Podcast (English): Play in new window |
Download (Duration: 25:35 — 17.1MB)
Subscribe:
Apple Podcasts |
Spotify |
Android |
Is AI More Than Just a Tool?
Many predictions regarding AI are influenced by the personal agendas—often political or ideological—of their authors, leading to biased interpretations. A more useful analytical approach is rooted in facts. This article will focus on two significant reports pertaining to AI.
The first report is Fred Cavazza’s analysis of AI’s societal and economic impact (original post in French), which characterizes AI as a force of drastic disruption. Having known Fred for years, I appreciate his extensive knowledge in this domain, making his insights particularly relevant. With his permission, I have translated his work from French to illuminate this topic.
The other report is authored by JP Gownder from Forrester, whom I plan to interview shortly. This discussion will help evaluate Fred’s assertions about AI’s disruptive capabilities. Our goal is to equip readers, especially students with pressing questions, to discern essential truths amid the noise.
AI is Reshaping Society and the Economy
We cannot view AI merely as a technological update. As a critical catalyst for productivity and automation, AI is fundamentally altering the economic and social fabric of our society. The productivity enhancements associated with AI are transforming office roles and creating disparities between users who leverage its capabilities and those who don’t. A pressing question arises: how do we effectively incorporate these synthetic entities within our collective frameworks? The conversations surrounding taxation, legal status, and psychological acceptance are critical as we navigate this new terrain.
AI IN SUMMARY — Key Takeaways
- AI is catalyzing a transformation in our civilization, marking the beginning of the fourth industrial revolution by transferring human cognition to machines.
- The productivity benefits of AI are significant yet unevenly distributed.
- AI agents are redefining how white-collar professions deliver value.
- We must develop a new legal and fiscal framework to accommodate the integration of AI.
- AI’s socio-economic consequences extend well beyond employment concerns.
AI on the World Stage: Davos Agenda
This week, the world’s leaders convened at the Davos Economic Forum, where AI, alongside Big Tech, took center stage. AI, Big Tech, and Trump Shine Most Brightly at the Davos Show.
AI is shaping discussions that extend far beyond mere technology:
“Artificial intelligence will displace so many jobs that it will eliminate the need for mass immigration.”
While I won’t delve into everyone’s viewpoints, each carrying varying biases, it is evident that major shifts are imminent:
AI experts are the prominent voices at this year’s Davos Forum, sharing insights on anticipated changes: DeepMind and Anthropic CEOs anticipate AI’s disruption of entry-level jobs in 2026.
A passive approach to AI’s evolution seems unreasonable while its revolutionary impact unfolds. Yet, opportunities remain; countries in the global south are already adapting: The AI Revolution Needs Plumbers After All.
Nuanced Productivity Gains of AI
I have often discussed the implications of generative AI (Superintelligence will exponentially enhance our capabilities and Digital divide is an undeniable issue). Although consensus exists around the impact of generative models, disagreements persist regarding the timeline for AI’s widespread adoption. The prevalent narrative remains that general AI remains a distant ambition, with human intelligence staying ahead of machines.
The Nature of Intelligence
This is where ambiguities arise: first, intelligence takes on multiple forms (Theory of multiple intelligences and What’s your intelligence type?); second, not all office work entails emotional or social intelligence. Most service sector roles primarily involve managing information and data—a task AI can execute proficiently.
For further insight into the rapid transformation AI is inducing in office roles, I recommend examining Claude’s publisher’s latest macroeconomic report: Anthropic Economic Index 2026.
Anthropic’s Economic Index 2026
This study, in its fourth edition, analyzed various activities using precise indicators: New building blocks for understanding AI use.
Significantly, the research illustrates a notable increase in the adoption and proficiency of generative models. Specifically, they recorded a 30% growth in Claude usage primarily attributed to the API—indicating rapid acceptance among advanced users (e.g., IT professionals) contrasted with slower adaptation by typical office workers.
The Growing Divide
A widening gap is apparent between those adapting to new practices (collaborating with AI) and those continuing with outdated methods from previous decades. This divide is concerning, as Claude’s latest version (Opus 4.5) possesses abilities comparable to individuals with extensive formal education.
This raises the pressing issue: how long can employers justify salaries for recent graduates when much of the work could be delegated to AI? Despite modest average productivity improvements (1.8% according to recent figures), AI’s efficiency in specific tasks is remarkable:
- Approximately 14 minutes to produce a lengthy article, compared to 3 hours without AI support;
- About 5 minutes to analyze a complex data set, versus 1 hour and 45 minutes without AI aid.
You might argue that these metrics are skewed, as they derive from highly proficient AI users, but this is not the case—the study encompasses regular employees who achieved a 67% success rate with outsourced tasks.
In essence, for a substantial portion of tasks, AI can reduce processing times by 10 to 20 times. Basic mathematics suggests that AI may enhance efficiency threefold, cutting average task completion times to a third of previous durations.
(Hint: McKinsey challenges graduates to leverage AI chatbots during recruitment updates)
The Era of Agentic Workers is Approaching
To clarify, the productivity gains mentioned apply to advanced AI usage beyond merely conducting basic searches or using applications like ChatGPT. We are discussing the full potential of generative models, specifically intelligent agents (see Agentic Web: the revolution that won’t wait for you).
Intelligent Agents
The promise of intelligent agents has become increasingly apparent to everyday employees, particularly following the introduction of Claude Cowork, which vividly illustrates the capabilities of agentic AI: Claude is Making Waves, Even Among Non-Technical Users.
This realization echoes in financial markets, which brace for declining revenues among traditional software providers, with one of France’s largest IT service firms downsizing and European banks poised to follow suit.
Adoption Levels and Their Implications
This topic demands careful consideration, as adoption rates can be contentious (as previously mentioned, it’s not a binary situation), and benefits can vary significantly (Why AI Enhances Creativity for Some Employees but Not Others). However, it is undeniable that AI agents are prompting a reevaluation of value creation for white-collar workers and the tertiary sector, which constitutes a substantial portion of France’s GDP.
Regardless of opinions or awareness, we are amidst a profound societal transformation, as AI’s proliferation accelerates the fourth industrial revolution, prompting upheavals whose full impact we have yet to comprehend.
Indeed, grappling with AI is complex (We don’t need more advanced AI, but a better grasp of AI). Tools founded on generative models necessitate behavioral changes, which will take time to implement. Nevertheless, we must prepare ourselves for the forthcoming transformations, recognizing that they are already in motion.
AI Represents More Than a Tool: A Fundamental Shift
The advent of generative AI and the pursuit of superintelligence signify more than technological progress initiated by computers and smartphones. We are witnessing a transformation that signifies the dawn of the fourth industrial revolution (Waves of change: Recognizing the driving forces behind cycles of innovation).
We are facing not merely a new technological phase, but an essential restructuring of our economic and social foundation: this time, we delegate not physical capacity but cognitive functions and creativity to machines. Regardless of when artificial general intelligence arrives, we live alongside autonomous entities capable of independent decision-making—be they digital (AI agents) or physical (robots).
This situation raises an unprecedented question: how do we incorporate artificial entities that significantly contribute to wealth generation while consuming substantial resources into our societal structures? Historical precedents—particularly regarding our gradual integration of domesticated animals—provide valuable insights.
AI is Not a Tool: From Analogy to Legal Reality
Humans cohabitate harmoniously with domesticated animals that have significantly shaped our evolution:
- Horses have facilitated exploration, warfare, agriculture, and transportation…
- Dogs have aided in hunting and security…
As animals contribute tangibly to our society, they are afforded rights and services:
- Guide dogs for the visually impaired receive recognition and support (a job and a purpose);
- Police dogs play pivotal roles in combating narcotics and are assured retirement security.
If animals offer significant contributions, could we similarly integrate AI that provides equal or greater societal value?
While it may be tempting to equate AI agents with newly adopted species, this analogy confronts ethical and legal challenges. Unlike domesticated animals, AI represents an information processing system devoid of a conscience or sentience.
The appropriate comparison may instead be with corporate entities (companies) since, like corporations, an AI:
- Contributes to wealth generation (through task automation, content creation, etc.);
- Utilizes infrastructure and consumes essential resources (energy, rare materials, water);
- Has rights (intellectual property) and responsibilities (transparency, accountability);
- Acts autonomously.
This comparison is relevant as it helps us shape the legal and social framework needed for integration.
The New Social Contract of the Synthetic Era: Taxation and Responsibility
We must approach the integration of intelligent agents not by granting them human-like rights—an absurd notion for a computer system—but by affording them legal personalities (similar to companies, NGOs, or governmental bodies).
The pressing question remains not about rights for AI but how to assign legal frameworks that clarify accountability. The concept of electronic personality, which was discussed in the European Parliament in 2017, aims to establish a framework for integrating these entities into our legal system, ensuring human protection and equitable distribution of benefits to avoid exacerbating wealth concentration.
As robots and AI systems begin to replace human jobs, they erode the base of social contributions derived from salaries. However, because they still contribute to economic activities and incur community costs (energy, e-waste management), it is reasonable to integrate them into our tax systems.
This doesn’t mean taxing AI agents as individuals. Instead, we should apply taxes on the economic value they generate through their operations. In return for this contribution, AI (or its creator) would not be entitled to social benefits (like pensions or healthcare), but they would gain a civil responsibility framework (fiscal, legal, and social). Such a system would allow for damages caused by AI to be addressed without always tracing responsibility back to the creator, who may not control the actions of the model.
Broader Socio-Economic Implications of AI
It’s important to recognize that issues surrounding AI extend beyond economic matters.
A Comparison of Domesticated Animals to AI
Revisiting the analogy of domesticated animals, we note that dogs serve not just as pets but also as emotional support animals. The term “emotional support animals” refers to those that provide comfort for individuals, especially the elderly or those suffering from chronic depression.
Likewise, domestic robots constitute key elements of Japan’s Society 5.0 initiative—providing care for the elderly physically (assisting with daily tasks) and mentally (stimulating memory through conversation).
While this prospect may be daunting for many in the West, it represents a necessary solution to Japan’s demographic challenges. In China, parents burdened with work-offer AI-enhanced toys that tell stories and answer questions, addressing their children’s inquisitiveness.
Furry Companions
This notion originated in Japan (Casio launches AI-powered furry robot pet that aspires to replace your dog), but it has gained traction in the West too (‘I love you too!’ My family’s unsettling week with an AI toy).
While this all may sound like the stuff of science fiction, these sociotechnological explorations have evolved over many years (Sony’s Aibo was launched in 1999).
Is the debate about the merits of emotional support robots our foremost concern? Clearly, we must prioritize tackling the immediate challenges. However, engaging in this discourse is essential as it prepares us for the stark realities of AI adoption, which currently lags in Europe—not due to functional obstacles, but emotional resistance stemming from a misunderstanding of AI (evidenced by less than 15% adoption rate among enterprises): EU Digital economy and society statistics.
Ultimately, we must initiate conversations and engage in thorough debates to come to terms with the looming changes, foresee upheavals that will severely test our social contracts, and begin reimagining our societal agreements (From Web 4.0 to Society 5.0).
Regulation as a Key Factor in Integration
Rest assured, I’m not launching into a lengthy discourse on the advantages of universal basic income (a non-viable economic proposition), but it is essential to touch on regulation.
In fact, coexisting with synthetic entities (AI and robots) shouldn’t be likened to domestication, as with animals. Instead, we need to regulate the synthetic workforce that we can no longer afford to overlook.
The fundamental issue isn’t whether robots or AI deserve social benefits, but how the wealth generated by their creations can sustain our social systems while managing resource consumption—which introduces economic tensions (fluctuating energy prices) and geopolitical concerns (China’s monopolyDefining the market is crucial to any marketing effort. on rare minerals).
AI Disrupting Our Civilization?
Fred Cavazza has laid out a compelling narrative about the impending civilizational revolution driven by AI. I find substantial merit in his insights into the future of AI and its broader impacts. Some aspects may resemble science fiction, but real-world events frequently align with these speculative narratives (think of Altman’s fascination with Jonze’s film *Her*). Fred’s observations suggest that the influence of AI will extend beyond current technological advancements.
However, viewing this through the lens of time is crucial. I can conceive of the potential of Anthropic’s Cowork, but witnessing its full realization is still on the horizon, despite my extensive experience using Claude.
Time is of the Essence
Integrating these technologies for effective workflows—rather than just simplifying tasks—will require time. Agentic software holds immense promise, and glimpses of its capabilities are emerging. Nevertheless, the productivity improvements introduced by these technologies often manifest inconsistently, even among proficient users.
Recently, after a strategic mentoring session that spanned an hour and a half, I utilized my usual Claude project to craft an exceptional executive summary. However, achieving this involved navigating three complex software platforms while multitasking in the kitchen.
Don’t misunderstand: one day, we will completely integrate these advancements. Yet, the timing remains uncertain—it’s not happening just yet. Innovation demands patience and effort. As Fred points out, resistance to change is a perennial barrier to innovation; this lag in adoption is not unique to Europe, though the continent’s traditionality certainly plays a role.
The overarching impact of AI on employment will undeniably be profound, but it may take years to materialize in statistical data, as Robert Solow opined.
In summary, Forrester’s perspectives offer a more layered understanding, which we will explore with JP Gownder soon. Only time will reveal where the truth lies, but I have a feeling it exists somewhere in the middle. Such a realization may not be as romantic or alarming (depending on your viewpoint), yet four decades of witnessing technological advancements have instilled in me a sense of resilience.
