Listen To The Show
Transcript
Welcome to The Prompt by Kuro House, your daily AI update. Today, we’ve got some big moves in AI, from supercomputers to controversial new encyclopedias. Let’s dive right in.
First up, Elon Musk’s new project Grokipedia is stirring up quite a debate. According to The Verge, Grokipedia launched as an AI-powered alternative to Wikipedia but it’s been found to contain copied Wikipedia pages. The site looks basic, with a search bar and simple entries, but many pages are word-for-word replicas of Wikipedia content, complete with attribution notices. Users can’t edit pages like on Wikipedia, and the AI claims to fact-check, though some facts are outdated or controversial. Despite promises of a “massive improvement,” Grokipedia currently has under a million articles compared to Wikipedia’s 7 million. So, it seems even Grokipedia needs Wikipedia to exist.
Staying with Grokipedia, Wired reports it’s pushing far-right talking points and misinformation. The encyclopedia falsely links pornography to worsening the AIDS epidemic and suggests social media is driving a rise in transgender identities. Entries denounce mainstream media and emphasize conservative perspectives, often distorting historical facts. For example, the slavery entry criticizes The 1619 Project and frames slavery in ideological terms favoring certain narratives. Even high-profile figures like Elon Musk and Donald Trump have entries loaded with political bias. This raises serious concerns about the reliability and neutrality of AI-generated knowledge bases.
On a different note, AMD and the US Department of Energy are teaming up on a $1 billion AI supercomputer project. The Verge reports that Oak Ridge National Laboratory will host two new supercomputers, Lux and Discovery. Lux, launching in early 2026, is billed as the nation’s first dedicated AI Factory, optimized for training and deploying AI models to accelerate scientific discovery. Discovery, arriving in 2029, focuses on high bandwidth and energy efficiency to support breakthroughs in energy, biology, materials, and national security. This partnership builds on AMD’s previous work with the Frontier supercomputer, pushing US capabilities in AI research and innovation.
Zoom’s CEO Eric Yuan shared a bold vision for AI at TechCrunch Disrupt 2025. He told TechCrunch that AI assistants could soon shorten our workweek significantly. Zoom is integrating AI features like digital twins that can represent you in meetings or negotiate contracts autonomously. Yuan also envisions AI managing emails and helping users focus on what truly matters. His goal? To reduce the traditional five-day workweek to three or four days within five years, thanks to AI efficiency.
Finally, OpenAI revealed some sobering data about mental health conversations with ChatGPT. TechCrunch reports that over a million users weekly discuss suicidal thoughts with the AI, representing 0.15% of active users. OpenAI says its latest GPT-5 model responds more appropriately to mental health issues, improving compliance with safety guidelines by 65% over previous versions. The company consulted over 170 mental health experts to enhance these responses. Despite improvements, challenges remain, and OpenAI is also building better age detection and parental controls to protect vulnerable users.
So, AI is advancing rapidly, but with progress comes responsibility and controversy. From supercomputers driving innovation to AI encyclopedias raising questions about bias, it’s clear the AI landscape is complex. Thanks for tuning in to The Prompt by Kuro House. We’ll keep bringing you the stories that matter in AI every day.
