New obligations for publishing AI-generated or AI-edited media from August 2nd onwards in the EU. When sharing artificially generated people, places, events or text in matters of public interest, transparency must be ensured (e.g. via disclaimers, image labels). hoerndlein-consulting.de/en/blog/0027

The EU AI Act and what actions you need to take

The regulatory framework for AI inside the EU is coming into effect until 2030. What does this mean for your AI applications, which obligations are imposed on you and until when?

hoerndlein-consulting.de

Which traits differentiate good from bad managers - and how does that evolve with AI getting omnipresent? www.fastcompany.com/91559339/are...

Are managers becoming obsolete—or more important than ever?

Eight experts weigh in on the debate.

fastcompany.com

AI labs warn about the threats of their own models - yet their safety efforts may overlook the most robust method. New research shows that filtering training data is over 10x more resistant to misuse than post-training fixes. #AISafety #DataQuality hoerndlein-consulting.de/en/blog/0031...

Watch how your AI grows - what do you feed it?

While leading AI labs are warning about the potential threats of their AI models, their efforts to make them safer might focus on the wrong methods. Recent research shows what they are missing out - a...

hoerndlein-consulting.de

AI's leaders spent years warning that AI would erase white-collar jobs. Now they suddenly highlight that the "human part" of work can't be outsourced to machines. A change of mind? Rather a change of where the money flows. open.substack.com/pub/intellig...

Follow the money, not the mood

Why AI's loudest doomsayers suddenly stopped warning us about our jobs - it's not because they've changed their deeply held beliefs.

open.substack.com

The undisputed leader in web search is rewriting how we discover content online — what does this mean for digital businesses and the open web? Meanwhile, Europe's leading AI companies are making a very different bet: industrial customers, not consumers. open.substack.com/pub/intellig...

We aren't being outsmarted by super intelligent AI, but outpaced. Significant changes in how we treat application security lie ahead of us. To prepare, we need to understand the bigger picture of changes, instead of chasing - and fearing - latest AI models only. open.substack.com/pub/intellig...

Decoding the AI Security Hype: 5 Realities Beyond the Headlines

AI is setting out to crack your software. What that really means for the future of secure software.

open.substack.com

The end of SORA only means that producing AI slop is not a sustainable business model. Various #GenAI video tools are still available and powerful, even as #OpenSource, supporting content creators. Providing fancy tech alone is not sustainable business. www.heise.de/en/backgroun...

OpenAI discontinues Sora: The end of ever more and bigger

The end of Sora marks OpenAI's strategic shift. Developers and companies are now to ensure revenue, analyzes Eva-Maria Weiß.

heise.de

Der EU AI Act wurde durch jahrelangen Lobbyismus gekapert, um die Geschäftsmodelle großer Tech-Konzerne zu schützen. Start-ups und Open-Source haben das nachsehen. Regulierung an sich ist nicht das Problem - sondern falsche Regulierung. #EUAIAct

Warum wir den Erfolg des EU AI Acts an den falschen Maßstäben messen

Was, wenn Europa seine KI-Industrie nicht überreguliert, sondern einfach nur falsch? Was mit dem Ziel begann, Menschen vor Risiken zu schützen und Vertrauen in der Nutzung von KI zu schaffen, wurde du...

hoerndlein-consulting.de

In the 90s, they said you had to learn HTML to survive the internet. Today, they say you must become an AI expert to keep your job. Both panics are wrong. Because good UX always wins, domain expertise and AI literacy will beat technical mastery for most of us. #FutureOfWork

We've been here before: What the 1990s HTML panic teaches us about the future of work

Many people face a wall of anxiety on AI: become an expert or irrelevant. But history tells a different story. Looking at the early web shows us how the future of work with AI will unfold.

hoerndlein-consulting.de

Many people claim that Anthropic is more responsible than OpenAI because of their conflict with US Department of Defense. A closer read of the fine print shows a different narrative. Anthropic does not have a better moral stance than other TechCos. But at least they ponder what tech is capable of.