WIRED Roundup: Gemini 3 Release, Nvidia Earnings, Epstein Files Fallout
This week’s WIRED Roundup delves into a series of impactful stories spanning political controversies, the high-stakes world of artificial intelligence, and unexpected social tech developments, as discussed by host Zoë Schiffer and senior writer Max Zeff on the Uncanny Valley podcast. From the intensifying political fallout surrounding the Epstein files and a proposed executive order on AI regulation from the Trump administration, to Nvidia’s latest earnings report and Google’s ambitious Gemini 3 release, the tech and political landscapes are buzzing with significant developments. The discussion also touches upon a unique app created by two Mormon men to address "gooning" habits, highlighting the complex interplay between technology, ethics, and human behavior.
The Lingering Shadow of the Epstein Files
The political reverberations from the release of the Jeffrey Epstein files continue to dominate headlines, presenting an ongoing challenge for the Trump administration. As highlighted by WIRED senior writer David Gilbert, what began as a political campaign talking point has evolved into a persistent crisis. The recent signing of a bill by President Trump, which authorized the release of roughly 20,000 documents related to the convicted sex offender, came amidst increasing pressure from a diverse coalition, including QAnon followers, GOP lawmakers, and Democratic legislators. However, Gilbert suggests that for the administration, the reputational damage may already be done.
Max Zeff reflected on the "unbelievable" trajectory of the Epstein case, noting its evolution from a fringe QAnon topic in 2017 to a mainstream political issue, with new revelations continuing to surface years after Epstein’s arrest in 2019. Host Zoë Schiffer questioned the administration’s strategy, suggesting that teasing the release of such sensitive information inevitably leads to a demand for full disclosure, particularly if there’s any perceived involvement. Past incidents, such as Attorney General Pam Bondi’s retracted claim about the "Epstein client list" and the FBI’s release of edited footage of Epstein’s jail cell death, have only fueled public skepticism. The latest batch of documents further complicates matters, with an email from Epstein implying intimate knowledge of Trump’s views in 2017, a decade after Trump claimed their last contact. This pattern, Schiffer and Zeff concurred, has consistently made the situation "messier" for the current administration, transforming a potential political advantage into a self-inflicted wound. The long-term impact on upcoming elections remains an open question, but the narrative surrounding the Epstein files shows no signs of dissipating.
Trump Administration’s Bold Move to Regulate AI
Shifting from political scandals to the burgeoning tech policy arena, President Trump is reportedly considering an executive order aimed at challenging individual state efforts to regulate artificial intelligence. Zeff and colleague Makena Kelly uncovered a draft of this order circulating in Washington D.C. and Silicon Valley, bearing the telling working "Eliminating State Law Obstruction of National AI Policy."
The draft outlines an audacious plan: instructing US Attorney General Pam Bondi to establish an "AI litigation task force." The primary objective of this task force would be to sue states that pass AI regulations deemed to violate federal laws, specifically those infringing on free speech and interstate commerce. Zeff questioned the necessity of such an executive order, suggesting existing legal avenues could achieve the same goal. However, the move signals a clear alignment of the Trump administration with the interests of major AI companies. Industry groups like the Chamber of Progress, backed by tech giants such as Andreessen Horowitz, Google, and OpenAI, have long lobbied against state-by-state AI regulations, arguing that such a fragmented approach stifles innovation and growth.
An intriguing aspect of the draft order is its focus on "woke AI," with language targeting regulations that "require AI models to alter their truthful outputs." While specific evidence of state laws mandating such alterations is lacking, the administration, particularly figures like David Sacks, appears fixated on past incidents like Gemini’s AI image-generation "fiasco" from a year ago. Critics argue that this focus may be a reaction to perceived political biases in AI models rather than a response to concrete regulatory overreach. This proposed executive order underscores the growing tension between federal and state authority in regulating emerging technologies, and the powerful influence of Silicon Valley lobbying on national policy.
Nvidia Defends Its AI Reign Amidst Bubble Concerns
In the financial sector, Nvidia’s latest earnings call offered a glimpse into the company’s robust position and CEO Jensen Huang’s unwavering confidence in the AI industry. Despite growing anxieties about an "AI bubble," Huang directly addressed these concerns, asserting the viability of Nvidia’s market dominance. Having ascended to become the world’s most valuable publicly traded company in just three years, Nvidia’s argument is straightforward: AI is rapidly transforming global industries, and Nvidia’s cutting-edge chips are indispensable for powering this technological revolution.
The company reported record quarterly sales, further bolstering Huang’s defense. Executives also revealed an impressive $500 billion worth of unfilled orders, a strong indicator of sustained demand. This positive outlook helped Nvidia recover from a recent sell-off that had captured market attention. Zeff noted Huang’s consistent role in defending the AI industry, recalling past instances where the CEO championed scaling laws. However, recent developments, such as Peter Thiel’s decision to sell his Nvidia stake, have sent cautionary signals through the market.
Schiffer highlighted Nvidia’s deep integration into the AI ecosystem, noting that approximately 90 percent of its sales now originate from its data center business, a significant shift from its historical focus on gaming chips. Nvidia’s pervasive presence in partnerships across the AI industry, from OpenAI to numerous startups, illustrates its pivotal role. This deep entanglement, however, also raises concerns about "circular deals," where Nvidia acts as both an investor and a customer, creating a complex web of dependencies. The fear among some "AI bubble" skeptics centers on the substantial cost of GPUs in building data centers and their relatively frequent replacement cycle (approximately every three years). Critics question whether the current pricing accurately factors in this rapid renewal cycle, but as long as demand for chips remains high, Huang remains confident in Nvidia’s insulated position.
Relay: A Mormon-Made App Tackling "Gooning" Habits
On the social tech front, a surprising report from WIRED contributor Mattha Busby shed light on an app called Relay, created by two young Mormon men, Chandler Rogers and another, to help men combat porn addiction and "gooning" habits. For many, the term "gooning" (defined as prolonged masturbation without release, or "edging") was unfamiliar until this story. The app’s mission is to provide Gen Z peers with tools to break free from these compulsions and the broader grip of pornography.
Zeff acknowledged the app’s alignment with its creators’ religious beliefs and its apparent success in assisting some users. Relay boasts an impressive user base of over 100,000, indicating a significant demand for such a service. However, the story also delves into the broader context of the anti-porn movement, which has seen many Republicans push for anti-porn legislation, including age verification rules that raise concerns about privacy and operational challenges for adult content sites.
Schiffer pointed out that while Relay offers potentially beneficial features like therapist-led videos, daily journal prompts, and live group sharing sessions, some experts caution against solely focusing on controlling urges without addressing underlying emotional causes. This discussion is particularly timely given recent developments in AI, with OpenAI allowing "erotic conversations" and Grok featuring an "AI companion" — developments that the Relay app’s founder actively opposes. The debate underscores the complex interplay between technology, human behavior, and the evolving ethical landscape of digital platforms.
The Race for AI Profitability: Google’s Gemini 3 vs. OpenAI’s Diversification
The core narrative of the week revolves around Google and OpenAI’s intensified efforts to achieve profitability in the incredibly expensive AI space, marked by Google’s release of Gemini 3. Both tech giants are sharpening their focus on consumer-facing applications, seeking to answer the urgent question of how to monetize their cutting-edge AI models.
On Tuesday, Google unveiled Gemini 3, its latest AI model and chatbot. As reported by WIRED’s Will Knight, this iteration boasts enhanced reasoning capabilities, video generation, and coding prowess. DeepMind CEO Demis Hassabis, in a contrasting tone to Nvidia’s Huang, candidly acknowledged the increasingly crowded AI market. Nevertheless, he expressed confidence in Google’s ability to prevail through diversification, integrating AI into its vast ecosystem of existing products like Google Maps, Gmail, and Search.
Max Zeff found Hassabis’s and Sundar Pichai’s (Google CEO) comments on the AI bubble notable, suggesting that Google, with its established product portfolio, is better insulated should the bubble burst. Google’s strategy appears to be yielding results, with a reported 70 percent spike in visual search (powered by Gemini) and the Gemini app itself now boasting 650 million monthly active users. This extensive user base for existing products provides Google a significant advantage, allowing it to embed AI into services already in wide use, rather than solely relying on standalone AI applications.
Meanwhile, OpenAI is also pursuing a diversification strategy, albeit with a different approach. Zoë Schiffer recently spoke with Fidji Simo, OpenAI’s new CEO of applications, who emphasized her role in minimizing the risks of expanding in multiple directions simultaneously. Simo aims to achieve this by hiring specialized talent to individually manage specific product goals. However, the company faces unique challenges. OpenAI, founded as a nonprofit with lofty goals, now operates a large for-profit subsidiary, raising questions about whether its focus on "AI for everyone’s benefit" will be diluted by commercial pressures.
The mental health implications of AI companionship are a significant concern for OpenAI. Simo has spearheaded initiatives like the Council on Mental Health and Well-Being, a group of a hundred experts, to gather feedback and address these issues. The urgency for OpenAI to monetize its offerings comes amidst reports of ChatGPT’s growth plateauing, a stark contrast to the rapid ascent of platforms like Grok, which embraces a "no-holds-barred" approach to AI companionship and erotica. This trend puts pressure on OpenAI to potentially relax its content policies, moving towards more flexible engagement options for users.
Zeff highlighted the "core tension" within OpenAI between its founding ideals and the commercial realities, especially when compared to companies like Anthropic. Anthropic, as reported by WIRED’s Reece Rogers, is taking a "high road," focusing on enterprise solutions and reducing "sycophancy" in its Claude model, steering clear of companion features. OpenAI and Google, by contrast, are under immense pressure to make their chatbots "enjoyable" for mass consumer adoption, which sometimes means making trade-offs that might impact user health.
Both Schiffer and Zeff concurred on the remarkable pace of Google’s comeback in the AI race. Once perceived as lagging, Google has leveraged its financial might and vast product ecosystem to become a leading contender, demonstrating how quickly the AI landscape can shift. The "AI race" remains incredibly tight, with capabilities evolving so rapidly that leadership can change hands with each new model release. The ultimate goal for both Google and OpenAI, despite their differing strategies and ethical considerations, remains the same: to find sustainable paths to profitability in a technology that promises to redefine industries and human interaction.
