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.)
POSSE
for a German speaking audience POSSE stands for a farsical popular drama of the 18th and 19th centuries. Nowadays POSSE has particularly in the English speaking world a completely different meaning. POSSE is short for Publish on your Own Site, Syndicate Everywhere. In other words publish your own creative work, text, tech or melody on your own website first before spreading the content on other platforms. By this strategy you make sure your content stays your content even if a platform like eXitTwitter disappears or the hype has passed like with facebook. And if at any point in time you decide the whole internet is a farce, you will be able to edit the content from your own webpage as a book or audio version as an edited volume. Do not make yourself a modern slave of tobacco, drugs, alcohol or platforms for whom you create (mostly unpaid) content they use to sell advertisements with huge untaxed profits. With POSSE you can make an end to this farce. We still rely on search engines like QWANT, Ecosia, duckduckgo or Mozilla’s Firefox, to name a few less well known web searching tools. They ensure that other people find your webpage beyond the elephant in the room called Google.
Mastodon
In the Post-Twitter-world there is room for other platforms to take over the information function and opinion exchange on an international level. Mastodon has made a good start to fill this gap. The number of registered users has surpassed the 10 million users by large. If the anger about the union-busting and Trump supporter Elon Musk increases further, more and more people will join Mastodon as users and content providers. Advertisers will follow where they can reach their appropriate audiences. If you sell seats on a spacecraft to the moon you will advertise on eXit-Twitter. If you care about responsible investment you advertise on elsewhere or on Mastodon, because that is where your audience will be. Ethically responsible advertisers started to make such moves already. The decentralised structure to potentially filter human rights violating hate messages will need to grow accordingly on this platform. This is a challenge for all platforms and involves costs. To better understand the idea behind Mastodon, think of a platform which connects many other local networks or platforms. Yes, it is a platform of platforms. Decentralised instances provide and can remove access to the local platform. Even whole sub-platforms might be shut out. This increases the potential to build a “caring platform”. Decentralised organisation often incurs higher costs than a central instance that decides and administers a one-size fits all solution. But it is worth a try.
An example makes is more clear, what are the advantages of such decentral organisation to achieve more shielded privacy. For example, a small local chess club targeting primarily youth can open its own server on mastodon and admit only its known members to the network. It is easy to control who gets access to your posts of solutions to riddles or innovative game and learning strategies. Access to other Mastodon instances is an additional option. For enterprises or larger organisations, it needs the communication department to talk to the IT-service to get such a structure on its way. A dedicated person needs to monitor what content gets posted and what kind of “traffic” is generated. But this can largely be automated. In case you know your members of the local platform it is much less likely to occur than on the eXit-Twitter or other centralised huge international platforms.
Many federated structures will find it useful and part of their social-DNA to join the FEDIVERSE. We shall have to wait for some more time to get the so-called social media platforms civilised. These are social investments and learning processes which deviate from the profit-maximising rationale of other platforms. (Image: screenshot from @mastodon-user on 2023-11-30). 
Over a generation the association with the word mail or AI has changed. Younger generations will automatically think of email as the obvious association with “you have got mail”. Asking people for their mail address, most people will respond immediately with their email account. That has certainly changed over the last 20 years. Email has become extraordinarily important for contacts, content and “crime”. The original set-up of mail servers were supposed to exchange data and information between trusted and trusting individuals. Nowadays we have become “anyone” on an email-list or part of a cascading email-chain. Pishing emails that try to lure us to potentially fake webpages to enter personal information is widespread. A whole new industry of cybersecurity has evolved in parallel with ever faster pingpong of new threats and costly remedies. Most critical remains the human factor to protect email and vital information from abuse. All training to better manage email should therefore begin with awareness building on the need to take cybersecurity seriously right from the beginning. It is not an issue to deal with towards the end of learning about it. Some general points have become common practice. Think carefully if you need to open the mail. Check whether there are external links in it. Do you really need this extra information? Be careful about the number of persons you forward or put in the copy field of your mail.
Unfortunately the simple mail transfer protocol (SMTP) makes it technically fairly easy to abuse email addresses and send an email from a domain name which you do not own. Same holds true for the content of emails. Originally only basic letters and numbers were allowed, no Umlaut etc. But then the “multipurpose internet mail extensions” (MIME) enabled more formats of attachments and illicit links and files within the email content.
In consequence it became necessary to scan emails en masse for malware and obnoxious content. Hence the quality of your email programs reveals itself with the amount of spam you still receive. Next come the email filters that sort your inbox for spam and other modern plagues like unwanted newsletters or notifications.
In consequence we sometimes wonder that we never received certain emails albeit the person sending it reiterated that the email was sent. Even checking your spam folder in the email program might not solve the issue because your receiving email server has been instructed not to accept emails from previously unknown mail servers. Not much you can do about this as an individual person, you mail server is just not on the so-called white list from which to accept incoming emails.
With a mail address from the big national or international companies you have little to worry about in this respect. Smaller companies or organizations might suddenly face a thorny issue if their i@xy.z address appears on a dubious list. You will end up sending but nobody receives your emails. It might take you quite some time to find out about it and even more to fix the issue. Hence the take home message is, have more than one email address registered on different mail servers to check that sending out and receiving emails works properly for you. Managing an own webpage with email service for members or employees is a nice service, but beware of the implications for cyber security as well.
Peace and AI
Rather than asking AI to draft a peace treaty, I used AI to generate images to illustrate my blog entry on strategic thinking and peace deals. My own bias for impressionistic images in art have guided my choice previously. The alternative suggestions from AI based on BING reveals the progressive as well as stereotypical creation of images through algorithms. Same gender in all images, even if the women only image is rather progressive, but as a matter of fact women still tend to be involved less in warfare. The racial stereotypes of AI in image creation also needs attention as the 2 POC persons are depicted in an unfavorable way, not one of strength as for the caucasian stereotype. Living with AI is a joint learning process, likely to be a long one, too. Critical assessment of output remains a human task and we need to train people how to critically and carefully analyze the flood of images in addition to photos.
Eyes
So-called deep learning algorithms can assist us in lots of routine tasks. Their applications seem to be spreading more rapidly than we commonly believe. The recent paper published in “Lancet Digit Health 2023; 5: e257–64” shows the relevance for medical screening again. Beyond skin or breast cancer, digital images have for years or decades been subject to studies using deep learning algorithms to early detection of cancer and other diseases. Now Google has published an additional potential application of algorithmic learning that assists in diagnosing blood sugar levels among others. For diabetes scanning photographs of the external eye with professional equipment (Zeiss Cirrus Photo 600) was used in the study, but the outlook hints already at the potential for most recent smartphone cameras to also be able to capture images of even better resolutions.
Just like measuring our weight and muscles as a weekly routine on scales, we shall have a more scientific look at our selfies soon. Spotting early onsets of several diseases will become a normal feature.
Early adopters of the new technology and screening potential of the smartphones might live even longer. Reminders to adopt or keep a healthy life style will be all around us. In some cultures, it is deeply offensive to look straight into the eyes of another person. We always knew that our eyes might tell a lot about us. Beyond mental health, they tell a lot about physical health, too.
Artists have always told us fascinating stories about eyes or how and what they see with their eyes. Scientists confirm that there is a lot of information in the eyes. (Image extract from MAD Paris, Picasso, Schiaparelli). 