Posts

Blog: Beyond the Chat Box: Rethinking How We Interact with AI

Image
Chat has become the default way of interacting with artificial intelligence. We ask a question, receive an answer, refine the prompt, and continue the conversation. Whether we are searching for information, writing an email, debugging code, analyzing data, or summarizing a paper, the interaction increasingly takes the same form: a text box waiting for us to type/say something. There are good reasons for this. Conversation is familiar. It requires almost no training; natural language lowers the barrier to using sophisticated systems; and a single interface can appear to support tens of different tasks. But as chat becomes the dominant way we interact with AI, a more fundamental question emerges: should a chatting interface be the only way we engage with AI systems? The question is not whether chatbots are useful. Clearly, they are. The more interesting issue is what happens when we treat conversation as the natural form of interaction for almost every AI system, including tasks where u...
Read more

Blog: Climate Democracy: When Participation Strengthens Climate Action and When Conflict Delays It

How climate policy is contested, reshaped, and strengthened through democratic participation Key question: When does democratic participation accelerate climate action, and when can political conflict contribute to delay or backlash? This blog looks at climate change through the intersecting lenses of politics, society, and climate change and forestry, not only as a scientific or technological problem, but also as a democratic one. Climate change is also a democratic challenge What happens when governments know that climate action is necessary, but citizens disagree about how quickly change should happen, who should pay, and who should bear the costs? The transition to a low-carbon, climate-resilient society is a political process as much as a technical one. Governments must decide how quickly emissions should fall, which sectors change first, how communities adapt, and how costs and benefits are distributed. These choices are made through elections, legislatures, local government,...
Read more

Blog: Grokking: When Models defy how we think they learn

Image
A few months ago, in a conversation with some colleagues, was when I first heard the term “grokking”. For a brief moment I thought it had something to do with the AI model Grok, but learned that it is actually a well-known phenomenon in the area of machine learning so I took some time to read about it.  Most of the time, when we are training a neural network model, we expect that the test accuracy to follow up the after training accuracy after some few hundred steps. And when it happens that the test accuracies are not improving over a few more hundred steps we give up. The idea of Grokking suggests that maybe the model is not done yet so give some more time. Grokking is the phenomenon when a model memorizes its training data quickly, sits at chance-level performance on test data for a long period, and then with no change in the setup, just more steps of training and suddenly it generalizes. Figure 1 depicts what grokking looks like. We observe that the training accuracy shoots up ...
Read more

Blog: Finding signal in the noise - Why some forecasts improve and others don‘t

Image
Humans have always tried to predict the future. During the hunter-gatherer era, people predicted things such as plant and food availability, weather, and animal movement. Today, with the rise of AI, prediction has entered a completely new dimension. We are no longer only making predictions ourselves but we are also teaching machines to make increasingly sophisticated predictions. Figure A. Popularity of the search term prediction in Google Trends With the rise of big data, one question becomes central for the art of prediction: How can we find the signal in this vast amount of data and use it to improve our predictions? As I was reading Nate Silver's book "The Signal and the Noise", I became more interested in this question. First, it would be good to define what a prediction is. A prediction can be defined as a specific, testable claim about a future observable outcome.​ A prediction usually has four important attributes: Target: What observable outcome are we predic...
Read more

Blog: Is she cultured: a short intro to cultural studies

Today we talk about CS. Not computer science, but cultural studies. First, let me ask, what is culture? Some consider drinking wine to be more cultured than drinking beer. Or what do you think of when you hear 'she is cultured'? Is it that 'she' is educated, classy and listens to the opera? Or that 'she' is traditional and does not follow western practices? Whatever notion you think of, that is a descendant of your perception of the word culture . Now, what exactly is culture and how do you study it? There are different fields which study aspects of culture, such as border studies, anthropology or history. But today, we focus on Cultural Studies. At its core, cultural studies is interdisciplinary, politically engaged and denies the existence of 'high' and 'low' culture. This field considers culture as something that we do, not something we have or we are. It inherently deals with the questions of power, focusing on identities of gender, sex, clas...
Read more

Blog: vibe code and token maxx like a ceo

Vibe-coding has, of late, transformed from a novelty among programmers into a mainstay of the software industry. Garry Tan, Y Combinator’s president, has returned to hands-on coding for the first time since his early days of founding a software company, and has praised AI’s ability to quickly reproduce his early software product. Mark Zuckerberg has similarly returned to coding , making his first substantive pull requests to Meta’s production code for the first time in 20 years. Some may dismiss vibe-coding as an advancement in slop-making or a phenomenon in C-suite micro-management, but I find myself in partial agreement with Jensen Huang’s comment that he would go “ape something else” if an engineer on a $500,000 salary were to spend only $5,000 in tokens—there have to be some ways in which AI can save your precious time. On a similar note, a resident AI expert at I2SC has similarly cautioned us against sparing tokens: if 10 hours of your work can be eliminated by $100 in tokens, ...
Read more

Blog: Can an AI assistant become part of a research team instead of living in a browser tab?

Image
Fig.1. From OpenClaw to Your Favorite Messaging App At the Interdisciplinary Institute for Societal Computing (I2SC), we recently explored OpenClaw —an open-source, self-hosted AI assistant. It connects to platforms such as Telegram, WhatsApp, Microsoft Teams, Slack, and Discord. Unlike most AI chatbots, it also lets you choose the AI model that best fits your task. One of OpenClaw’s biggest advantages is control . Conversations, memory, API keys, and configuration stay on infrastructure that you manage. The assistant can also run continuously on a server, even when your laptop is turned off. One assistant, multiple channels OpenClaw uses a central gateway between messaging platforms and AI models. The gateway keeps conversation history, runs scheduled tasks, uses browser tools, and executes reusable skills. It also supports multiple AI providers. You can use Claude for complex reasoning, GPT for general tasks, Gemini for Google integrations, or switch models depending on cost and ...
Read more