← Home
About
This project is my (Joel Galvez) latest attempt (out of three) to find a way to know what going on on Saturday. I quit Facebook in 2018 and for the most part it was a relief, but I missed having an overview of what was going on.
The events are collected in two ways, one good and one easy. The good way is what I have been trying in the past years, that is to collect event data through the open protocol .ics (iCalendar). It was almost successful at times. The easy way (added now in 2024) is to run some newsletters through a locally hosted language model (AI). This way I get a good overview of what happens tomorrow and the day after, and so on. This is the first moment I can say that it truly works for me, personally.
AI has some drawbacks, but it solves the biggest problem: There is no threshold to get started. One aspect of this approach, that I personally like, is that I can't easily force anything. If an organisation deletes my email from their recipients list, that's the end of it.
The question I have now is what to do with this. Partly, I wonder if it's okay to publish these events extracted with AI here. I suspect it depends on how its done. Taking something that is meant for one context and putting it another can be problematic. On the other hand, event-data is so sparse and "non-contextual" (essentially, title, date, image) it might be okay. Depending on the reactions I get, my plan might become to make this into a "funnel" to the fediverse. This would allow me to both share and view events from other sources.
This particular website is my own collection of events and event sources within the arts. It is centred around Amsterdam, since that's where I live. It's not meant to be a big collection. If you would like to make your own collection, let me know. You can also host your own server.
The events are collected in two ways, one good and one easy. The good way is what I have been trying in the past years, that is to collect event data through the open protocol .ics (iCalendar). It was almost successful at times. The easy way (added now in 2024) is to run some newsletters through a locally hosted language model (AI). This way I get a good overview of what happens tomorrow and the day after, and so on. This is the first moment I can say that it truly works for me, personally.
AI has some drawbacks, but it solves the biggest problem: There is no threshold to get started. One aspect of this approach, that I personally like, is that I can't easily force anything. If an organisation deletes my email from their recipients list, that's the end of it.
The question I have now is what to do with this. Partly, I wonder if it's okay to publish these events extracted with AI here. I suspect it depends on how its done. Taking something that is meant for one context and putting it another can be problematic. On the other hand, event-data is so sparse and "non-contextual" (essentially, title, date, image) it might be okay. Depending on the reactions I get, my plan might become to make this into a "funnel" to the fediverse. This would allow me to both share and view events from other sources.
This particular website is my own collection of events and event sources within the arts. It is centred around Amsterdam, since that's where I live. It's not meant to be a big collection. If you would like to make your own collection, let me know. You can also host your own server.