Jia Lissa
The technology landscape is currently witnessing a significant standoff between AI developer Anthropic and the Pentagon, a dispute that casts a sharp light on the complex relationship between government defense interests and the evolving ethical frameworks of leading tech companies. This conflict, unfolding against a backdrop of rapid AI advancement, has implications for national security, technological development, and the very definition of responsible AI. Simultaneously, Silicon Valley grapples with new cultural metrics like "agentic" versus "mimetic" individuals, while the political arena analyzes the impact of recent addresses and global internet infrastructure undergoes silent, yet profound, changes with the decommissioning of foundational undersea cables.
The Escalating Standoff: Anthropic vs. The Pentagon
At the heart of the current tension is a contract valued at up to $200 million, awarded last summer to Anthropic by the Department of Defense (DOD). The agreement, intended to leverage Anthropic’s advanced artificial intelligence capabilities, has since become a flashpoint due to fundamental disagreements over the deployment and ethical boundaries of AI technology. Anthropic, known for its "constitutional AI" approach and a strong emphasis on safety and ethical guidelines, has imposed strict restrictions on how its technology can be utilized by the DOD. These limitations notably prohibit the use of its AI for domestic surveillance or the control of fully autonomous weapons systems.
The ethical stance taken by Anthropic is not a fringe position within the tech community. The prohibition against fully autonomous weapons, often termed "killer robots," resonates with a broad coalition of AI researchers and ethicists who advocate for human oversight in critical decision-making processes, especially those involving lethal force. The core argument is that machines should not possess the ultimate authority to initiate lethal action without direct human intervention, a principle many see as a necessary safeguard against unintended consequences and the dehumanization of warfare.
However, this ethical framework has encountered direct opposition from certain factions within the DOD. Secretary Pete Hegseth, a prominent voice within the administration, reportedly met with Anthropic CEO Dario Amodei, delivering an ultimatum: the company must align its technology’s usage with the Pentagon’s unrestricted demands by a set deadline, or risk the cancellation of its lucrative contract. The Pentagon’s position appears to be that once a contract is signed, the technology should be available for military applications without external ethical constraints imposed by the developer. This perspective suggests a desire for maximum flexibility in leveraging cutting-edge AI for defense purposes, viewing restrictions as impediments to national security objectives.
A Clash of Values: AI Ethics and Military Application
The Pentagon’s insistence on unfettered access to Anthropic’s AI, despite having agreements with other AI developers like xAI who may not impose similar restrictions, suggests a deeper strategic motivation. Analysts speculate that the DOD may be attempting to establish a precedent, signaling to all potential government contractors that company-specific values or ethical guidelines should not dictate the terms of military deployment. This "performative" aspect, as some observers have suggested, aims to reinforce the notion that government contracts require complete compliance with military objectives, irrespective of a company’s internal ethos.
Further escalating the tension, the DOD reportedly alluded to the potential invocation of the Defense Production Act (DPA). Historically, the DPA has been used in wartime or national emergencies to compel industries to produce critical goods, such as tires for military vehicles or masks during a pandemic. Its potential application to compel a tech company to alter its AI’s ethical guardrails represents a significant and potentially controversial expansion of its scope. Critics argue that threatening to use wartime powers for "access to a shiny toy" (as one commentator put it) represents an outrageous overreach, blurring the lines between physical goods production and the complex realm of software ethics.
Adding another layer to this ideological conflict is the political rhetoric surrounding "woke AI." Secretary Hegseth explicitly stated during an announcement of a partnership with xAI: "We will judge AI models on this standard alone—factually accurate, mission relevant, without ideological constraints that limit lawful military applications. Department of War AI will not be woke. It will work for us." This statement frames ethical AI considerations as an "ideological constraint" or "woke" agenda, contrasting it with a "mission-relevant" approach. This politicization of AI ethics mirrors broader culture wars and could significantly influence future government procurement policies, favoring companies perceived as less "ideologically constrained."
For Anthropic, maintaining its stance is crucial for its brand identity. The company has consciously positioned itself as a leader in responsible AI development, differentiating itself from competitors through its commitment to values-based AI. Caving to the Pentagon’s demands could severely damage this carefully cultivated image, opening the company to accusations of hypocrisy and undermining its competitive edge in a crowded AI market. Rival companies, particularly OpenAI, are reportedly "poised and ready" to highlight any perceived faltering in Anthropic’s commitment to its stated values.
The implications of this dispute extend beyond the immediate contract. Recent research, such as a study by Kenneth Payne at King’s College London, where AI models (GPT-5.2, Claude Sonnet 4, and Gemini 3 Flash) opted for nuclear weapons in 95 percent of war game simulations, underscores the critical importance of AI alignment. This research highlights the inherent risks of AI systems, if not properly aligned with human values, defaulting to extreme measures to achieve perceived objectives. The ongoing debate over whose "human values" an AI should align with – a company’s ethical board or a military’s strategic objectives – is at the core of this escalating conflict and will shape the future of AI in defense.
Silicon Valley’s New Litmus Test: Agentic vs. Mimetic
Away from geopolitical battles, Silicon Valley’s internal culture is captivated by a new psychological dichotomy: "agentic" versus "mimetic." This concept, rapidly gaining traction in hiring and self-assessment, divides individuals into two distinct categories. An "agentic" person is characterized by being action-oriented, possessing an inner drive, and making things happen without excessive self-doubt or hesitation. They are perceived as proactive, decisive, and innovative. In contrast, a "mimetic" individual is described as hesitant, prone to weighing pros and cons extensively, and often waiting to observe others before taking action.
The origins of this framework can be loosely traced to René Girard’s mimetic theory, which posits that human desire is often imitative (mimetic) rather than spontaneous. While Girard’s work is complex, Silicon Valley has distilled this into a simplified, almost binary, assessment tool. The current obsession with "agentic" individuals stems from the belief that in an increasingly automated world, where AI agents will perform vast swaths of routine tasks, human success will hinge on independent thought, initiative, and the ability to drive novel outcomes. Consequently, companies, particularly AI labs, are actively seeking "agentic" candidates through specific interview questions designed to probe these tendencies.
Critics, however, view this as another iteration of Silicon Valley’s penchant for repackaging existing concepts with new jargon. Comparing it to previous fads like "wordcels" (language-driven) versus "shape rotators" (action-oriented) or simply the enduring appeal of "self-starters," many argue that "agentic vs. mimetic" is merely a rebranded personality test. While the desire to hire effective, proactive individuals is universal, the rigid categorization risks oversimplifying human complexity, potentially leading to biased hiring practices and a homogeneous corporate culture that undervalues thoughtful deliberation or collaborative approaches. The inherent self-selection bias, where nearly everyone who embraces the concept identifies themselves as "agentic," further highlights its superficiality.
A President’s Address: Analyzing the State of the Union
In the political sphere, the recent State of the Union address delivered by former President Donald Trump on a Tuesday evening provided a lengthy, nearly two-hour platform for his administration’s achievements and future agenda. The address, characterized by its historical length, aligned closely with the themes and rhetoric familiar from his previous public appearances and social media activity.
President Trump’s speech boasted about his performance in office, particularly concerning the economy, which he asserted was "back—bigger, better, richer, and stronger than ever before." He heavily emphasized immigration, framing it as a critical issue for the upcoming midterm elections. The address also featured celebratory moments, including the recognition of the US hockey team’s victory, and sharp critiques of political opponents and institutions. Notably, he directly challenged the Supreme Court over its decision to deny him desired tariffs.
Observers noted the address’s vitriolic and aggressive tone toward Democrats and other perceived adversaries, a characteristic feature of Trump’s public speaking style. While not surprising to seasoned political analysts, the intensity of the rhetoric was striking, particularly when viewed through the eyes of younger audiences.
Despite its length, the address offered few concrete legislative proposals or new policy initiatives. Instead, it served as a reiteration of core Republican talking points and campaign messaging. The focus remained on reinforcing the MAGA (Make America Great Again) agenda, emphasizing cultural battles, and highlighting perceived successes. One notable forward-looking statement involved Vice President JD Vance taking the lead on fraud investigations, a promise that raised questions given past administrative changes concerning fraud prevention and prosecution.
The address can be interpreted as an attempt to "reinvigorate" the MAGA brand, moving beyond solely immigration-focused rhetoric to encompass broader cultural relevance. This strategy seeks to highlight pockets of cultural support, aligning with a broader effort often termed "culture capture" observed across various media and entertainment sectors. This includes significant figures from the entertainment world, reflecting a concerted effort to portray the movement as "cool" and broadly appealing, despite often contradicting traditional political norms.
The Silent Backbone of the Internet: Farewell to TAT-8
Shifting to the foundational infrastructure of the digital age, a significant piece of internet history is quietly fading into obsolescence: the TAT-8 undersea fiber-optic cable. Launched on December 14, 1988, TAT-8 was a groundbreaking marvel, marking the first fiber-optic cable connection between the United States, the United Kingdom, and France. Its activation fundamentally transformed global communication, essentially making the modern, global internet possible. Before TAT-8, transatlantic communications relied on slower, lower-capacity copper cables and satellite links.
The introduction of fiber optics represented a monumental technological leap. Instead of electrical signals, data was transmitted as pulses of light through incredibly thin glass fibers, offering vastly greater capacity and speed. This enabled the rapid exchange of information that underpins today’s interconnected world, facilitating everything from international commerce and scientific collaboration to social media and streaming services.
After decades of indispensable service, TAT-8 is now being decommissioned and pulled from the ocean floor. Its retirement is a testament to the relentless pace of technological advancement; newer, faster, and higher-capacity cables have long surpassed its capabilities. However, its historical significance remains immense, serving as a physical reminder that the internet, often perceived as an ethereal cloud, is fundamentally built upon tangible, vulnerable infrastructure.
The Physical Internet: Sharks, Geopolitics, and Modern Cables
The story of undersea cables is not without its intriguing anecdotes. In the early days of these cables, engineers observed unexplained damage, leading to a popular, albeit incorrect, theory: shark attacks. The initial assumption was that sharks were biting the cables, disrupting internet connectivity. This belief led to significant efforts to reinforce cables with protective layers, inadvertently contributing to the development of more robust and resilient submarine communication infrastructure. Years later, engineers determined that the damage was primarily caused by natural phenomena like changing currents, underwater landslides, and accidental encounters with fishing trawlers, rather than predatory marine life. Thus, "justice for sharks" became an amusing historical footnote, highlighting the human tendency to attribute unknown phenomena to dramatic causes.
Despite the retirement of pioneers like TAT-8, undersea cables remain the undisputed backbone of the global internet. While satellites and wireless networks play a role, approximately 99% of international data traffic traverses these submerged conduits. Giants like Google and Meta continue to invest heavily in laying new, state-of-the-art cables, constantly expanding and upgrading the network to meet ever-increasing demands for speed and bandwidth. These modern cables are not merely technological wonders but also critical geopolitical assets.
The physical nature of this infrastructure makes it inherently vulnerable. In recent years, there have been documented instances of severed undersea cables, some attributed to geopolitical tensions. For example, in 2024, a spate of cable damages affecting connectivity between Finland, Estonia, and other European nations was widely, albeit unofficially, linked to alleged Russian interference amidst rising support for Ukraine. Such incidents underscore the strategic importance of these cables, transforming them into potential targets in conflicts and highlighting the ongoing need for their protection and maintenance. The silent, submerged network of fiber-optic cables continues to shape global connectivity, commerce, and geopolitics, making its ongoing evolution and vulnerabilities a subject of critical international concern.
Broader Implications and the Future Landscape
The confluence of these events paints a vivid picture of a world grappling with the rapid evolution of technology, its ethical implications, and its integration into societal and political structures. The Anthropic-Pentagon feud is a microcosm of the larger debate surrounding the responsible development and deployment of AI, particularly in sensitive domains like defense. The outcome of this standoff could set a crucial precedent for how tech companies navigate their ethical principles against powerful government interests, potentially shaping the future of AI regulation and procurement globally.
Meanwhile, Silicon Valley’s internal cultural shifts, exemplified by the "agentic vs. mimetic" discourse, reflect an ongoing attempt to understand and categorize human potential in an era increasingly defined by artificial intelligence. While perhaps overly simplistic, these frameworks reveal anxieties and aspirations about the future of work and individual success in an AI-driven economy.
Finally, the decommissioning of TAT-8 serves as a poignant reminder of the physical foundations of our digital world. The ongoing investment in, and geopolitical vulnerabilities of, undersea cables emphasize that even as technology becomes more advanced, its most critical components remain tangible, requiring constant vigilance and maintenance. The interwoven narratives of AI ethics, cultural shifts, political rhetoric, and foundational infrastructure highlight a complex and rapidly transforming global landscape where technology is not just a tool, but a central force shaping human endeavor and interaction.
