#47 Society and Technology

It is now quite evident that AI-supported technologies are systematically replacing certain types of tasks. In this new era, many executives say that the most valuable candidates are those with high-quality skills that AI has not yet been able to replicate. Indeed, these candidates not only stand out in areas where AI falls short but also excel significantly in AI-assisted tasks, producing high-level work that others cannot match.

Amodei points out that with AI-assisted tools, people can now produce high-quality work without being hindered by technical barriers, thereby expanding opportunities for everyone. However, he also notes that in light of this shift, possessing technical knowledge no longer plays as distinctive a role as it once did. Amodei is a graduate of literature and possesses deep familiarity with the humanities, making him well aware of the value of the education being offered or pursued.

In an interview with ABC News, he said: “Today at Anthropic, when hiring, we are looking for individuals with exceptional communication skills, high emotional intelligence, excellent interpersonal abilities, kindness, compassion, curiosity, and a genuine desire to help others.” Similar statements were made by the CEO of JP Morgan: “Although AI will eliminate some jobs, young people should focus on critical thinking, communication skills, strong writing, and performing well in meetings. If they master these, they will find plenty of work.”

The growing centrality of so-called soft skills is not limited to Anthropic or JP Morgan. Today, many companies emphasize the importance of abilities that cannot be easily quantified. Having a mindset that values what it means to be human, being able to highlight human distinction, and sustaining classical methods have become key attributes of the new era.

The cognitive abilities we are handing over to chatbots must be preserved and developed before they disappear from our skill inventory. Therefore, even as AI usage continues to rise, we must safeguard our ability to understand texts without chatbots, write original content, and generate ideas.

The issue of skill loss, which we have repeatedly highlighted in our societal and technological bulletins, has now entered the agenda of global corporations. While making definitive predictions about the future is difficult, it is not hard to foresee that critical thinking, high-quality text production, and intellectual depth will become more valuable than ever.

Elon Musk announced that he has brought together SpaceX, which operates in space transportation and satellite internet, with his AI venture xAI under a single umbrella. This merger, valued at approximately $1.25 trillion, has been recorded as one of the largest corporate acquisitions to date and has sparked new debates about the future of space, AI, and data infrastructure. While SpaceX is a profitable company known for its rockets and the Starlink satellite network, xAI was an immature AI venture with an advantage in data through its chatbot Grok and the X platform, but one that incurred massive expenses without yet establishing a sustainable business model. The reasons and mechanisms behind this merger must be understood not only as a technological development but also as a financial and strategic move.

The primary justification Musk has emphasized is that the rapidly growing energy and computational demands of AI cannot be sustainably met by terrestrial data centers. According to Musk, current AI progress remains dependent on massive data centers that consume enormous amounts of electricity and require intensive cooling. His proposed solution is a global connectivity network based on Starlink, satellites powered by solar energy, and, in the long term, distributed data centers in space. In this framework, SpaceX is no longer merely a company that launches rockets; it is positioned as a platform providing the physical infrastructure for the age of artificial intelligence.

According to Musk’s vision, future computing systems in orbit will enable AI models to process data in space, creating a more cost-effective infrastructure over time. This approach appears to move space beyond its traditional roles as a medium for communication or transportation. Moreover, it signals a paradigm shift that transforms space into a direct site for computation and data generation.

At this point, the merger offers an early example of how space and data will become increasingly intertwined. In fields such as communications, defense, surveillance, finance, and AI, the location where data is processed has become as strategic as the data itself. The SpaceX–xAI example is shaping the perception that satellite networks, AI, and data centers will collaborate more closely across diverse sectors in the coming years.

However, explaining this move solely as a technological vision is insufficient. xAI’s current state reveals a strong financial dimension to the merger. xAI is a structure that spends billions of dollars to compete with rivals, burns through cash rapidly, and has not yet developed a sustainable revenue model. Therefore, its integration into the profitable and prestigious SpaceX provides xAI with both financial resilience and greater credibility among investors. Some commentators have interpreted this merger not merely as a “synergy” but as a lifeline for xAI, which makes sense in this context.

Particularly when considering SpaceX’s upcoming public offering, the timing is striking. The AI label and the “AI + space” narrative create strong appeal for investors. Musk has repeatedly used this strategy in Tesla, where value is created not based on current financial data but on promises of the future.

Ultimately, the SpaceX–xAI merger may present a compelling narrative about the future of space-based data and AI infrastructure, but this narrative could be nothing more than a marketing strategy. While parts of this story may be intended for implementation, the primary purpose of the entire narrative appears to be a pragmatic maneuver aimed at attracting investors.

Google plans to invest $40 billion in a new data center in Texas. This move once again highlights that the physical boundaries of digital technology are being redrawn, and AI competition has transformed from a battle of algorithms into a heavy industry contest based on concrete, steel, and gigawatts of energy. On November 14, 2025, Google announced that its massive infrastructure project would involve building new campuses in the Armstrong and Haskell regions, making it one of the largest capital expenditures in the United States. On Wednesday, Google further surpassed this commitment by increasing its projected AI infrastructure spending for 2026 to $185 billion. It has been noted that revenue generated from Google’s cloud technologies has also contributed to this increased spending.

The timing and scale of this investment are clearly a response within the industry. The technology world has been shaken for some time by Microsoft and OpenAI’s initiatives, particularly the reverberations of the Stargate project. Just as the Stargate project, with its multi-billion-dollar budget, aims to fundamentally transform existing computing infrastructure, Google has declared with its Texas project that it will not fall behind in this race. The result is a “compute” arms race, where companies compete by consuming more processing power, data, and energy to develop stronger models.

The geographic choice of investment is also noteworthy. The state of Texas has become a hub for companies in this new era due to its vast land availability and, more importantly, its flexible energy regulations. Google’s specific selection of Haskell County and its plan to directly link one of its facilities to a solar energy and battery storage plant appear designed to address the immense energy demands of AI operations. The energy required to train and run AI models far exceeds that consumed by traditional data centers. Thus, these facilities are not merely information processing centers but colossal energy consumers. Google’s commitment to adding 6,200 megawatts of new generation capacity to the grid to meet this demand underscores the sheer scale of the issue.

Energy is the most critical and controversial aspect of this investment. Texas’s electrical grid, managed by ERCOT, has proven vulnerable in the past, as seen during winter storms. Although Google has established a $30 million Energy Impact Fund to address local community concerns, this amount remains symbolic compared to the $40 billion investment. The use of local water resources for cooling systems and the potential pressure on regional electricity prices are shattering the illusion that this technology is clean and cost-free.

When examining competitors, the picture becomes even clearer. Other giants such as Amazon and Oracle are also seeking gigawatt-scale power sources, including nuclear energy options. However, Google’s Texas initiative has raised the stakes of this race to a new level. While the Stargate project promises a single centralized hub of immense computing power, Google is pursuing a more dispersed but collectively massive capacity strategy. This approach may be more sustainable both in terms of risk management and diversification of energy supply. Yet the ultimate goal remains nearly identical across all companies: to meet the scaling demands of AI models and produce stronger models than competitors. Economically speaking, $40 billion exceeds the annual GDP of many countries. The fact that this sum will be spent by 2027 reveals just how alarming the concentration of capital in the technology sector has become.

In conclusion, Google’s $40 billion fortress in Texas signals the end of the romantic phase of the AI era and the beginning of its industrial phase. The stories of startups built in garages have given way to multi-billion-dollar energy agreements and infrastructure projects on a national scale. Projects like Stargate and Google’s move demonstrate that the future of technology will not be determined solely by the quality of code, but by who controls the electricity and chips that run that code.