Tag: technology

Posted on

AI Racing

AI has entered the racing of cars after we have been racing horses, dogs and camels for many decades. The fact behind all these races is the huge market for gambling. Anything you can bet on will do for juicy profits in that industry. The recent “Abu Dhabi Autonomous Racing League” is the latest addition to the racing craze. Moving online with 600000 spectators at its peak on video and gaming platforms the investment seems promising. The only problem, AI is not yet ready to really compete with the world of real drivers. The progress, however, is astonishing. Just one lap of 2 minutes on the circuit yields 15 Terrabyte of data from 50 sensors. These are closed circuits so no person can enter or animal can get in their way. The challenge to integrate more data and faster processing as well as algorithms for fast decision making is steep. Great learning opportunities for advances in robotics. The hype has not been able to live up to the expectations as no real racing took place yet. We have replaced the gladiators of the Roman empire with Formula 1 drivers. It is only fair to retire those drivers soon and let AI race cars against each other. It feels like a computer game on screen and it is as we shall most likely watch these races on a screen as well. Hence, what is the point. Watching youth on TWITCH play racing games will probably not change the viewing behavior of the masses. The programmers have nevertheless great learning opportunities and will find their way rapidly into the job market. The other challenges of ASPIRE seem more important for humanity like human rescue and food for the growing world population. In the meantime let the boys play around with cars and learn about potentials as well as failures of AI-programmers and dealing with both.

Posted on

AI Disruption

Many scientists started to question the disruptive potential of AI in, for example, the military’s domain. The Journal of Strategic Studies featured 3 papers on AI and autonomous systems more generally. The major argument by Anthony King is the reliance of autonomous systems on other systems mainly human operators even in the background to get these systems off the ground and maybe back again. Not only logistic support but also satellite communication is needed to guide and protect the operations. In quoting Clausewitz, Anthony King stated that war is a “collision of two living forces”. Strategy and counter-strategy will co-evolve as will attack and defence.
Jackie G. Schneider and Julia Macdonald (2024) advocate the use of autonomous and unmanned systems for their cost effectiveness. Economic costs as well as political costs are lower for these new strategic weapons. Mass fire power from swarms of drones is much cheaper than nuclear warheads and the home electorate is assumed to be more willing to accept and support limited and more precisely targeted unmanned missions. The disruption potential of AI is huge but it is most likely an addition to the arsenals than replacing them. (Image 2 swarms of drones fly in the air above tanks, created by AI – copilot-designer 2024-4-29).

Posted on

Hannover Fair

The annual science fair at Hannover is a kind of a show of things to touch and of those things that come to the public market in the near future. Most of the annual hype is about potentials of production. Rationalization, using few resources or innovative solutions of digitization are high on the agenda. Create your digital twin, save energy, make production more safe or cyber secured.
Robotics is another reason to visit the fair. Some 7 years ago I had my Sputnik experience there. The robotics company KUKA had demonstrated live the that assembling a car from pre-manufactured components takes just 10 minutes for the robots. Shortly afterwards the whole company was bought by Chinese investors. Roughly 5 years later we are swamped by cars from China. It was not that difficult to predict this at that time. Okay, we need to focus on more value added production and take our workforces (not only) in Europe along on the way. Reclaiming well-paid, unionized jobs in manufacturing, as Joe Biden does, will not be an easy task. Robots and their programming is expensive, but skilled workers, too. Hence, the solution is likely to be robot-assisted manufacturing as a kind of hybrid solution for cost-effective production systems.
Following the proceedings of the 2024 fair we are astonished to realize that visiting the fair is still a rather “physical exercise” walking through the halls. After the Covid-19 shock we expected a lot more “online content”. Instead we keep referring to webpages and newletters rather than virtual visits and tours. The preparation of the visit in advance remains a laborious adventure. However, the in-person networking activities in the industry are largely advanced by ease of exchanging virtual business cards and the “FEMWORX” activities.
This year’s Sputnik moment at Hannover is probably most likely related to the pervasive applications of AI across all areas of the industry and along the whole supply chain. Repairing and recycling have become mainstream activities (www.festo.com). Robotics for learning purposes can also be found to get you started with automating boring household tasks (www.igus.eu).
Visiting Hannover in person still involves lengthy road travel or expensive public transport (DB with ICE). Autonomous driving and ride sharing solutions might be a worthwhile topic for next year’s fair. Last year I thought we would meet in the “metaverse fair” rather than in Hannover 2024. Be prepared for another Sputnik moment next year, maybe.
(Image: Consumer’s Rest by Stiletto, Frank Schreiner, 1983)

Posted on

AI Defence

For those following the development in robotics we have been astonished by the progress of, for example, rescue robots. After an earthquake such robots could enter a building that is about to collapse and search the rooms for survivors. A recent article in “Foreign Affairs” by Michèle A. Flournoy has started its thinking about the use of AI in the military with a similar 20 year old example. A small drone flying through a building and inspecting the dangers of entering for persons or soldiers. Since then technology has advanced and the use of AI for automatic detection of dangers and “neutralising” it, is no longer science fiction. The wars of today are a testing ground for AI enhanced military strategies. It is about time that social scientists get involved as well.
Warfare left to robots and AI is unlikely to respect human values unless we implement such thoughts right from the be beginning into the new technology. An advanced comprehension of what algorithms do and what data they are trained on are crucial elements to watch out for. According to Flourney, AI will assist in planning as well as logistics of the military. Additionally, AI will allow a “better understanding of what its potential adversaries might be thinking”. Checking through hours of surveillance videos is also likely to be taken over by AI as the time consuming nature of the task binds a lot of staff, that may be put to work on other tasks. Training of people and the armed forces become a crucial part of any AI strategy. The chances to develop a “responsible AI” are high in the free world that cherishes human rights and democratic values. Raising curiosity about AI and an awareness of the dangers are two sides of the same coin or bullet. Both need to grow together.
(Image created by Dall-E Copilot Prompt: “5 Robots disguised as soldiers with dash cams on helmet encircle a small house where another robot is hiding” on 2024-4-23)

Posted on

Digital Estonia

The progress of Estonia in going digital is quite advanced. The electronic identity card which allows data to be linked to health data and accounts or banking gives an impression of how far-reaching digitalization may go. Great steps have been taken to guide the population on the way to move towards the digital (only) world. Learning and coaching of a huge amount need to take place so that people do not abandon or get lost on the path towards “everything digital”. For the so-called digital natives, who have grown up with the sound of their smartphone at the bedside all the time, this move feels “natural”. Some experienced or silver workers got on track, if they were accompanied in suitable forms. The 65+ population might find it harder to adapt to the permanent use of digital devices for not only getting around in your city, but also to do your tax declaration, pay your dues and vote in elections.
Digitalization is not a goal in itself. It has advantages to reach communities in remote places or islands, but it might alienate older persons that have no other person around to assist them in the digital only world. An easy way to get some social science data to inform the debate is to refer to Eurostat and the surveys with information about the “overall life satisfaction” of people (EU-SILC). Checking for some major countries of the EU and neighbours of Estonia with less digitalization the differences are rather small. In terms of overall life satisfaction (16+ years old) Estonia has been catching up to the EU-average mainly between 2013 and 2021. Since then, stagnation at the EU-average is what the data tell. A quick testing of the hypothesis that the older persons (65+) might not see the past evolution as rosy is reflected in the EU-data as well. Good pensions seem to drive the “happiness” of older persons in the EU more than good digitalization. Eventually the two features of a society will have to go hand in hand to improve life satisfaction to higher levels. (Image: Data Eurostat EU-SILC Life satisfaction 65+, selected countries 2013-2023, retrieved on 2024-4-23, comparison with table all ages here, Data source)

Posted on

AI Reader

In the middle of the hype around AI it is useful to take stock of the reflection and evolution of AI. In my own analyses and writings on AI it evident that a narrowing of focus has taken place. Whereas before 2022 the writing dealt more with digital technologies in general. The links to the literature on the social construction of technologies was obvious. Algorithms and AI was a part of the broader topic of society and technology.
This has changed. The public debate is focused on “everything AI now”. We look at technological developments largely through the lens of AI now. Hence, my focus of assessments of technology from a societal perspective follows this trend. In a collection of blog entries on AI we try to demonstrate the far reaching changes that have started to have an impact on us. In the last few months the all encompassing concern about AI’s effect on us needs full attention of social scientists, policy makers, companies and the public at large. We can no longer leave this topic to the software engineers alone. By the way, they themselves ask us to get involved and take the latest advances in AI more seriously.
As a “flipbook” the online reading is rather comfortable (Link to flipbook publisher MPL). The pdf or epub files of the blog entries allow to directly follow the links to sources in webpages or other publications (AI and Society 2p 2024-4-18). The cycles of analyses and comments have become faster. Traditional book writing suffers from time lags that risk to make pubications outdated rather quickly. Dynamic ebook writing might bridge the gap between time to reflect and speed to publish or inform the wider public. The first update as .pdf-file is available here: AI and Society(2).

Posted on

AI and PS

AI like in ChatGPT is guided by so-called prompts. After the entry of “what is AI” the machine returns a definition of itself. If you continue the chat with ChatGPT and enter: “Is it useful for public services” (PS), you receive an opinion of AI on its own usefulness (of course positive) and some examples in which AI in the public services have a good potential to improve the state of affairs. The AI ChatGPT is advocating AI for the PS for mainly 4 reasons: (1) efficiency purposes; (2) personalisation of services; (3) citizen engagement; (4) citizen satisfaction. (See image below). The perspective of employees of the public services is not really part of the answer by ChatGPT. This is a more ambiguous part of the answer and would probably need more space and additional explicit prompts to solicit an explicit answer on the issue. With all the know issues of concern of AI like gender bias or biased data as input, the introduction of AI in public services has to be accompanied by a thorough monitoring process. The legal limits to applications of AI are more severe in public services as the production of official documents is subject to additional security concerns.
This does certainly not preclude the use of AI in PS, but it requires more ample and rigorous testing of AI-applications in the PS. Such testing frameworks are still in development even in informatics as the sources of bias a manifold and sometimes tricky to detect even for experts in the field. Prior training with specific data sets (for example of thousands of possible prompts) has to be performed or sets of images for testing adapted to avoid bias. The task is big, but step by step building and testing promise useful results. It remains a challenge to find the right balance between the risks and the potentials of AI in PS.

Posted on

AI and text

The performance of large language models (LLMs) with respect to text recognition and drafting texts is impressive. All those professions that draft a lot of texts have often decades of experience with using word-processing software. The assistance of software in the field of texts ranges from immediate typo corrections to suggestions of synonyms or grammatical corrections in previous word-processing software.
The improvement of AI stems for example from the potential to suggest alternative drafts of the text according to predefined styles. A very useful style is the “use of easy language”. This rewriting of texts simplifies texts in the sense that longer and more structured sentences are split into shorter ones, lesser-known words or acronyms are replaced by more common or simpler words. Some languages like German have a particular need to use easy language when it comes to administrative regulations and procedures. Public services that aim for inclusiveness of for example older persons or youth can become much more accessible if the use of easy language is spread more widely. Just keep in mind the large numbers of so-called “functional illiterates” (OECD study “PIAAC”) in all OCED countries.
AI can do a great job in assisting to reach a broader public with texts adapted to their level of literacy and numeracy competences. Webpage Designers have made use of Search Engine Optimization (SEO) for years now. The most common way is to use frequently searched keywords more often on your website in order to be found more often by search engines like GOOGLE et al. Additionally, AI allows to explain keywords, sentences or even jokes to you (Spriestersbach 2023 p.111). This may help in situations when cross-cultural understanding is important.
We have made use of optical character recognition (OCR) for a long time now in public services as well as firms and for private archives. AI is taking this “learning experience” to the next level by making use of the content of the recognized text. Predicting the following word or suggesting the next sentence was only the beginning of AI with respect to texts. AI can draft your speech to plead guilty or not guilty in a court. But we shall have to live with the consequences of making exclusive use of it rather than referring back to experts in the field. AI please shorten this entry, please!

Posted on

AI by AI

It has become a common starting point to use electronic devices and online encyclopedias to search for definitions. Let us just do this for artificial intelligence. The open platform of Wikipedia returns on the query of “artificial intelligence” the following statement as a definition: “AI … is intelligence exhibited by machines, particularly computer systems …“. It is not like human intelligence, but tries to emulate it or even tries to improve on it. Part of any definition is also the range of applications of it in a broad range of scientific fields, economic sectors or public and private spheres of life. This shows the enormous scope of applications that keeps rapidly growing with the ease of access to software and applications of AI.
How does AI define itself? How is AI defined by AI? Putting the question to ChatGPT 3.5 in April 2024 I got the following fast return. (See image). ChatGPT provides a more careful definition as the “crowd” or networked intelligence of Wikipedia. AI only “refers to the simulation” of HI processes by machines”. Examples of such HI processes include the solving of problems and understanding of language. In doing this AI creates systems and performs tasks that usually or until now required HI. There seems to be a technological openness embedded in the definition of AI by AI that is not bound to legal restrictions of its use. The learning systems approach might or might not allow to respect the restrictions set to the systems by HI. Or, do such systems also learn how to circumvent the restrictions set by HI systems to limit AI systems? For the time being we test the boundaries of such systems in multiple fields of application from autonomous driving systems, video surveillance, marketing tools or public services. Potentials as well as risks will be defined in more detail in this process of technological development. Society has to accompany this process with high priority since fundamental human rights are at issue. Potentials for assistance of humans are equally large. The balance will be crucial.

Posted on

AI Sorting

Algorithms do the work behind AI systems. Therefore a basic understanding of how algorithms work is helpful to gauge the potential, risks and performance of such systems. The speed of computers determines the for example the amount of data you can sort at a reasonable time. Efficiency of the algorithm is an other factor. Here we go, we are already a bit absorbed into the the sorting as purely intellectual exercise. The website of Darryl Nester shows a playful programming exercise to sort numbers from 1 to 15 in a fast way (Link to play sorting). If you watch the sorting as it runs you realize that programs are much faster than us in such simple numeric tasks. Now think of applying this sorting routine or algorithm to a process of social sorting. The machine will sort social desirability scores of people’s behavior in the same simple fashion even for thousands of people. Whether proposed AI systems in human interaction or of human resource departments make use of such sorting algorithms we do not know. Sorting applicants is a computational task, but the data input of personal characteristics is derived from another more or less reliable source. Hence, the use of existing and newly available databases will create or eliminate bias. Watching sorting algorithms perform is an important learning experience to be able to critically assess what is likely to happen behind the curtains of AI.

Posted on

AI and dialect

The training of Large Language Models (LLM) uses large data sets to learn about conventions of which words are combined with each other and which ones are less frequently employed in conjunction. Therefore, it does not really come as a surprise that training which uses standardised languages of American English might not be as valid for applications that receive input from minority languages or dialects. The study forthcoming in the field of Computer science and Language by Hofmann et al. (Link) provides evidence of the systematic bias against African American dialects in these models. Dialect prejudice remains a major concern in AI, just like in the day-to-day experiences of many people speaking a dialect. The study highlights that dialect speakers are more likely to be assigned less prestigious jobs if AI is used to sort applicants. Similarly, criminal sentences will harsher for speakers of African American. Even the more frequent attribution of death sentences for dialect speakers was evidenced.
If we translate this evidence to wide-spread applications of AI in the workplace, we realise that there are severe issues to resolve. The European Trade Union Congress (ETUC) has flagged the issue for some time (Link) and made recommendations of how to address these shortcomings. Human control and co-determination by employees are crucial in these applications to the world of work and employment. The need to justify decision-making concerning hiring and firing limit discrimination in the work place. This needs to be preserved in the 21st century collaborating with AI. The language barriers like dialects or multiple official languages in a country ask for a reconsideration of AI to avoid discrimination. Legal systems have to clarify the responsibilities of AI applications before too much harm has been caused.
There are huge potentials of AI as well in the preservation of dialects or interacting in a dialect. The cultural diversity may be preserved more easily, but discriminatory practices have to be eliminated from the basis of these models otherwise they become a severe legal risk for people, companies or public services who apply these large language models without careful scrutiny.
(Image AI BING Designer: 3 robots are in an office. 2 wear suits. 1 wears folklore dress. All speak to each other in a meeting. Cartoon-like style in futuristic setting)

Posted on

Energy Storage

On a sunny and windy day, even in winter or spring, renewable energy is abundant. If demand is stable prices will drop. Prices will rise again as demand for energy picks up. Hence, this is an obvious case for trading opportunities. All you need is … energy storage. All so-called prosumers, short for producers and simultaneously consumers have a lot to gain if they are able to store energy when it’s abundant and cheap. Sell it when it is expensive or use it yourself if needed. Just keep an eye on the costs of energy storage. A stylish insulated carafe is a well known example of storing hot water for astonishingly long time. Insulation is key to store transformed electric energy here. Other options use kinetic energy like pumping water to a higher level and then generate electricity again when the water returns to the lower level. Of course, batteries are a simple way for energy storage as well. Costs seem to come down rapidly and less environmentally hazardous materials leave the laboratory almost every month. It is about time to consider this seriously. More and more cities have understood that energy storage can generate cash for them (Example Feuchtwangen) and appears to be a worthwhile investment for a local power generating community. For the time being my favorite energy storage is the insulated carafe. It is often the beginning of energizing conversations.

Old school energy storage
Posted on

AI Collusion

In most applications of AI there is one system of AI, for example a specialized service, that performs in isolation from other services. More powerful systems, however, allow for the combination of AI services. This may be useful in case of integrating services focusing on specialized sensors to gain a more complete impression of the performance of a system. As soon as two and more AI systems become integrated the risk of unwanted or illegal collusion may occur.
Collusion is defined in the realm of economic theory as the secret, undocumented, often illegal, restriction of competition originating from at least two otherwise rival competitors. In the realm of AI collusion has been defined by Motwani et al. (2024) as “teams of communicating generative AI agents solve joint tasks”. The cooperation of agents as well as the sharing of of previously exclusive information increase the risks of violation of rights of privacy or security. The AI related risks consist also in the dilution of responsibility. It becomes more difficult to identify the origin of fraudulent use of data like personal information or contacts. Just imagine using Alexa and Siri talking to each other to develop another integrated service as a simplified example.
The use of steganography techniques, i.e. the secret embedding of code into an AI system or image distribution, can protect authorship as well as open doors for fraudulent applications. The collusion of AI systems will blur legal borders and create multiple new issues to resolve in the construction and implementation of AI agents. New issues of trust in technologies will arise if no common standards and regulations will be defined. We seem to be just at the entry of the new brave world or 1984 in 2024.
(Image: KI MS-Copilot: Three smartphones in form of different robots stand upright on a desk in a circle. Each displays text on a computer image.)

Posted on

AI input

AI is crucially dependent on the input it is built on. This has been already the foundation principle of the powerful search engines like Google that have become to dominate the commercial part of the internet. The crawling of pages on the world wide web and classifying/ranking them with a number of criteria has been the successful business model. The content production was and is done by billions of people across the globe. Open access facilitates the amount of data available.
The business case for AI is not much different. At the 30th anniversary of the “Robots Exclusion Standard” we have to build on these original ideas to rethink our input strategies for AI as well. If there are parts of our input we do not AI to use in its algorithms we have to put up red flags in form of unlisting parts of the information we allow for public access. This is standard routine we might believe, but everything on the cloud might have made it much easier for owners of the cloud space to “crawl” your information, pictures or media files. Some owners of big data collections have decided to sell the access and use to their treasures. AI can then learn from these data.
Restrictions become also clear. More up-to-date information might not be available for AI-treatment. AI might lack the most recent information, if it a kind of breaking news. The strength of AI lies in the size of data input it can handle and treat or recombine. The deficiency of AI is not to know whether the information it uses (is in the data base) is valid or trustworthy. Wrong or outdated input due to a legal change or just-in-time change will be beyond its scope. Therefore, the algorithms have a latent risk involved, i.e. a bias towards the status quo. But the learning algorithms can deal with this and come up with a continued learning or improvement of routines. In such a process it is crucial to have ample feedback on the valid or invalid outcome of the algorithm. Controlling and evaluating outcomes becomes the complementary task for humans as well as AI. Checks and balances like in democratic political systems become more and more important.