Mercury Automates Banking. Amazon Automates Warehouse Jobs. Ex-Cohere Lead Bets Against Scaling. AI Answers Contain Errors.

Show notes

The AI news for October 23rd, 2025

--- This episode is sponsored by ---

Find our more about our today's sponsor Airia at airia.com.

---

Would you like to create your own AI-generated and 100% automated podcast on your chosen topic? --> Reach out to us, and we’ll make it happen.

Here are the details of the day's selected top stories:

Project Mercury replaces hundreds of bankers.
Source: https://www.all-ai.de/news/news24/openai-mercury
Why did we choose this article?
High-impact, concrete example of AI moving from general assistants to specialized, high-value automation. Important for readers who need to understand enterprise use cases, near-term job and cost effects, and why financial services is a strategic target for AI vendors.

Amazon's robot army is supposed to replace 600,000 jobs.
Source: https://www.all-ai.de/news/news24/amazon-ki-strategie
Why did we choose this article?
Major, well-sourced claim about large-scale automation with clear implications for labor markets, supply chains, and corporate strategy. Useful for listeners tracking policy, workforce planning, and the social consequences of AI-driven automation.

Why Cohere’s ex-AI research lead is betting against the scaling race
Source: https://techcrunch.com/2025/10/22/why-coheres-ex-ai-research-lead-is-betting-against-the-scaling-race/
Why did we choose this article?
Provides a strategic counterpoint to the dominant 'scale-up' narrative. For builders and decision-makers it highlights alternative research directions—adaptive, efficient models—that could be more robust, cheaper to run, and better aligned to real-world deployment constraints.

Misinformation by AI: 45 percent of the answers are incorrect.
Source: https://www.heise.de/news/Europaeische-Rundfunkunion-KI-Systeme-geben-Nachrichteninhalte-oft-falsch-wider-10796779.html?wt_mc=rss.red.ho.themen.k%C3%BCnstliche+intelligenz.beitrag.beitrag
Why did we choose this article?
Concrete empirical finding on model reliability in the news domain. Important practical takeaway: current models make frequent interpretation errors, so newsrooms, product teams, and policymakers must treat LLM outputs as untrusted and layer verification and human oversight into workflows.

Do you have any questions, comments, or suggestions for improvement? We welcome your feedback at podcast@pickert.de.