SAN FRANCISCO, April 4, 2025 – In the era of ​great linguistic models ​and rapid progress in artificial ​intelligence (AI), the Silicon Valley elite increasingly ​believes that the next great transformation of humanity is not only imminent – ​it is already underway. But as a flow of predictions, manifestos and scenario reports have an impact on the ​public domain, some ​voices are beginning to wonder whether this technological ​optimism is rooted in science or part of a new ​kind of mythology.

Scientific journalist Adam Becker’s ​latest book, More Everything Forever, offers a sober but richly detailed exploration of what happens ​when unproven faith in AI confronts the ​disorderly ​realities of ​power, governance and historical repetition. His work dissects ​the ​ideology ​that feeds ​many of Silicon Valley’s most influential futurists – Ray Kurzweil and Eliezer ​Yudkowsky to risk capitalist Marc Andreessen – revealing a cosmopolitan where immortality is only ​a mental burden and spatial colonization is a solution to the finite resources of the Earth.

Becker describes this belief system as “technological salvation”, a secular extension of the theological narratives that promise eternal life and transcendent perfection. It is an ideology that raises all social problems – ethics, ​inequality, governance, even death – as solvent ​by code. And although such technological utopianism may seem aspirational to the surface, Becker’s vision ​is getting colder: these ideas are only experiences of thought. They influence AI’s policy, wealth distribution and global governance architecture.

Futuristic predictions ​of Ideology or Reality?

The report “AI 2027” – ​a fictional prognosis but based on data written ​by former OpenAI researcher Daniel Kokotajlo and AI analyst Eli Lifland – embodies Becker’s criticism. The report describes a near future in which AI systems quickly prevail over human intelligence, automate at exponential speeds and potentially ​threaten human survival. The authors, both ​in line with the movement of Effective Altruism, paint a living image of an ​AI firm called “OpenBrain” which releases successive generations of superintelligence agents culminating in ​Agent- 4, a ​system capable of making a year of progress of artificial intelligence within a week.

Asked whether humanity will ​still recognize the world by 2030, Kokotajlo hesitated, imagining utopias of robot factory or dark skies by the catastrophic collapse. This speculative ​prognosis covers the line between responsible future ​preparation and a new ​kind of technological ​mysticism – science fiction part, Silicon Valley gospel part.

What’s Driving the ​AI Elites to Think in Extremes?

According to Becker, and ​has resonated in criticism by means such as KCRW and Nature, what motivates this kind of thought is ​not only technological ​optimism, ​but a deep echo, often unconscious of religious desire. The ​transhumanist movement, for example, reinterprets classical spiritual themes ​through ​a ​technological lens. The mind becomes ​a digital resurrection; AI becomes the oracle of moral truth.

This convergence of ​capitalism and techno-messianism is shown ​in a surprising way. Becker argues ​that multi-millionaire technocrats have taken these visions not ​only to ​form cultural ​narratives, but also to ​consolidate economic and political ​power. “Expand their influence now,” he writes, “is often the real goal behind the claims to want ​to ​save humanity later. Therefore, the pressure for IA domination is not limited to innovation; It involves institutionalizing a world view that focuses on central control, predictability ​and benefits over the adaptability, equity and complexity of the real world.

To what extent are these ​forecasts based ​on historical models?

Critics like security expert Davi Ottenheimer, writing for FlyingPenguin, draw historical parallels ​that warn against the type of rigid prediction seen in AI 2027. By comparing its predictive logic with ​Napoleon’s catastrophic naval strategy ​- where centralized ​control and overtrust have led to the destruction of the French fleet by the more agile British navy – it argues that ​the industry of the AI falls into the same trap. Napoleon built ​massive ships decorated like ​the Orient; today’s AI giants are ​building mass computing centers. According to him, both monuments are vulnerable to over-centralization and arrogance.

Ottenheimer continues this analogy with ​references to the Maginot line – the ​defence system of the ​First World War was inflexible and ​outmoded ​in France that was overcome by German innovation in the ​Second World ​War. Just as the French generals ignored the signs of change and were based on outdated war patterns, he warned that modern AI centralists can ​ignore the subtle but powerful rise of AI’s open, ​distributed and decentralized technologies.

Is “Open Source” ​Still the Best Path Forward for AI?

Interestingly, IA 2027 authors contradict that “you have what you ​pay,” suggesting that expensive owner models will ​dominate. However, his own work – freely distributed – supported this statement. Ottenheimer reminds readers ​that almost all transformation technologies that support modern computing – from ​TCP/IP and HTML ​to Linux and HTTP – have ​been opened, not patented.

In fact, ​technologies such as Mistral, LLAMA and DeepSeek already fill the performance gap between closed and open models. And the historical lesson ​is ​clear: open protocols, ​precisely because they are accessible and adaptable, tend to save with time. This undermines the fundamental logic of the report and ​suggests that we can witness the ​increase of a new decentralized AI ecosystem ​despite the attempts of large companies to monopolize development.

Can former OpenAI employees predict the future of intelligence?

While open AI is used by the authors of AI 2027 as a ​title, critics argue that membership in a tumultuous organization should not automatically translate into prophetic authority. ​OpenAI has had to deal with repeated crises ​of ​governance, resignation and ​failure ​of deadlines, including IGA promises ​not kept in just a few years, a demand repeatedly ​made over the ​past decade with little to show for it.

Becker and others argue that “past affiliation is not a qualification; ​It’s a context.” Actually, I should speed up control. Have these people helped put in place strategies that have failed? Were they spectators ​or participants? Failure ​to respect a ​clear causal framework for their predictions weakens ​their ​authority. Citing David Hume’s criticism of inductive reasoning, ​Ottenheimer warns against ​the confused correlation of the past with future causality – a logical falsehood that is repeated ​throughout the report.

Is There a Better Framework for Thinking About AI’s Future?

Instead of ​accepting AI’s future as inevitable, ​Becker argues that we need to rethink these debates. What if AI was simply a tool to ​strengthen ​human collaboration and democratic self-government? What if communities, ​not businesses, ​determine how these systems are integrated into society?

It recommends a ​strong public dialogue, not led by billionaires or former OpenAI collaborators, but by ethics, historians and community leaders who understand technology not only as ​a technical challenge, but as a ​social and moral ​challenge. Becker even suggests radical political interventions such ​as the punitive imposition ​of billionaires to put an end ​to the widespread influence of technocratic elites.

Why ​historical illiteracy makes technical predictions dangerous

The deepest criticism, rooted in ​all sources, is that much of ​today’s AI predictions are affected by a dangerous combination of history and overconfidence. ​The same mistakes that led to disasters ​such as the cold of the port or the radio silence of the French generals before the fall of France during the Second World War are ​repeated in the meeting rooms of Silicon Valley. The trend is the same: deference to defective authority, inflexible assumptions ​and contempt for the lessons of the past.

When institutions isolate critical comments, they make bad predictions and ​worse policies. This is particularly dangerous when betting involves controlling systems that could shape everything from the labour market to the fundamental nature of human ​cognition. ​It is not ​enough to have technical skills, ​a deep understanding of history, ​philosophy and ethics is also ​essential.

That is why Becker’s call for balance, nuance and ​humility is so vital. His warning is not only to be torn apart by AI, but also by human institutions that do not rule ​wisely in the ​face of rapid change. The future is not something we can ​model with ​perfect ​precision ​- it’s something we build together with open eyes and open ​minds.

While technoelites publish more “scenarios” and science fiction forecasts disguised as a political guide, Becker’s voice reminds us that the ​real danger is not that machines become human, but that humans can forget how to ​be human in the process.