Mario Angst

My profession (research) is just becoming increasingly unhinged in terms of "AI" use. Last three days:

• Day 1: Meeting on a grant proposal. Colleague has not read the call, but has Google Gemini Chatbot open and asks the Chatbot during our meeting constantly about call specifics.
• Day 2: Another meeting about a possible grant proposal (yes, my life). Colleague has not read the call, but brings an "AI" summary they read while discussing it with me.
• Day 3: Preparation for a kick-off meeting on a project. Colleague has prepared a list of tasks and timeline with Claude. I dread looking at it.

#academia #research #academicchatter

And you know, I even use LLMs to generate text for computer code sometimes, so I am not generally opposed to tightly scoped use of generated text in some places, but this is just an avalanche of carelessness by now.

RE: https://social.anoxinon.de/@SNF_ch/116000038278544878

"More demand, but less money available" is hardly ever what you want to hear if you are trying to find funding for research. This is how the Swiss National Science Foundation deals with it in its most popular funding scheme (project funding).

Should be interesting news here for researchers in #Switzerland. Essentially, the already announced chances considered to project funding are now formally here:

• each researcher can have only one funded project running anymore at a time (plus one Lead/ Weave/ ICIS)
• for re-applying to the project funding after rejection, you now need to wait a year
• funding for salaries in the project capped at 200k/year (250k/year without salaries)

I do think, given the circumstances, these are sensible changes by the @SNF_ch@social.anoxinon.de - putting a bit of a brake on the overheating submission carousel. Some things that I find interesting to see now are:

• Will we be seeing more shenanigans with people being designated PIs for projects others basically write for them (already a problem anyhow with the eligibility rules and not being able to fund your own salary)?
• Will we be seeing more international collaboration applications within the Weave and Lead Agency schemes in particular?
• Will overall project sums for disciplines that are mainly requiring personnel, not equipment, actually go down? For example, I am a social scientist - beyond computing and travel, I mostly need people for projects, not microscopes.

#research #swissresearch #academia

Hey Fedi #academicchatter / #research people,

What are you using (or considering) as Overleaf alternatives? The writing has been on the wall with overleaf in terms of enshitification for a while now, and I'd rather have a headstart. Also, I am not publishing or reviewing with Elsevier/RELX, so should get rid of Overleaf at some point anyhow.

I've dabbled in hosting https://github.com/overleaf/overleaf myself, but I'd rather pay an ok company to stand up this type of tooling for me, to be honest.

#academicwriting #LaTex #overleaf

If I see actually good research published on MDPI, I just feel sad by now. Don't do it, people, I'd love to cite your stuff, but I won't.

Morgen sind Wahlen in der Stadt Zürich. Als Politikwissenschaftler bin ich objektiv ziemlich gespannt wie eine Stadt wählt in diesen Zeiten.

Subjektiv zittere ich mit einigen Bekannten mit. Hoffentlich kommen wir einen Schritt weiter in Richtung Mobilitätswende, Wohnen und nachhaltiger Stadtentwicklung. ❤️ Züri ❤️

Oh wow... tears in my eyes? How did this happen?
#StarFleetAcademy Episode 8

Es ist 2026, Menschen hungern immer noch auf der Welt, aber immerhin kann ich Betty Bossi fragen, ob sie mir Python Code generieren kann.

My first real crush as a teen and all the color in her scarves. The human rights activist who was so soft-spoken and strong. My colleague who smiles through the pain and keeps on doing the best science. The PhD student who is so smart you feel frightened.

The Iranian people I have met in my life have only ever been the kindest and most beautiful people.

And I know it makes no sense asking it, but I just cannot understand why the world would make them suffer like this. Not that anyone deserves suffering, but I cannot help but think of them.

Good news, as well as blog post title of the year :)?

Crew and Ground Teams Successfully Troubleshoot Orion’s Toilet

https://www.nasa.gov/blogs/missions/2026/04/02/artemis-ii-flight-update-crew-and-ground-teams-successfully-troubleshoot-orions-toilet/

I like that preparing for the solar-powered website/ "how much energy does the internet use" workshop I am teaching tomorrow at the UZH children's university has also given me a reason to add a low-resource version of my own homepage, which only serves a simple about page.

I've implemented it as an initial choice on the root address, which now gives the options to visit a minimal, the full English, or the full German version of the website.
https://marioangst.com

This has actually gotten me thinking about this design pattern (if it can be called that) - an internet that serves a minimal page with 20% of the information of the website that 80% percent of visitors are probably after as a default might not be such a bad idea.

Technically, I am using a #Quarto profile that only renders the index page at / and a separate quarto website rendered to /minimal.

Gerade wieder einen #AIHype Artikel in einer grossen Schweizer Tageszeitung überflogen. Inhalt nicht der Rede wert, ging darum, dass KI für Hobbymusiker irgendwie mega toll sein soll, weil "du singst und dann macht die KI Backing-Musik" oder sonst etwas, was auch nur Leute mit eigentümlichem Bezug zu Musik schreiben können. Wohl Leute, die auch Freude an einem Fussballspiel unter Robotern hätten oder so, wenn die das cool finden, warum nicht.
Spannender war, dass schon wieder dieses eigentümliche Wort "Sparring Partner" auftaucht. Die "KI als Sparring Partner", ich sehe das in allen möglichen Kontexten momentan (inklusive in der Lehre), woher kommt diese plötzlich Fixierung auf Boxen oder gegen wen kämpft man als Hobby-Musiker denn nun genau?

Sehr toll aufgearbeiteter Beitrag zu Abriss versus Neubau vom SRF: https://www.srf.ch/play/radio/redirect/detail/8bf8cecd-cc61-3472-a460-8782aeb28b33