CEO/Principal Consultant at INNOQ, he/him, software architect, RESTafarian, conference tourist. Works at innoq.com. Fediverse: @stilkov@innoq.social

Germany
Joined April 2007
Wie wirken sich LLM-Technologien auf wichtige Architekturthemen aus? Welche neuen Ansätze für Design und Architektur werden möglich? Wie verändern sich Architekturaufgaben? In dieser Keynote zeichnet @st_toth ein Zukunftsbild und liefert jede Menge Inspiration 🙌
4
8
Stefan Tilkov retweeted
“Stop treating your frontend as some generic API client, and start treating it as a half of your app.”
a common comment/question: if you use hypermedia as your web application API, what about other clients? good news, we've got some essays on that: htmx.org/essays/splitting-yo… htmx.org/essays/hypermedia-a… TLDR: your webapp & general API needs are different, so splitting them up is good!
1
10
91
Stefan Tilkov retweeted
This is weird to say out loud, but I actually am kinda an expert in rate limiting, so I'm gonna explain some stuff. About half of incidents in large-scale production systems involve having more requests than you can serve. There are two categories of this kind of incident:
164
2,274
374
10,650
Stefan Tilkov retweeted
Ein Journalist fragte mich in Sachen neuer #Antisemitismus-Studie, ob ich ihm die Top 10 meiner schlimmsten antisemitischen Begegnungen von 2022 nennen könne. Wie bizarr und traurig eine solche Top 10 wäre... Okay, here we go! 1/11
133
533
67
1,740
Stefan Tilkov retweeted
Replying to @bettersafetynet
literally yes
7
249
27
1,203
Literally every single thing he has done just sucks completely, and worst of all, is being copied by others. To even think I had respect for that guy once makes me cringe
1
17
I really shouldn’t care, but I’m still mad at what that fool @elonmusk did and continues to do to this site. It feels a bit as if a friend got a spouse that I just can’t stand, and they are changing them for the worse
2
5
1
33
Stefan Tilkov retweeted
this is cool
Today, we released the first piece of our vision: Timecraft, an application runtime built on top of #WebAssembly and with a Time Machine at its core. There is a lot to say but for now checkout the readme, try it and let us know what you think about it! github.com/stealthrocket/tim…
2
12
Stefan Tilkov retweeted
I think this is mostly right. - LLMs created a whole new layer of abstraction and profession. - I've so far called this role "Prompt Engineer" but agree it is misleading. It's not just prompting alone, there's a lot of glue code/infra around it. Maybe "AI Engineer" is ~usable, though it takes something a bit too specific and makes it a bit too broad. - ML people train algorithms/networks, usually from scratch, usually at lower capability. - LLM training is becoming sufficently different from ML because of its systems-heavy workloads, and is also splitting off into a new kind of role, focused on very large scale training of transformers on supercomputers. - In numbers, there's probably going to be significantly more AI Engineers than there are ML engineers / LLM engineers. - One can be quite successful in this role without ever training anything. - I don't fully follow the Software 1.0/2.0 framing. Software 3.0 (imo ~prompting LLMs) is amusing because prompts are human-designed "code", but in English, and interpreted by an LLM (itself now a Software 2.0 artifact). AI Engineers simultaneously program in all 3 paradigms. It's a bit 😵‍💫
🆕 Essay: The Rise of the AI Engineer latent.space/p/ai-engineer Keeping up on AI is becoming a full time job. Let's get together and define it.
157
715
74
4,117
Stefan Tilkov retweeted
Lerne mit @kammer_anja Wege zur Realisierung dynamischer Cloud-Native-Architekturen. Ihr #iSAQB CPSA-A-Training #CLOUDINFRA findet vom 01.-03.08. online statt – Infos und Anmeldung: socreatory.com/de/trainings/… #softwarearchitektur
1
5
5
#Sonneberg #AfD #Neonazis #NSU Mir fehlen immer noch die Worte! Simon Pearce hat sie bereits vor Jahren gefunden: Gedicht: „Bei Hitlers brennt noch Licht“ #BeiHitlersBrenntNochLicht #NoAfD #fckafd 1/4
16
525
36
1,078
0
Stefan Tilkov retweeted
This, then, is the challenge: LLMs give us the feeling of understanding. Sometimes, that is enough. But, when it is not enough, and we persist in the illusion, we enter dangerous territory.
29
38
2
346
Stefan Tilkov retweeted
"Do LLM understand?" is a question that yields passionate answers. As for me and my house: no, LLMs do not reason and in fact are architectural incapable of reasoning. Let's unpack that.
62
336
50
2,138
Stefan Tilkov retweeted
Large Language Models: verändern sie alles? In Folge 130 sprechen @stilkov und @marcelweiss über Chancen für die Wirtschaft, Einsatzmöglichkeiten, aber auch Risiken. Hört rein 👉 innoq.com/de/podcast/130-lar…
3
9
0
Stefan Tilkov retweeted
Repeat after me: Blocking paste on a form textbox is not a security feature.
24
111
9
713
Stefan Tilkov retweeted
BTW, Twitter, if you're looking for someone to run an engineering org or a DevRel org, I should also mention that my own hat is in the ring. Funemployment has been... um... fun... but I'm ready to start on something new.
4
1
6
Stefan Tilkov retweeted
Wir kriegen gerade endlos viele wütende Mails von AfD-Fans und das gefällt uns außerordentlich gut 😂 #fckafd #zsk #hassliebe #keinbockaufnazis
543
1,573
115
11,534
0