iiinova — where your digital self keeps evolving

Tim König @koenigt.bsky.social · 09/04/2025

"All else being equal, being able to squeeze more predictive juice out of flawed theories and inadequate paradigms will help them to stick around for longer, impeding true scientific progress."

020

Tim König @koenigt.bsky.social · 09/04/2025

"Actual scientific progress happens through theories, which explain a collection of findings, and paradigms [...] We suspect that the rapid proliferation of scientific findings based on AI has not accelerated — and might even have inhibited — these higher levels of progress."

nature.com

Why an overreliance on AI-driven modelling is bad for science

Without clear protocols to catch errors, artificial intelligence’s growing role in science could do more harm than good.

120

Reposted by Tim König

FragDenStaat @fragdenstaat.de · 28/03/2025

Moment mal... ...so schnell werdet ihr uns nicht los. Zusammen Informationsfreiheit stärken: fragdenstaat.de/jetzt-spenden/

Screenshot eines Artikels der Frankfurter Rundschau, der lautet: "Philipp Amthors Vorschlag? CDU plant, das Informationsfreiheitsgesetz zu kippen: FragDenStaat vor Aus"

136782

Tim König @koenigt.bsky.social · 25/03/2025

Give it a shot and let me know if you run into any issues. It should come in handy when you want your agent to retrieve literature for you ("What articles do I have on LLMs and computational social sciences?"), or for full-blown RAG systems, where access to your library may improve generated answers

100

Tim König @koenigt.bsky.social · 25/03/2025

While playing around with the LangChain Framework, I figured it might be convenient to let LLM Agents access your Zotero library, so I built a retriever It should become part of the langchain_community package soon, but it in the meantime it's available via Github or pip github.com/TimBMK/langc...

github.com

GitHub - TimBMK/langchain-zotero-retriever: A Retriever for the Langchain eco system to access the Zotero API

A Retriever for the Langchain eco system to access the Zotero API - TimBMK/langchain-zotero-retriever

150

Reposted by Tim König

Jeff (Hot Type) Jarvis @jeffjarvis.bsky.social · 12/03/2025

In @zeit.de, @ewo.name's devastating report from #SXSW: Like it's 2023, "back when one could still innocently predict the future at tech conferences without taking into account burgeoning fascism." Again Zeit's coverage of our fall is better than our own.

04716

Tim König @koenigt.bsky.social · 10/03/2025

It doesn't matter if people use GPT, Llama, or Gemini, or if their chatbots lose money. It doesn't matter if Deepseek's architecture reduces compute requirements, as the cost decrease will likely increase demand. To integrate LLMs at scale you need compute, and the money is in providing it. 10/10

000

Tim König @koenigt.bsky.social · 10/03/2025

So the large players like AWS, Microsoft etc. don't invest into these data centers to push the development of AI. They're smarter than that. Their play is that with the proliferation of LLMs across businesses (and again, applications are aplenty), the demand for compute will rise. 9/10

100

Tim König @koenigt.bsky.social · 10/03/2025

When deployed at large, however, even the smaller models require compute often unattainable for all but the largest providers. Due to economics of scale, only large data centers can provide at a price point profitable for both provider and (third-party) user. Hence, the need for data centers. 8/10

100

Tim König @koenigt.bsky.social · 10/03/2025

However, there's not just the proprietary models, but also a great open source LLM ecosystem. You could argue that the investments into proprietary models will not pay off, as there are equally good, open source solutions (think windows/linux). This remains to be seen. 7/10

110

Tim König @koenigt.bsky.social · 10/03/2025

The other day, somebody in the industry compared LLMs to operating systems, and I think this is a very good way to think about it. Operating systems are only viable at scale. And they're a means to an end, not the endpoint. They're only there to enable other tasks. Adaptation pays off at scale. 6/10

100

Tim König @koenigt.bsky.social · 10/03/2025

You could even argue that ChatGPT and other chatbots are merely there to showcase the model's capabilities. They're there to show a large user base what these models can do in terms of language understanding. It does not matter if the companies lose money on them - they're a proof of concept. 5/10

100

Tim König @koenigt.bsky.social · 10/03/2025

Most of these applications are not chat bots though. They're all types of tasks where you'd like your computer to be better at understanding natural language - and there's a lot of those. So part of the money is in third parties offering their clients "AI" applications for their businesses. 4/10

110

Tim König @koenigt.bsky.social · 10/03/2025

LLMs excell at language understanding, more so than any other technology available. And this capability can be used in a plethora of tasks - from information retrieval to human-computer-interaction to code completion. This is evidenced most strongly by third parties offering LLM integrations. 3/10

110

Tim König @koenigt.bsky.social · 10/03/2025

First of all, I agree with a lot of the issues Ed raises. "AI" is a marketing term. "AGI" even more so - and it can not happen within the framework of large language models. And yes, chatbots severly underdeliver in the majority of tasks they're advertised for. But that is besides the point. 2/10

100

Tim König @koenigt.bsky.social · 10/03/2025

A week ago or so I stumbled across this piece by @edzitron.com. He raises some compelling points and is asking the right question - where's the money? - but ultimately I think he is wrong. The money is there. And it's in the data centers. Let me explain. 1/10 www.wheresyoured.at/wheres-the-m...

wheresyoured.at

There Is No AI Revolution

Soundtrack: Mack Glocky - Chasing Cars Last week, I spent a great deal of time and words framing the generative AI industry as a cynical con where OpenAI's Sam Altman and Anthropic's Dario Amodei hav...

111

Tim König @koenigt.bsky.social · 06/03/2025

At least the author answers their own question of "fragmented or shallow reviewer feedback". It's reviewer 2 - now with 2 pages of AI slop to give it the semblance of meaningful feedback

010

Reposted by Tim König

Adam Bonica @adambonica.bsky.social · 04/03/2025

As another Trump-induced market crisis unfolds from tariffs, it's time we ask "What's the Matter with Billionaires?" I've run the numbers. Billionaires do better under Dems. So why do billionaires keep supporting GOP policies against their economic interests? See my latest piece on why. (TLDR 🧵⬇️)

open.substack.com

What's the Matter with Billionaires?

How America's wealthiest sabotage their own prosperity (and ours) to avoid paying taxes

982215788

Reposted by Tim König

Der Volksverpetzer @volksverpetzer.de · 04/03/2025

München, Solingen, Magdeburg: Zu allen Anschlägen gab es einen Brennpunkt in der ARD. Zur gestrigen Amokfahrt in #Mannheim bisher nicht. Warum wohl? www.volksverpetzer.de/aktuelles/ma...

volksverpetzer.de

Mannheim: Als die Herkunft bekannt wurde, machte man die Livestreams aus

München, Solingen, Magdeburg: Zu allen Anschlägen gab es einen Brennpunkt in der ARD. Zur gestrigen Amokfahrt in Mannheim bisher nicht.

611859580

Tim König @koenigt.bsky.social · 03/03/2025

Thanks to everyone who participated - as always, I learned a lot preparing the workshop and from all the engagement and smart questions!

000

Tim König @koenigt.bsky.social · 03/03/2025

Last week, I held a two-day workshop on NLP techniques, from Bag of Words to LLMs. Most materials are in English and include code and hands-on exercises. We covered static embeddings, classification with BERT/RoBERTa and LLMs, Model Evaluation, and finally built our own multi-agent and RAG systems!

github.com

GitHub - TimBMK/LLM-workshop-HI: Slides and Materials for a workshop on NLP, Transformers, LLMs and Agents. Taught at the University of Hildesheim, February 2025

Slides and Materials for a workshop on NLP, Transformers, LLMs and Agents. Taught at the University of Hildesheim, February 2025 - TimBMK/LLM-workshop-HI

120

Reposted by Tim König

CORRECTIV @correctiv.org · 11/02/2025

Nachdem Friedrich Merz mit der AfD gestimmt hat, drängen Wirtschaftslobbyisten und CDU-nahe Akteure zum nächsten Tabubruch: Eine Minderheitsregierung. CORRECTIV deckt auf, wer die CDU nach rechts treibt.

correctiv.org

Die Rechtstreiber der CDU

Einige Wirtschaftslobbyisten und Denkfabriken untergraben aktiv die Brandmauer. Der nächste mögliche Tabubruch: eine Minderheitsregierung

451314606

Reposted by Tim König

Alessandro Rigolon @alerigolon.bsky.social · 10/02/2025

I was just told that I have to remove “climate” from the title of an ongoing grant if I want to keep it. And publications from that grant cannot include “climate” and other forbidden words. I can’t believe I’m writing this from the United States of America. #AcademicSky

549129045481

Reposted by Tim König

Zentrum für Politische Schönheit @politicalbeauty.bsky.social · 10/02/2025

Wir haben die Schikanen der Polizei Berlin der letzten Wochen nicht an die große Glocke gehängt: Weil wir keine Absicht dahinter vermuteten. Seit gestern wissen wir: Die Polizei hat systematisch ALLES daran gesetzt in den letzten 30 Tagen, den Adenauer SRP+ zu verhindern! Ein Thread zum SKANDAL! 🧵

571776716

Reposted by Tim König

Arne Semsrott @arnesemsrott.bsky.social · 07/02/2025

die cdu kopiert jetzt wirklich einfach nur noch afd-forderungen

: AfD will Strafmündigkeit ab zwölf Jahren

CDU-Generalsekretär Linnemann will Strafmündigkeit ab zwölf Jahren
Ein 13-Jähriger soll einen 12-Jährigen in Stuttgart getötet haben. Baden-Württembergs CDU-Justizministerin Marion Gentges fordert eine wissenschaftliche Untersuchung. Ihr Generalsekretär denkt an härtere Konsequenzen.

27805195

Reposted by Tim König

Curd Knüpfer @curdknupfer.bsky.social · 06/02/2025

Take note: The infrastructures of democratic deliberation are under attack. This may start in the US but will have fallout across the globe. One way to strengthen democracy: support and protect public information access in whatever form you can.

08848

Reposted by Tim König

Ami Fields-Meyer @fieldsmeyer.com · 06/02/2025

1/ I suspect we are days away from hearing Elon is "using AI" to “streamline” benefits programs. You should know why fully automating decisions about who gets critical assistance is a terrible idea that will have disastrous consequences. A 🧵 based on work I’ve done with the great @awhf.bsky.social

nytimes.com

Musk Allies Discuss Deploying A.I. to Find Budget Savings

A top official at the General Services Administration said artificial intelligence could be used to identify waste and redundancies in federal contracts.

16235106

Reposted by Tim König

Dr Abeba Birhane @abeba.blacksky.app · 05/02/2025

this is such a good piece

721057

Reposted by Tim König

Margaret Mitchell @mmitchell.bsky.social · 04/02/2025

With Google removing their Responsible AI Principles, they no longer state that they will *not* engage in "Technologies whose purpose contravenes widely accepted principles of international law and human rights". Concerns about surveillance and injury are also erased. ai.google/responsibili...

ai.google

AI Principles – Google AI - Google AI

A guiding framework for our responsible development and use of AI, alongside transparency and accountability in our AI development process.

8269173

Reposted by Tim König

Gary Marcus @garymarcus.bsky.social · 04/02/2025

ChatGPT in Shambles. • can’t properly count to 50 • sometimes forgets US states • think the letter h can be a vowel • can’t count vowels to save its electronic life • issued numerous corrections that were wrong • lied about having a subconscious At open.substack.com/pub/garymarc...

open.substack.com

ChatGPT in Shambles

After two years of massive investment and endless hype, GPT’s reliability problems persist

1412935

Reposted by Tim König

Ronen Tamari @ronentk.me · 04/02/2025

Looking forward to reading this! 📚 www.taylorfrancis.com/books/oa-edi...

052

Reposted by Tim König

Social Media Lab @socialmedialab.ca · 03/02/2025

"Trump orders creation of US sovereign wealth fund, says it could buy TikTok" finance.yahoo.com/news/trump-s... There are currently 90 sovereign funds worldwide, and they are a hotbed for corruption and other governance risks. carnegieendowment.org/research/202...

finance.yahoo.com

Trump orders creation of US sovereign wealth fund, says it could buy TikTok

WASHINGTON (Reuters) -U.S. President Donald Trump signed an executive order on Monday ordering the creation of a sovereign wealth fund within the next year, saying it could potentially buy the short v...

64519

Reposted by Tim König

Paula Villa Braslavsky @paulita.bsky.social · 04/02/2025

This is censorship, full frontal attack on academic freedom by any means.

522470

Reposted by Tim König

FragDenStaat @fragdenstaat.de · 29/01/2025

Wir launchen den Real-O-Mat! Statt Wahlversprechen liefert der #RealOMat den Reality-Check zur #Bundestagswahl. Positionier‘ dich zu 20 politischen Forderungen und vergleiche wie die Parteien dazu tatsächlich abgestimmt haben – unabhängig von ihren Versprechen. Probier’s aus: real-o-mat.de

102970485

Reposted by Tim König

Anne Roth @anneroth.bsky.social · 29/01/2025

Wenn die AfD #Messer zieht, ist das Drama irgendwie nicht ganz so groß.

125583

Reposted by Tim König

Lisa Davidson @lisabdavidson.bsky.social · 27/01/2025

NIH study sections were paused last week, today NSF panels are pausing. And NYU sent out this email today. We all know how "temporary" these measures end up being (ahem, much like NYU put up "temporary" barriers to block access to publicly used plazas last year that are--surprise!--still there)

Notices of Cancellations of Federal Grants

Over the weekend, NYU’s Office of Sponsored Research (OSP) received notification from the US Department of State of two grants being terminated. The only reason given in each instance is that “the award does not meet the agency’s priorities.” We have reached out to the affected researchers, taken required steps, and offered support. It is certainly possible, if not likely, that more such notices will come.

The research activities of our scholars are central to NYU’s mission. The University values its faculty’s scholarship immeasurably and views these developments with the utmost seriousness, especially the sudden grant cancellations.

35440

Reposted by Tim König

Nolen Gertz @ethicistforhire.bsky.social · 24/01/2025

Soon we will learn whether the Humanities proving more difficult for AI to destroy (sorry, I mean "assist") than expected will mean that the Humanities get more respect ("Wow historians aren't just collecting facts!") or be seen as useless ("Why aren't historians just collecting facts?")...

13112

Reposted by Tim König

Zentrum für Politische Schönheit @politicalbeauty.bsky.social · 23/01/2025

Tesla Gigafactory, Berlin. – Buy a Tesla, support facism!

421807395

Reposted by Tim König

LobbyControl @lobbycontrol.bsky.social · 23/01/2025

Der jahrelange Druck von CDU-Politiker:innen auf die EU-Kommission hat gewirkt: Sie schwächt Nichtregierungsorganisationen über Änderungen im EU-Haushalt. Die Argumente der CDU sind fadenscheinig, scheinheilig und demokratiegefährdend. www.lobbycontrol.de/lobbyismus-i... 1/7

lobbycontrol.de

Konservative Kampagne: Wie Christdemokraten in Brüssel die Zivilgesellschaft schwächen

In Brüssel schwächt die EU-Kommission mit neuen Finanzierungsregeln den Einfluss von Umwelt-Organisationen. Dafür haben sich deutsche Christdemokraten seit Jahren eingesetzt. Sie setzen damit demokrat...

313747

Tim König @koenigt.bsky.social · 23/01/2025

It'd sure be a shame if somebody set up a bot to write daily emails about how the "Orange people in office program" has really gone too far...

040

Reposted by Tim König

Judd Legum @juddlegum.bsky.social · 21/01/2025

1. Donald Trump took the oath of office yesterday — the first person to be sworn in as president while hawking an eponymous meme coin It is brazen corruption $TRUMP is effectively a "For Sale" sign on the White House, facilitating anonymous bribes in any amount Follow this 🧵 for details

On Monday, Donald Trump took the oath of office — the first person to be sworn in as president while simultaneously hawking an eponymous meme coin.

451350319

Reposted by Tim König

Casey @mamasissiesays.bsky.social · 21/01/2025

Why will Meta let us search for Republican videos but not for Democrat videos?? What in the 1930’s Germany is this?!

1341520490

Reposted by Tim König

Mike Cowburn @mikecowburn.bsky.social · 21/01/2025

Elon Musk doing a Nazi salute is NOWHERE on the New York Times website this morning. But we do get a story about Melania Trump's hat, so...

New York Times front page 21st January 2025. No mention of Musk performing a Nazi salute at a Trump celebration rally. Melania Trump hat story bottom left.

49733