The UK’s Competition and Markets Authority (CMA) has issued a warning about the potential risks of artificial intelligence (AI) foundation models. These AI systems, trained on massive, unlabeled data sets, underpin large language models and can be used for various tasks. The CMA has proposed principles to guide the development and use of foundation models, including accountability, access, diversity, choice, flexibility, fair dealing, and transparency. The report warns that poorly developed AI models could lead to societal harm, such as exposure to false and misleading information and AI-enabled fraud. The CMA also warns that market dominance from a few firms could lead to anticompetition concerns, with established players using foundation models to entrench their position and deliver overpriced or poor quality products and services. The CMA will provide an update on its thinking in early 2024. The UK government has tasked the CMA with weighing in on the country’s AI policy, but has opted to give responsibility for AI governance to sectoral regulators.
Call for Papers
- Thu, November 30, 2023: Paper submission deadline
- Wed, January 31, 2024: Notification
- Thu, February 29, 2024: Camera-ready versions due
- Tue-Fri, May 21 – May 24, 2024: Conference dates
All dates are 23:59 Anywhere on Earth time
Possible topics across methodological approaches and digital contexts include but are not limited to:
Understanding the Web
- Automation and AI in all its manifestations relevant to the Web
- Trends in globalization, fragmentation, and polarization of the Web
- The architecture and philosophy of the Web
- Critical analyses of the Web and Web technologies
Making the Web Inclusive
- Issues of discrimination and fairness
- Intersectionality and design justice in questions of marginalization and inequality
- Ethical challenges of technologies, data, algorithms, platforms, and people on the Web
- Safeguarding and governance of the Web, including anonymity, security, and trust
- Inclusion, literacy, and the digital divide
The Web and Society
- Social machines, crowd computing, and collective intelligence
- Web economics, social entrepreneurship, and innovation
- Legal issues, including rights and accountability for AI actors
- Humanities, arts, and culture on the Web
- Politics and social activism on the Web
- Online education and remote learning
- Health and well-being online
- The role of the Web in the future of (augmented) work
- The Web as a source of news and information, and misinformation
Doing Web Science
- Data curation, Web archives and stewardship in Web Science
- Temporal and spatial dimensions of the Web as a repository of information
- Analysis and modeling of human vs. automatic behavior (e.g., bots)
- Analysis of online social and information networks
- Detecting, preventing, and predicting anomalies in Web data (e.g., fake content, spam)
2024 Emphasis: Reflecting on the Web, AI, and Society
In addition to the topics at the heart of Web Science, we also welcome submissions addressing the interplay between the Web, AI, and society. New advances in AI are revolutionizing the way in which people use the Web and interact through it. As these technologies develop, it is crucial to examine their effect on society and the socio-technical environment in which we find ourselves. We are nearing the crossroads wherein content on the Web will increasingly be automatically generated, blended with that created by humans. This creates new potential yet brings new challenges and exacerbates existing ones in relation to data quality and misinformation. Additionally, we need to consider the role of the Web as a source of data for AI, including privacy and copyright concerns, as well as bias and representativity of resulting systems. The potential impact of new AI tools on the nature of work may bring a transformation of some careers while creating whole new ones. This year’s conference especially encourages contributions documenting different uses of AI in relation to how people use the Web, and in the ways the Web affects the creation and deployment of AI tools.
Format of the submissions
Please upload your submissions via EasyChair:
There are two submission formats:
- Full papers should be between 6 and 10 pages (including references, appendices, etc.). Full papers typically report on mature and completed projects.
- Short papers should be up to 5 pages (including references, appendices, etc.). Short papers will primarily report on high-quality ongoing work not mature enough for a full-length publication.
All accepted submissions will be assigned an oral presentation (of two different lengths):
All papers should adopt the current ACM SIG Conference proceedings template (acmart.cls). Please submit papers as PDF files using the ACM template, either in Microsoft Word format (available at https://www.acm.org/publications/proceedings-templateunder “Word Authors”) or with the ACM LaTeX template on the Overleaf platform, which is available at https://www.overleaf.com/latex/templates/association-for-computing-machinery-acm-sig-proceedings-template/bmvfhcdnxfty. In particular; please ensure that you are using the two-column version of the appropriate template.
All contributions will be judged by the Program Committee upon rigorous peer review standards for quality and fit for the conference by at least three referees. Additionally, each paper will be assigned to a Senior Program Committee member to ensure review quality.
WebSci-2024 review is double-blind. Therefore, please anonymize your submission: do not put the author(s) names or affiliation(s) at the start of the paper, and do not include funding or other acknowledgments in papers submitted for review. References to authors’ own prior relevant work should be included but should not specify that this is the authors’ own work. It is up to the authors’ discretion how much to further modify the body of the paper to preserve anonymity. The requirement for anonymity does not extend outside of the review process, e.g., the authors can decide how widely to distribute their papers over the Internet. Even in cases where the author’s identity is known to a reviewer, the double-blind process will serve as a symbolic reminder of the importance of evaluating the submitted work on its own merits without regard to the authors’ reputation.
For authors who wish to opt-out of publication proceedings, this option will be made available upon acceptance. This will encourage the participation of researchers from the social sciences that prefer to publish their work as journal articles. All authors of accepted papers (including those who opt out of proceedings) are expected to present their work at the conference.
Program Committee Chairs:
Oshani Seneviratne (Rensselaer Polytechnic Institute)
Luca Maria Aiello (IT University of Copenhagen)
Yelena Mejova (ISI Foundation)
For any questions and queries regarding the paper submission, please contact the chairs at firstname.lastname@example.org.
Call for Posters
Call for Workshops and Tutorials
Ian: Matt, it feels strange to welcome you as our newest Lab Director when I think I’ve known you as part of the Web Science community for at least 10 years
Matt: Probably longer – I think my interest in Web Science and particularly Web data goes back to the very first Web Science conference in 2010 and perhaps before that.
Ian: So was Web Data your point of entry to Web Science?
Matt: Thats right, I’d spent a lot of time looking what was thought of as archived web data and trying to render those as large-scale researchable data collections. We went through a number of iterations from a system called Hub Zero through to Archives Unleashed and most recently that work was integrated into the Internet Archive research services by a team at University of Waterloo so that people who are looking to extract value and sound research conclusions from these data sets can find them and access them through well -supported high quality platforms and tools.
Ian: How hard it is to get everyone involved?
Matt: Well one of the major challenges is trying to get people to share and engage with these data sets outside of tightly controlled commercial offerings.
Ian: Well we’ve certainly seen Palantir, Recorded Future et al. work to derive interesting conclusions and predictions from large data sets like this.
Matt: I think the difference here is partly that many users (even if they are data rich) are much less interested in creating/curating data sets than they are in using them. We’ve seen humanities, CIS and engineering groups all derive huge benefits from well-curated third-party data. Getting those groups to create and share their own data too is tough without aligning the process with their academic objectives and the academic recognition system.
Ian: Has anyone cracked that problem in this space?
Matt: The Harvard Dataverse is an attractive platform which hosts data sets and generates benefits for both the contributors and the community as a whole by tracking/reporting which datasets are downloaded via a data DOI.
Ian: Which translates to recognisable impact in academe?
Matt: Absolutely, I had a data set which I was able to show had been downloaded more than 35’000 times. Thats significant impact.
Ian: So lets talk about the NetSci lab at Rutgers
Matt: This is a collaboration between a great team of leading academics in Communication, Information Science, and Journalism who are addressing a wider view of Human Networks interacting through Technological Networks as well as other contexts.
Ian: What is your current focus?
Matt: We are looking at systems of local information that feed/support their communities and how this intersects with the phenomena of misinformation. We’ve mapped the transition to more regional news structure and a steady decline in the production of quality local news (critical information, politics, education, disaster/safety) in favour of less substantial/serious content (sports, human interest etc) which, whilst potentially of interest, does little to support a local communities in more serious situations.
Ian: Do users simply live with less local content as a result?
Matt:In fact, this gap in local news coverage tends to increase the use of (local) social media such as Next Door and Facebook for new, where stories are largely unverified, not edited by a third party and, in some cases, anonymous. This leads to a greater risk that the information provided may be misinformation or even malicious.
Ian: How serious is the potential impact?
Matt: For example we have seen a troubling loss of local news connections between communities and infrastructure providers such that in the event of power outages in adverse weather events there is no longer a trusted independent local news source to disseminate news updates, timetables and disaster response information from the power company to the community but only what potentially poorly informed social media commentators may be saying. We are focused on better understanding the impact of the loss of a robust and trusted connection between physical systems and information systems.
Ian: What could be a potential response to address this disconnect?
Matt: We are considering the process of re-establishing a trust-based relationship between communities and service providers (industrial, government) via trusted intermediaries – a role that quality news/media organisations used to fill.
Ian: This sounds like really interesting work
Matt: We don’t believe we are even close to seeing the potential impact of mis-information – both inadvertent or even the weaponisation of (dis)information as it will continue to affect local and national news and our understanding of the truth.
Ian: Thanks for speaking to me today and welcome to the WSTNet.
The next Brave Conversation will take place on Thursday 20th July at Newspeak House in London.
Brave Conversations London 2023
This is our 22nd Brave Conversations event and our third in London.
At our first in 2018 we had only just launched Brave Conversations and our work with the Web Science Trust was gaining momentum.
Since that time we have been to many different places and spaces but the conversations remain the same – how can we help Humans become Smarter in a world stuffed full of smart technologies?
Brave Conversations London 2023 takes place within the maelstrom of the emergence of G-LLAMs – Gigantic Large Language Models (Tristan Harris and Aza Raskin, the AI Dilemma )
“The biggest threat facing humanity today is humanity in the age of the machines.” (Mo Gawdat)
So what are G-LLAMs and what should we be thinking about as they increasingly pervade our lives – underpinning everything from our waking moments to our sleep.
We will be exploring
- What is the definition of a human in the Age of the Smart Machine?
- What is the role of a human in this Age of the Smart Machine?
- How do we retain our humanity whilst harnessing the power of Smart Machines to improve life for all on Planet Earth?
We will be guided in our conversations by the use of Mark Moore’s Strategic Triangle asking:
- What Can We Do?
- What Should We Do?
- What May We Do?
“We assumed people were going to do good things with it. We really didn’t have any idea of what bad things people would do with it.
Everything that is bad within society is writ large within the Internet. They lose their moral compass in a way.”
Professor Dame Wendy Hall in Interview with Beth Ridley
The goal of Brave Conversations is to challenge everyone who participates – regardless of what background they come from, or what their skillset and expertise are – to more fully explore and understand the interplay between humans, the societies we live in, and the technologies we have created.
We want to empower people to proactively make decisions about how we live our everyday lives, how we participate as commercial actors within the economy, and how we operate as digital citizens and exercise our political rights. That empowerment comes from demystifying data and information and understanding how it informs the everyday decisions which gradually create the future.
Each of those decisions begin on an individual human level – our bodies and our minds – and then emanates out to our families, communities, societies and from there to nation states. We are all responsible for the world we are creating and never has there been a time when we have more potential to influence the changes around us. But we need to be given the space for robust debate and respectful curiosity, learning from each other, playing with ideas, and asking the questions that are both confronting and potentially will take us to uncomfortable places.
Dame Wendy Hall gives an interview on Sky News on the future of AI, the possibility of long-term dangers and how we can pragmatically engage with the risks of disinformation in the short term
“Modernity” is a social, cultural, or historical descriptor for a certain type of society or set of social arrangements. It is a contentious and disputed term, often understood implicitly. It is a way of describing and classifying highly complex, dynamic, and emergent aggregate social phenomena, and so dramatically simplifies such contexts. However, the language of modernity remains attractive to commentators, academics, and policymakers.
In this monograph, the author reviews the literature that characterises what is called digital modernity. Digital modernity narratives focus on the possibilities of the data gathered by an ambient data infrastructure, enabled by ubiquitous devices such as the smartphone, and activities such as social networking and e-commerce. It is characterised by (1) a subjunctive outlook where people’s choices can be anticipated and improved upon, (2) the valorisation of disruptive innovation on demand, and (3) control provided by data analysis within a virtual realm that can be extended and applied to the physical world. The author explored the synergies and tensions between these three aspects as well as the opportunities for and dilemmas posed by misinformation. The author identifies five principles that emerge from the study of relevant texts and business models and concludes by contrasting digital modernity with other theories of the 21st century information society.
Narratives of digital modernity are useful because they help explain the development of technology. It matters because many influential people accept, and often generate, the digital modernity narrative. Given digital modernity’s strong association with the Web, it is a central topic for Web Science as the interdisciplinary study of the World Wide Web from the technological, social, and individual points of view.
Publication Date: 17 Nov 2022