Deus ex machina

The term “deus ex machina” used to be applied more in its figurative meaning. With the rise of digital tools like chatbots, facilitated and enhanced through AI, God is speaking to us not only in multiple languages, but also from our pockets through our smartphones and headsets. This is a rather recent form of “deus ex machina”, which we did not expect some years ago. The bible as e-book or pdf-file has been around for some decades, but only more recently we can enter conversations with God through chatbots as another version of “deus ex machina “ about almost everything (and pay for it via digital credit card). Programming of such an AI-tool is easily achieved. AI will prepare a weekly or daily sermon or prayer for you, following your predilections of your favourite quotes of the bible. An interesting twist to the programming is to use authorized as well as unauthorized translations of the bible across several centuries.
Another interesting enlargement of the input data base is the inclusion of interpretations and discussions not only within your own religious community, but beyond. Maybe the discussion of several different religious chatbots with each other could prevent aggressions due to differences in basic beliefs. These “dei ex machina” might further our understanding of what makes us humans different from machines and machine-based solutions of human conflicts.
As genetic clones of ourselves have become already technically more feasible, our digital alter-egos (the comprehensive collection of traces in the internet and digital images, plus social scoring) help to empower those “dei ex machina”.
This kind of “Brave New World” asks us to be rather brave ourselves.
(Image: interior St Denis Basilique Cathedral Paris 2024)

Chatbot Me We

In order to dig deeper into the functioning of AI, I deemed it expedient to construct, for example, a simple chatbot on a limited knowledge base from my own writings on AI (link to reader in previous blog entry here).
A toolbox from Google offers powerful assistance in such an endeavour. The outcome uses only my input text and no other sources. It is dynamic in the sense that it interprets questions and searches within the text file provided only. The answers are edited with a LLM (large language model) and provide flawless English texts. You can try it here using catchat as magic formula and Google account so far.
With a bit of programming knowledge (htlm, python, Java) and related learning sites it is feasible to come up with a “static” chatbot hosted at a free of charge provider as well. For learning purposes this step by step building and coding of a chatbot is helpful. The outcome is rather limited or requires a lot of time to increase the scope of Q & A interactions and to move from a static (predefined Q & As) to dynamic ones.
Full control of answers, excluding any hallucinations and high-speed replies, come at a cost. Take a look here. It is a very basic version so far, just to get the idea of it. full web address:
https://schoemannchatbot.eu.pythonanywhere.com/

Chatbot Me

Chatbots are helpful to allow queries to larger data sets like the blog entries here. So here is a try of a Chatbot to query all entries on AI using ChatGPT to create a Chatbot that uses and references it source from www.schoemann.org/tag/ai and the AI reader in pdf-format.
Please send me an email if the hallucinations of this Chatbot 1.0 on AI from a social science perspective are giving strange results. I’ll get back to you. Please use at your own risk as I cannot guarantee for all answers. The usual disclaimer applies here.

ChatGPT proposed the following set of Questions and Answers on the blog for an entry into the chat: Example Q&A with the chatbot

Q: What are the social science concerns with AI?
A: Bias in results, job shifts, democracy risks, privacy, and new inequalities.

Q: What does the text say about reinforcement learning?
A: It’s seen as the next step for AI: focusing on learning and reasoning, not just predicting text. It also uses fewer resources.

Q: How are robots described in the document?
A: Robots are mostly assistants. They can follow people or carry small items, but more complex tasks need sensors and AI training.

Q: What about biased results?
A: Studies can be misleading if control groups are flawed. AI faces the same challenge — social scientists warn: “handle with care”.

Q: What is Schoemann’s blog view on AI?
A: He links AI to energy use, fairness, and its role in the “all-electric society” — stressing efficiency and responsibility.

More on the chatbot (in testing phase) and the Link to the coding help received from ChatGPT on this mini-test-project :
https://chatgpt.com/share/68c1d160-0cc0-8003-bf04-991b9e7c3b24

 

AI Podcasting Me

Content producers have lots of tools at their disposal to get their content across to very different audiences. For some time the traditional media of newspapers, radio and TV were the prime outlets for content distribution. Social media have changed this to many more senders of content than before.
In the 21st century, AI allows to automate media productions. In a trial run I just used Google’s NetbookLM to generate 3 podcasts based on my own writings on AI over more than a year by now. The result is available and using artificial voices it is possible to broadcast yourself without revealing your own personal voice. I am not done with the evaluation of the outcome(s) yet, but the first impression is an interesting other form to spread content.
More tests are necessary to check for hallucinations as well.
Here are the links to my virtual podcasts:
AI, intimacy and insecurity

AI, Society and the Human Spirit

AI and the Human Mosaic: Navigating Our Interconnected Future

Video Doku by AI

Based on my own blog on this webpage “schoemann.org” Google NotebookLM creates a video of about 7 minutes. Using Microsoft Clipchamp automatic subtitles with a slightly different storyline are produced based on the video data. In the end, the blog entries are re-modelled into something like a lecture on “AI in a wider social context” (see and play below). No voice layover so far, read by yourselves. A podcast format is another option.
It feels like walking across landscapes in my own mind. Content creators of today or the past never imagined the impact they might have through the powerful tools of AI. The only caveat, jokes I incorporated into the texts cannot really be handled by AI tools unless they are explicitly designated as such. These AI tools take me much more seriously as I do myself. This is serious.

Mind Map Me

AI tools are great to assist learners in the task to get more structure into larger documents or books. It is up to the teachers or lecturers to use the tools themselves to pre-structure content they want other persons to learn. Mind maps are useful to summarise larger content and offer a tree-like structure to a text moving from the general to more specific content and then into details by at the same time not loosing sight of the overall structure of the content. Basics can be provided by Google’s NotebookLM and you may rework this basic structure yourself linking the mind map to the detailed content. Learning may start with a comprehensive mind map at the beginning to move on to details. Alternative versions of a mind map are equally feasible to come up with new combinations of subjects. This can be done using the tags of the blog entries in addition to the categories and fast search keywords.
It is a fascinating way to mind map yourself based on longer texts written by yourself. This clarifies a bit what potential readers or learners are likely to retain from reading your document(s) or blog entries.
Its worth trying, as others will most likely use these tools from now onwards anyway, whether we like it or not.
Image: Google NotebookLM Mind Map of my AI blog entries 2025-9-9.

Own AI related entries

The digital drafting of blog posts facilitates the publishing of volumes based on a selection of entries. Pdf-files, or EPUB formats are quite common. The choice of a flipbook is more suitable for larger screens of tablets or notebooks. It also serves as a rapid stock taking of entries and missing topics to cover a broader field. In some fields AI is not yet popular, for example in agentic home applications, whereas in other fields text, translation or image production AI has grown into a threat to professional fields. Even computer languages are a major field of application with substantial negative employment effects.
Hence, enjoy the latest comments on AI towards the end of my latest Flipbook.
Just click here.

ChatGPT provides a summary of my AI-reader as follows:

From the uploaded text AI from a Social Science Perspective (Dr. Klaus Schoemann), social science concerns with AI focus on potentials and limitations of how AI interacts with society:

  • Labour and employment: AI impacts jobs, but adjustment processes are slow. For example, there’s a shortage of engineers and care workers, while robots are still limited to simple assistant tasks. This raises concerns about who benefits and who is left behind.
  • Bias and fairness: The book highlights issues of biased results in research and decision-making. When AI systems rely on incomplete or selective samples, results can be misleading — a challenge also familiar to social scientists.
  • Democracy and accountability: Journalists and scientists are framed as part of a “checks and balances” system. AI adds complexity here, as the fast spread of information without accountability can weaken democratic processes.
  • Human–machine relations: Topics like intimacy, housing, and memory are re-thought in the context of AI and digitalization, raising questions about privacy, social norms, and cultural heritage.
  • War and security: AI-driven technologies like drones change the nature of warfare, raising ethical and political questions.

On his blog, Schoemann also stresses the energy and resource costs of large AI models and how reinforcement learning offers a more efficient alternative. He warns that AI should be understood not just technically, but as part of an “all-electric society” where social sciences help assess risks, inequalities, and long-term consequences.

👉 In short: AI raises social science concerns about equity, bias, democracy, labour, cultural shifts, and global risks.

Former Crypto currency

The rationale behind a Crypto currency is quite an ancient idea. In the absence of a central bank or in addition to it, people may choose to establish their own currency or currencies. The central idea is to secure assets (currently and in the future) as well as to enable an exchange between a community of people. The Museum of the National Bank of Belgium (image below) has a “Yap stone” on display which originates from the Pacific Yap Islands. Even very heavy and big stones served as wealth storage. The idealistic value of a particular stone conveyed additional value. In the true value of the word these people were banking on stone money in a way not to distant from what Crypto currency embodies today.

AI in Central Banks

Yes of course, Central Banks will use AI, and some do so already (Kazinnik and Brynjolfsson, 2025). Beyond the standard application of AI by its employees, there are many potentials to use AI to analyse and publish data at a faster rate or in order to detect financial crimes. Similarly, data collection based on webpage harvesting might yield new indicators of inflation, expenditure for environmental risks (heat waves, flooding etc.) earlier and in addition to the normal set of indicators. Hence, Central banks might be better and faster in forecasting inflationary tendencies using more AI tools in their daily routines. Of course, it is difficult to predict a disruptive tariffs policy of a major economic player in the world economy, but the calculation of more, even hallucinatory scenarios become more feasible. It is feasible to weigh overall risks of different scenarios to the economy.
(Image: Celtic coins, Museum of the Belgian National Bank)

AI employment effects

The first robust empirical evidence about employment effects of AI in the USA has been published by the Stanford Digital Economy Lab by  Chen, Chandar and Brynjolfsson (2025). A previous paper by Wang and Wang (2025) highlighted the comparative advantage of persons who use AI in their work compared to others and the authors coined the term “learning by using technology”.  The prediction of the model was that there might be job losses of more than 20% in the long run and half of this already in the first 5 years of the introduction of the technology. The Stanford economists have estimated with real world data these effects in the USA and find quite surprisingly that the negative employment effects of AI have the strongest impact on young labor market entrants with few years of labor market experience. Middle-aged and more senior employees seem to benefit from “tacit knowledge” about the work, which is more difficult to replace with AI, at least for the time being of the early days of AI. This evidence is based on recent payroll data from the largest payroll processing firm “ADP” in the USA which has firms overrepresented from the manufacturing and services industries as reported in another paper  (Firm size maybe another source of bias).  However, the effect that youth 22-25 years of age suffered the most calls into question the common belief that older workers are more likely to suffer the consequences as during in the rise of the digital economy around the year 2000. (AI Image created with Canva)