lags. They, too, perceive implementation lags to be the “biggest contributor to the clash of expectations [from AI]”. Quoting this research, Ek writes [3],
“These explanations suggest that AI capabilities have yet to be diffused widely and full effects will not be realised until waves of complementary innovations are developed and implemented.”
It is expected that the development and adoption of AI will follow different trajectories in different regions — each creating a unique and fascinating story — but, as history shows, the measurable impact of this ‘fourth industrial revolution’, like that of all industrial revolutions, will take time to manifest itself.
There is also an understanding that the technology isn’t quite there yet and we may be expecting too much from it (more “clash of expectations”) at this stage of its genesis.
These we-are-early-to-the-party conclusions become clearer when we look at AI in newsrooms and news media — which is the broader focus of this project.
AI in the Newsroom
Being in a digital newsroom in India at the time of ChatGPT’s fated release and the AI wave completely washing over us was exciting and stressful. Product, marketing, innovation, and business teams scrambled to come up with the next big AI idea — tap into the ‘first-mover advantage’, as we are taught.
The optimism was reflected in newsroom research at the time with the Reuters Institute for the Study of Journalism, the University of Oxford announcing in its ‘Digital News Report 2023’,
“New technological disruption from Artificial Intelligence (AI) is just around the corner, threatening to release a further wave of personalised, but potentially unreliable content.”
Doubts, then, were already cropping up, with some newsrooms making the sane decision to begin drafting internal AI guidelines and policies — an essential step in the absence of other regulatory frameworks addressing (what is called) ‘AI ethics’.
Some brakes were pulled and more time was invested in considering the suitability and sustainability of using AI for content in a fractured world where trust in news was already declining. The latest ‘Digital News Report 2024’ by the Reuters Institute highlights these concerns —
“[A]cross all countries, only a minority currently feels comfortable using news made by humans with the help of AI (36%), and an even smaller proportion is comfortable using news made mostly by AI with human oversight (19%).”
And,
“Overall, we are still at the early stages of journalists’ usage of AI, but this also makes it a time of maximum risk for news organisations…”
Another research on newsrooms with a small data set but mighty observations — ‘Generative AI in Journalism’ — details the ‘aspirational usage of AI’ with respondents’ interests lying in the use of AI in news discovery, ideation, sense-making, research, news aggregation and curation, personalisation, coding, fake news detection, and more — use cases that are currently in the realm of possibility because the tech (at least that of Generative AI or GenAI) is found to be wanting.
AI Literacy
There is another issue at play here — another contributing factor to the AI productivity paradox — and the Reuters’ ‘Digital News Report 2024’ is hyper-aware of it: AI awareness and literacy —
“Across 28 markets where AI questions were included, selfreported awareness is relatively low, with less than half (45%) of respondents saying they have heard or read a large or moderate amount about it.”
The report further suggests that while people with greater AI awareness tend to feel relatively more comfortable with the use of AI in journalism, most audiences have not put much, if any, thought into how AI could be used for news.
Without AI literacy, people’s starting point is generally that of resistance, suspicion, and fear.
With AI literacy, they may be empowered to — as a respondent in a 2023 research by the London School of Economics’ JournalismAI project says — “adopt a power-conscious framing of global AI development and adoption.”
The world’s first comprehensive AI law, the EU Artificial Intelligence Act (EU AI Act) — formally adopted in May 2024 — lists AI literacy as an organisational requirement (in Article 4) —
“Providers and deployers of AI systems shall take measures to ensure, to their best extent, a sufficient level of AI literacy of their staff and other persons dealing with the operation and use of AI systems on their behalf, taking into account their technical knowledge, experience, education and training and the context the AI systems are to be used in, and considering the persons or groups of persons on whom the AI systems are to be used.”
EducAIted — An Introduction
Sense-making, for me, is the first step towards AI literacy — the main aim of this project, which I am calling EducAIted.
Through a series of posts and complementary resources, I hope to unpack seemingly opaque and complicated AI-related ideas and issues — and make learning about AI feel less overwhelming.
This essay is an example of the kind of posts that you can expect from me — and some concepts mentioned here will be elaborated in upcoming posts. I am also creating a database of organisations that are working towards AI literacy and open-to-all resources (coming soon). Eventually, I hope to bring in diverse voices to cover areas of AI literacy and impact that I may not be well-versed in.
The inspiration for this project came from the early-to-the-AI-party camp — though late or early, either way, I am here — and a part of the effort is also to document my learning process. Friedman argues that to be a good columnist, one needs a ‘worldview’ — “some “take” on the biggest forces shaping the world in which we live and how to influence them”. I am in search of this worldview.
[1] This book — Thank You For Being Late: An Optimist’s Guide to Thriving in the Age of Accelerations by Thomas L Friedman — is told from a Western perspective, arguably for a Western audience, though it claims to be “radically inclusive”. It also promotes meritocracy in some chapters. So, I don’t recommend it, unless you are curious about how America perceives itself and its role in ‘the age of accelerations’.
[2] Paraphrased from the section ‘Difference Between Definitions of AI, with Irene Ek’ in Module 1: ‘Introduction to AI in work and business’ of the online course ‘AI, Business & the Future of Work’ by Lund University on Coursera.
[3] Ek, I. & Montagnier, P. (March 2021). AI measurement in ICT usage surveys: A review. OECD DIGITAL ECONOMY PAPERS №308.
