Recently we had the discussion, if a specialized KI with limited data access to "trusted information" only would be a desirable tool on the ELF website (which targets patients as primary audience).
After some cups of coffee I'm confessed, that this is not a solution which receives my support. This is a strong message, but it has been developed that way. And what are my arguments?

1. I'm not a specialist for LLMs like chatgpt, perplexity, droracle, or whatever someone uses. But from my few attempts to get answers as "simple patient", I have to confess, that all answers are pretty fair good, even if generated from non-specialized training data.
2. Results on specific questions are evolving quickly, which means, that training data are enhancing. A reasonable amount of these "training data" is generated by prompts to specific subjects, which allows the KI to better match expectations on content and structure of results. If we would have a specialized KI with only few traffic, we would not generate these additional training data and results would fall behind evolving standards.
3. For curiosity (and in preparation of our annual congresses) I have been asking chatgpt, how the care of patients in Austria could be improved. A have been asking at different times and this are the changing/evolving results (please have your browser to translate the German text): Question 3 (June 2025), Question 4 (Sept. 28th), Question 5 (Sept. 29th).
   (I know, that's not a typical patient's question, but the reason is to demonstrate evolution from this example in my files. I use the result of chatgpt as a reasonable description of the state of the art when discussing in our medical working group, therefore I have this kind of questions in my files.)
   There is really evolution of results for the community, because I always operate KIs in anonymous mode.
4. Even if results are increasingly well (and always better than I would be able to summarize), there are nearly always weaknesses or empty phrases, which arise from the publication bias in existing literature. But how to cope with that, when asked about sensitive subjects?

Basically a specialized KI would be a next generation of Doctor Google, which frustrated many doctors with reason. Nevertheless, I have a lot of rare disease patients at hand, which found their diagnose themselves with support from Doctor Google, despite of a long and ineffective journey at a long chain of physical doctors.
Therefore I'm confessed, that people will ask any tool easy at hand - and that's OK. A specialized KI on a site with few traffic will not attract a significant number of patients, even if results are good (find droracle.ai as example).

My credo is, that the majority of patients (or citizens) will ask common KIs (like chatgpt) and there are only two instances, which can or have to handle gaps and glitches of the results: doctors and patient organizations.
If we want to improve this situation, it would be highly recommended to foster visibility/competences/recognition of patient organizations. And, of course, everything which improves health competency of patients esp. in their condition, is highly assistant.

Concerning monetarization we are all fishing from the same pool. Historically grown we are operating in different claims, mostly defined by (national) borders. But there are also international sources of wealth and gratuity we are trying to tap.
There rises not only a potential competing situation, when a new child appears on the playground - like the situation with ERF could be interpreted.
It could also be a competitive situation, when established donations from ELF to patient organizations have to face some other work branch - a specialized KI - competing for resources and rising the question, if we would not have the resources better allocated somewhere else.

Don't get me wrong here. I don't see much competition for myself, because I'm competing only for travel bursaries and free event access. Additionally my activities tend to fade to resolution, a doctor would say.

This is a common phrase from business, because it would ruin the business model of your established customer.
The more ELF tries to reach out for patients directly, the more we have to discuss and define the intended relationship between patients, patient organizations and ELF.

Please recognize: ELF doesn't support patient organizations to reach out for patients, it supports only POs internals. You have to conclude that from the disclaimer on the ELF site, that all material is property of ELF (this includes even the materials we provided like the patient charter) and might be distributed only in POs internally or to be used by individuals for personal use only.

Again: Don't get me wrong. Everything, which happens in the field of the orphaned disease sarcoidosis, is very welcome. But we should be transparent and clear about the claims we are claiming.
A "don't care about the disclaimer" isn't helpful.

Last year I wanted and this year I really tried to create a poster for our annual congress of pneumonologists advertising for cooperation with or outreach for ERN Lung. And I failed miserably so far. There isn't a regular member of ERN Lung from Austria and accordingly there is some omnipresent distinct position. I thought, it would be instrumental to describe how doctors could benefit from ERN Lung for their patients.
And did I say - I failed miserably so far.

My idea was, that the upcomming CRC sarcoidosis would be a nice anchor to list options to benefit.
But in reality it's not that easy. The Sarcoidomics CRC was lauched by the ERS, but we tend to conect it to ERN Lung often without any further information (visible to me - should ask gpt?).
Every year, when we try to gather references, what ERN Lung has done, we face the problem, that ERN Lung doctors are also active members of ERS - and mostly their indicated affiliation is ERS and not ERN Lung, this problem isn't new.

Back to Sarcoidomics: I gues it is intended by activists as resource vehicle to foster activities of or in ERN Lung.
If that is true, I have to confess, that I cannot see the scope of activities.
While the medical part is told to be very exclusive (the circle of CN sarcoidosis members only), the outreach for individual patients seams infinite inclusive.
This activity spans are incompatible in operation view, because it neglects the structures in between. In addition to the elaboration of relationship between ELF and POs above, we have to add the complete medical society outside the ERN Lung CN members to our consideration and also the consideration customer of your customer-relation should receive attention.
Here - again - I am without idea, what relevant information I should put on my poster for the medical congress, because all potential viewers are excluded.
What I am afraid to do, is to make a poster just expressing that ERN Lung is so bright and shining without exhibits, saying with other words, that the reader is less and dump without a chance to get out of his/her dark.
Any ideas for my poster?