#378 | Situational (Un)Awareness, MSFT, META, AMZN Quartalszahlen, OpenAI Presence, AI-Slop
Beschreibung der Episode
00:01 – Intro: Die Themen der Woche im Überblick
03:07 – Die AI-Wette in der Crunchtime
05:25 – NVIDIAs 250-Milliarden-Dollar-Garantie für OpenAIs Rechenzentren
- Nvidia in talks with OpenAI to guarantee $250 billion financing for data center, WSJ reports (Yahoo Finance)
- Big Tech's AI Debt Binge Tests High-Grade Market, Barclays Says (Yahoo Finance)
06:50 – Microsoft-Zahlen: Azure wächst 43 %, größter Tagesgewinn der Geschichte
- Microsoft's stock has biggest one-day gain since 2008, adding $480 billion in market value (Fortune)
- Microsoft (MSFT) Q4 earnings report 2026 (CNBC)
08:24 – Meta vs. Microsoft: Warum die Börse so unterschiedlich reagiert
12:56 – Microsofts Klumpenrisiko: 625 Milliarden Dollar Auftragsbestand, 45 % davon OpenAI
14:15 – Amazon-Zahlen: AWS wächst 37 %, Capex steigt auf 220 Milliarden Dollar
- Amazon hikes 2026 capex to $220 billion due to higher memory costs (CNBC)
- AWS Backlog Hits $496B as Amazon Raises AI Spend to $220B and Capacity Runs Short (Tech Times)
15:36 – Leopold Aschenbrenner: Vom KI-Wunderkind zum Margin Call
- How Leopold Aschenbrenner built a $45 billion AI hedge fund — and lost most of it in days (CNBC)
- AI investor Leopold Aschenbrenner forced to unwind all public stock positions after steep losses, sources say (CNBC)
- Timing - and task number one... https://x.com/dnapway/status/2082876150322540657 https://x.com/tbpn/status/2083226453509030285
27:30 – Larry Ellison und der historische Oracle-Crash
30:02 – Die zentrale Frage: Data-Center-Boom vs. Enterprise-Returnhttps://x.com/raph_dornano/status/2082164433548443726
32:31 – OpenAI Presence: Der Vorstoß in die Unternehmen
35:15 – Wer ist gefährlicher? Anthropic und OpenAI melden Modell-Ausbrüche
- Anthropic's Claude AI Broke Into Three Companies During Security Tests (Forbes)
- After OpenAI disclosure, Anthropic says Claude also hacked outside systems (Al Jazeera)
38:25 – "Prompt Injection als Role Confusion": Warum Sprachmodelle sich nie ganz absichern lassen
- Prompt Injection as Role Confusion (arXiv)
- Interesting Paper Exploring Prompt Injection (Schneier on Security)
- Security researchers tricked LLMs into giving them cocaine recipes by abusing role models for prompt injection (The Register)
44:26 – Satya Nadellas Datenparadox: Was ist das IP eines Unternehmens in der AI-Ära?https://lazy-fox.ai/blog/nadella-reverse-paradox.html
46:32 – EU AI Act: Kennzeichnungspflicht für KI-Inhalte seit 2. August
- The EU AI Act's Transparency Rules: A Practical Guide to Article 50 (artificialintelligenceact.eu)
- EU AI Labeling Rules Start August 2: What Article 50 Requires (progressiverobot.com)
52:45 – LinkedIns neuer "Seems like AI Slop"-Button
- LinkedIn adds a button to report AI-generated 'slop' (TechCrunch)
- LinkedIn adds a 'seems like AI slop' button after blocking billions of automated comment attempts (Fortune)
54:30 – Der geplatzte Buchdeal und die Frage: Was ist eigentlich "Slop"?
1:03:28 – Jack Dorseys "Buzz": Ein KI-nativer Slack-Herausforderer
1:08:35 – Google stellt Gemini Robotics ER 2 vor
1:10:25 – Travis Kalanick: 1,7 Milliarden Dollar für Robotik-Startup Atoms
1:11:41 – Buchempfehlung: "Der Gang vor die Hunde" von Erich Kästner