Category: AI Ethics

Posted on

Biases and Stereotypes in LLMs

The paper “Revisiting the Trolley Problem for AI: Biases and Stereotypes in Large Language Models and their Impact on Ethical Decision-Making” by Sahan Hatemo, Christof Weickhardt, Luca Gisler (FHNW School of Computer Science), and Oliver Bendel (FHNW School of Business) was accepted at the AAAI 2025 Spring Symposium “Human-Compatible AI for Well-being: Harnessing Potential of GenAI for AI-Powered Science”. A year ago, Sahan Hatemo had already dedicated himself to the topic of “ETHICAL DECISION MAKING OF AI: An Investigation Using a Stereotyped Persona Approach in the Trolley Problem” in a so-called mini-challenge in the Data Science degree program. His supervisor, Oliver Bendel, had told the other scientists about the idea at the AAAI 2025 Spring Symposium “Impact of GenAI on Social and Individual Well-being” at Stanford University. This led to a lively discussion. The student recruited two colleagues, Christof Weickhardt and Luca Gisler, and worked on the topic in a much more complex form in a so-called Challenge X. This time, three different open-source language models were applied to the trolley problem. In each case, personalities were created with nationality, gender, and age. In addition, the data was compared with that of the MIT Moral Machine project. Sahan Hatemo, Christof Weickhardt, and Luca Gisler will present their results at the end of March or beginning of April 2025 in San Francisco, the venue of this year’s event.

Posted on

Androids for Different Areas of Application

RealDollX and Realbotix have been known for years for Harmony, one of the most advanced sex robots in the world, known not least for the movie “Hi, AI”. Prof. Dr. Oliver Bendel was able to meet and talk to the robot head at a workshop in 2019. GPT was used, as explained by Kino Coursey from Realbotix, who was present and co-developed Harmony. Oliver Bendel was able to persuade him to write the article “Speaking with Harmony“, which was published in his book “Maschinenliebe” in 2020. Realbotix has since broadened its focus. The company develops social robots – mainly female androids, but also male models – for interactive presentations, among other things. Movable arms and hands are used for this. The company’s new flagship is Aria, which caused a stir at a trade fair in 2024. The movable arms and hands could also benefit Harmony, for example when hugging. Above all, however, it enables gestures that support the already convincing facial expressions. In future, the speech capabilities of the general models will be created using several speech models. Further developments will be discussed at a meeting in March 2025, for which the teams from Texas, Nevada, and California will come together in one place (Photo: Realbotix).

Posted on

kAIxo says “kaixo”

The final presentation of the “kAIxo” project took place on January 9, 2025. Nicolas Lluis Araya was the project team member. The FHNW School of Business has been developing chatbots for dead, endangered, and extinct languages for several years. A well-known example is @llegra, a chatbot for Vallader. In the spring of 2024, Oliver Bendel tested the reach of GPTs for endangered languages such as Irish (Irish Gaelic), Maori, and Basque. According to ChatGPT, there is a relatively large amount of training material for them. On May 12, 2024 – after Irish Girl and Maori Girl – a first version of Adelina, a chatbot for Basque, was created. It was later improved in a second version. As part of the “kAIxo” project (the Basque “kaixo” corresponds to the English “hello”), the chatbot kAIxo was built, which speaks Basque. Its purpose is to keep users practicing written or spoken language or to develop the desire to learn the endangered language. The chatbot is based on GPT-4o and Gemini 1.5 Flash, and the user can select his or her preferred large language model (LLM). Retrieval-augmented Generation (RAG) plays a central role. The ChatSubs dataset is used, which contains subtitles of movie dialogs in Basque. Thanks to a text-to-speech engine, the chatbot can also speak. At the final presentation, Nicolas Lluis Araya presented a working prototype that can be accessed via www.kaixo.ch.

Posted on

Role Player Nina

Around the turn of the millennium, there were already numerous pedagogical agents, i.e., chatbots, voice assistants, and early social robots in learning environments. Names such as Virtual Learning Companions (VLCs) were also commonly used for the purely virtual versions. One of the companies that was ahead of its time was Extempo Systems, Inc. based in Redwood City, California. The company had evolved out of the Stanford Engineering School and designed and developed commercial agents for entertainment, business, and education: “Extempo makes e-learning products that help corporate em­ployees perfect their people skills. The company’s approach allows every learner to achieve mastery of the people skills they need for effective management, teamwork, sales, customer service, and other critical business functions. Extempo’s products give learners authentic practice in a variety of job-specific conversational role-plays, along with expert individualized coaching throughout the learning process.” (Extempo Systems) One example was the virtual Nina, with whom you could communicate and who you were supposed to motivate to behave in a certain way. It wasn’t just the functionality that was impressive at the time, but also the design of the characters.

Posted on

The kAIxo Project

The interim presentation of the kAIxo project took place on 11 November 2024. Nicolas Lluis Araya is the project collaborator. Chatbots for dead, endangered, and extinct languages are being developed at the FHNW School of Business. One well-known example is @llegra, a chatbot for Vallader. Oliver Bendel recently tested the reach of GPTs for endangered languages such as Irish (Irish Gaelic), Maori, and Basque. According to ChatGPT, there is a relatively large amount of training material available for them. On 12 May 2024 – after Irish Girl and Maori Girl – a first version of Adelina, a chatbot for Basque, was created. It was later improved in a second version. As part of the “kAIxo” project (the Basque “kaixo” corresponds to the English “hello”), the chatbot or voice assistant kAIxo is being built to speak Basque. Its purpose is to keep users practising written or spoken language or to develop the desire to learn the endangered language. The chatbot is based on GPT-4o. Retrieval-Augmented Generation (RAG) plays a central role. A ChatSubs dataset is used, which contains dialogues in Spanish and three other official Spanish languages (Catalan, Basque, and Galician). Nicolas Lluis Araya presented a functioning prototype at the interim presentation. This is now to be expanded step by step.

Posted on

AI Ethics at the FHNW

Prof. Dr. Oliver Bendel has been teaching information ethics, AI ethics, robot ethics, and machine ethics at the FHNW for around 15 years. He is responsible for the “Ethik und Technologiefolgenabschätzung” (“Ethics and Technology Assessment”) module in the new Business AI degree program at the FHNW School of Business in Olten. Here, the focus is on AI ethics, but students will also learn about robot ethics and machine ethics approaches – including annotated decision trees and moral prompt engineering. And they will use information ethics, including data ethics, to analyze and evaluate the origins and flows of data and information and engage in bias discussions. Last but not least, they will delve into technology assessment. Oliver Bendel also teaches the “Ethik und Recht” (“Ethics and Law”) module in the Business Information Systems degree program at the FHNW School of Business in Olten (which he took over in 2010 as “Informatik, Ethik und Gesellschaft”, later renamed “Informationsethik”), the “Recht und Ethik” (“Law and Ethics”) module in the Geomatics degree program at the FHNW School of Architecture, Construction and Geomatics in Muttenz, and “Ethisches Reflektieren” (“Ethical Reflecting”) and “Ethisches Implementieren” (“Ethical Implementing”) in the Data Science degree program at the FHNW School of Engineering in Brugg-Windisch. His elective modules on social robotics are very popular (Photo: Pati Grabowicz).

Posted on

Continuation of the AAAI Spring Symposia

On 27 August 2024, AAAI announced the continuation of the AAAI Spring Symposium Series, to be held March 31 – April 2, 2025, at San Francisco Airport Marriott Waterfront in Burlingame, CA. The Call for Proposals for the Spring Symposium Series is available on the Spring Symposium Series website. According to the organizers, proposals are due October 4, 2024, and early submissions are encouraged. “The Spring Symposium Series is an annual set of meetings run in parallel at a common site. It is designed to bring colleagues together in an intimate forum while at the same time providing a significant gathering point for the AI community.” (Website AAAI) The traditional conference will therefore not be held at Stanford University in 2025 – as it was in 2023. It returned there in 2024 to the delight of all participants. The Covid-19 pandemic had hit the conference hard before. The AAAI can only be advised to return to Stanford in 2026. Only there will the conference live up to its promise.

Posted on

Keynote by Anthony Elliott

On the second day of Robophilosophy 2024, Anthony Elliott, Distinguished Professor of Sociology at the University of South Australia, gave a keynote speech entitled “May AI Be With You: Agency and Automation in the Age of Algorithmic Modernity”. From the abstract: “From industrial robots to ChatGPT, and from driverless cars to military drones: AI is transforming all aspects of our lives, from the changing nature of work, employment and unemployment to the most intimate aspects of personal relationships. In this presentation, Anthony Elliott focuses on the complex systems of AI – spanning intelligent machines, chatbots, advanced robotics, accelerating automation, big data – and their centrality to new forms of social interaction, organizational life and governance. He argues, provocatively, that today modernity has come to mean smartphones, tablets, cloud computing, big data, automated recommendation systems and predictive analytics. This has heralded the arrival of what he terms ‘algorithmic modernity’, an altogether new ‘stage’ in the ordering techniques of envisioned human mastery. In this automated order of algorithmic modernity, human agency is increasingly outsourced to smart machines. We should understand this phenomenon, Elliott argues, in terms of a containment of both uncertainty and complexity which the digital revolution in social relations poses, but which ultimately denies answers.” (Website Robophilosophy 2024) In his presentation, the sociologist emphasised the risks of AI rather than the opportunities. The photo shows David Gunkel announcing Anthony Elliott’s lecture – he is sitting on the far left.

Posted on

Keynote by Wendell Wallach

On the first day of Robophilosophy 2024, Wendell Wallach, one of the fathers of machine ethics, gave a keynote speech entitled “Re-envisioning Ethics: From Moral Machine to Extensive Regulation”. From the abstract: “Have we been underestimating the socio-technical challenges posed by ro(bot)s – physical systems and virtual bots? Many of the complexities inherent in managing intelligent systems can not be adequately met by scientific innovation, existing ethical constraints, or weak regulations forged by legislatures under the capture of the AI oligopoly.  In spite of naive future projections, the science we have, and are likely to have in the near future, will not produce AI systems capable of making even satisfactory choices in complex situations where uncertainty reigns, multiple values converge, and the information available is inadequate to project meaningful consequences for various courses of action. AI will pose safety and security risks far beyond those being addressed by the pittance of investment presently directed to build trustworthy systems. Scientific humility is needed. Ethics must be reenvisioned and empowered to work through the plethora of socio-technical obstacles and trials ahead. A vast infrastruture to ensure AI safety will be required.” (Website Robophilosophy 2024) Wendell Wallach went far beyond these hints and gave an overview of the advances and setbacks in machine ethics and AI ethics in recent decades.

Posted on

An AI-generated Teen Collection

Spanish fashion chain Mango has launched an advertising campaign created using AI models. First, all the clothes were photographed, then an AI model was trained to place the images on artificially generated models. The images were then retouched and edited. This was reported by Golem in an article dated July 30, 2024. Digital models are not new. They have been used for decades in various contexts, from computer games to mobile phone applications. Cameron-James Wilson founded an agency for digital models in London in 2019. He is the creator of the digital supermodel Shudu. Her sisters are Noonoouri and Lil Miquela. Since the triumph of generative AI, models have moved into another league. There are now beauty pageants for AI-generated models. The winners of Miss AI 2024 are Kenza Layli (Morocco), Lalina (France), and Olivia C (Portugal). They are already successfully represented on Instagram, showing themselves in various poses and dresses. According to Golem, Mango plans to sell the advertised collection in 95 countries. With this step, the company aims to reduce the costs that would otherwise be incurred for photographers, models and the entire production process (Image: DALL-E 3).