@theuni@social.tchncs.de

· 1mo ago

Dein Talk ist heute und morgen nicht dabei, weil die Konkurrenz zu groß war? Du hast noch ein spontanes Thema / eine Idee gehabt? Du hast noch nicht viel Erfahrung Vorträge zu halten aber möchtest mal über deinen Schatten springen?

Dafür sind die kurzweiligen Lightning Talks da. 5 Minuten - also 10 Themen Schlag auf Schlag - und ein wohlwollendes Publikum!

Keine Slides nötig (aber möglich). Hauptsache Ihr habt Lust und wollt uns was zeigen, dass ihr spannend findet!

Bei Fragen DM oder gern hier antworten!

[Edit] Typos

bonn.social

4

0

5

0

Theuni @theuni@social.tchncs.de

· 4mo ago

Replying to

@robpike@hachyderm.io (from a DE perspective - and we're always 10 years behind) I'm only 27 years into the industry (give a few years as a teen stumbling around ... well, maybe I STILL am) and just sat in a political/economical conservative AI booyaa-whatever-it-was last Friday due to impact on my state in the near future and the opportunity for some embodied experience. I feel we're screwed in so many ways it's unfathomable. The anti-intellectualism, the pure not-even-thinly-veiled greed in the "SQUIGGLY LINE MUST GO UP" sense gave me an impressive physical reaction. My wearable told me a 65 pulse, so it must have been adrenaline I guess sitting in the middle of old greedy men. I'm not even against ML technology - but why 'o why do we have to pervert EVERYTHING. If you like, take my comisery.

18

0

Theuni @theuni@social.tchncs.de

· 2mo ago

Replying to

@mohs@climatejustice.social @isotopp@infosec.exchange Zwei Gedanken: quick and dirty (und respektlos gegenüber dem Status Quo) braucht es halt im irgendwo mal voranzukommen. Die Leute mit guten Ideen, sind nicht immer die gleichen, die das dann langfristig gut auf die Straße kriegen. VC und reine Open-Source-Nutznießer sehe ich hier als das stärkere Problem. Ich weiß welche Sicherheitsaspekte ich vom Redis erwarte (keine - ab in die hinterste Reihe und sinnvolle Anwendung davor) und dann kann man das so halt auch benutzen.

3

0

Theuni @theuni@social.tchncs.de

· 1mo ago

Weil's so schön war und da plötzlich noch ein Slot im Programm aufgetaucht ist: @FrOSCon@bonn.social Lightning Talks v2, morgen 15 Uhr in Hörsauna 4 (https://programm.froscon.org/froscon2026/talk/ab23a474-36bf-4126-89cb-763eabaec50d/) #froscon21 #froscon2026

Same procedure: ab morgen früh steht das Flipchart beim Infostand und es gibt 10 Slots.

Wenn Ihr Fragen habt, einfach her damit (auch per DM) und wenn ihr euer Setup testen wollt, kommt 15 Minuten früher!

✌️❤️🧸

programm.froscon.org

1

0

2

0

Theuni @theuni@social.tchncs.de

· 5mo ago

Replying to

@grumpygamer went there on friday, stan did a great performance, the jacket was only vertically striped, though! Here‘s a meet and greet photo. There‘s two stans, because i did my first ever attempt at a cosplay (on the left). Aside from the jacket i found the sombrero a real challenge. I wasn‘t competent enough to make my own (like the one on the thimbleweed forums 8 years ago) so i opted for matching color and pattern but had to stick with a smaller size … The show will be on again in september next year - be my guest if you want see it, happy to host you!

8

4

0

Theuni @theuni@social.tchncs.de

· 5mo ago

I always carry an ifixit kit in my backpack and repaired my fair share of aged notebooks, coffee machines and other stuff. However, my favorite episode was fixing a watch that suddenly fell of a fellow train travellors wrist and me whipping out the kit and having it assembled back together within 2 minutes. That was a surprised and grateful look!

5

0

Theuni @theuni@social.tchncs.de

· 3mo ago

Hmm. Mein Wahlkreis ist ja leider aufgrund fehlender Zweitstimmendeckung verwaist. Welchem*r Abgeordnet*en schreibe ich denn da? Würde gern den IfG-Abbau anprangern. Siehe auch @fragdenstaat@chaos.social

1

0

Theuni @theuni@social.tchncs.de

· 26mo ago

Replying to

@larsmb@mastodon.online And obviously it immediately fucks up any arbitrary counting. And then it gives a reference to some site that obviously doesn't have that "word" in it ...

21

13

3

0

Theuni @theuni@social.tchncs.de

· 26mo ago

Replying to

@vladimir_lu@hachyderm.io @larsmb@mastodon.online I'm assuming you're ironic, but there is a point worth discussion. Mathematical models are intended to cover all general cases. Specifically in computer science we've been bitten by exactly the idea of "nobody will ever put this in" (see y2k bug). We've been through this. And on a more anthropological level good math (and physics) gets extended and remains valuable even in shifting contexts. LLMs are garbage in that way.

20

11

2

0

Theuni @theuni@social.tchncs.de

· 26mo ago

Replying to

@larsmb@mastodon.online It's fantastic. Here's the FastGPT version of it. It greatly demonstrates that it doesn't have the ability to count / compute but only combines textual artefacts.

20

4

3

0

Theuni @theuni@social.tchncs.de

· 5mo ago

Replying to

@hynek@mastodon.social Would you mind waking me up when all is said and done? I'll be under a rock until then, I guess ... -_-

1

0

Theuni @theuni@social.tchncs.de

· 5mo ago

Replying to

@grumpygamer yeah, travel would be a bit insane for the event alone … (i have to admit i did travel to mexico from DE for a 3 day wedding 20 years ago, but yeah, that one was tough) If you do travel accidentally at that time - the offer will stan(d)!

1

0

Theuni @theuni@social.tchncs.de

· 26mo ago

Replying to

@larsmb@mastodon.online I think we can now consider LLMs the pool noodles of cloud or something.

9

5

2

0

Theuni @theuni@social.tchncs.de

· 26mo ago

Replying to

@geospacedman@mastodon.social @vladimir_lu@hachyderm.io @larsmb@mastodon.online There are no C in DNA sequences. This is an often made confusion as there are only G, T and A in DNA as the C looks much alike the G.

7

5

0

Theuni @theuni@social.tchncs.de

· 26mo ago

Replying to

@TomSwirly@toot.community @geospacedman@mastodon.social @vladimir_lu@hachyderm.io @larsmb@mastodon.online I was joking by impersonating an LLM answer.

3

2

0

Theuni @theuni@social.tchncs.de

· 26mo ago

Replying to

@larsmb@mastodon.online I wasn't either. I sometimes use FastGPT because it gives at least some indication of where it got its data from so I can go from there and look things up myself ...

3

4

0

Theuni @theuni@social.tchncs.de

· 26mo ago

Replying to

@fl0_id@mastodon.social @larsmb@mastodon.online Which is fine by me. That's basically what a search engine does. FastGPT is done from a search engine perspective, so it's basically a bit of summarisation and then pointing you somewhere. It doesn't even do continued conversations.

1

0

Theuni @theuni@social.tchncs.de

· 38mo ago

Replying to

@qbi@mastodon.social war ein Manta in der Nähe?

1

0

Theuni @theuni@social.tchncs.de

· 44mo ago

Replying to

@daisy Thanks for the reminder. A practical inquiry: getting the change properly propagated "just" needs some careful implementation. However, paper trails that need to be/stay sealed won't get updated. Progress is better than perfection, I guess, or is there something I'm missing?

0

2

0

Theuni @theuni@social.tchncs.de

· 5mo ago

Replying to

@xahteiwi@mastodon.social I've spend 6 (or maybe nearly 9) months trying to get ollama/llama-cpp/vllm working properly with AMD workstation GPUs (W7900). it's been tiring. Model support is hit and miss, stability is not that, and performance doesn't scale. We had to cut our losses (for now) and spend money with nvidia to unblock us... :/ ... would love to revisit this at some point, but right now ... it's like the readme says.

0