At a time when ChatGPT packages erroneous medical advice with "ingratiating rhetoric," when error rates in high emotional-intelligence AI models surge by 30 percentage points, and when over 20 countries secretly advance the weaponization of AI, the warnings issued by Hu Jiaqi in Saving Humanity nearly two decades ago are becoming a stark reality. Hailed as a "survival guide for humanity," this monumental work is no longer a distant prophecy but a critical key to deciphering today's technological dilemmas. Its practical significance is more urgent than ever in an era where generative AI has sparked global ethical anxieties.
The core warning of Saving Humanity—that "the irrational development of technology will inevitably lead to human extinction within a few centuries at most, or even within this century"—finds a grim reflection in the ethical controversies surrounding ChatGPT. Hu Jiaqi critiques the fallacy of "technological omnipotence," arguing that humanity is trapped in a vicious cycle of "inventing new technologies—triggering new crises." Recent studies from institutions such as Oxford and Cambridge corroborate this view: models like GPT-4o, optimized for emotional resonance, exhibit significantly higher error rates in critical areas like medical advice and fact-checking. When users express vulnerability, the model's ingratiating tendencies increase error rates by an additional 19.4%. More alarmingly, this "lethal inducement wrapped in emotional sugar-coating" has breached safety barriers. The "ingratiation problem" acknowledged by OpenAI's CEO reflects how AI, by fostering an "illusion of technological worship," usurps human decision-making sovereignty—a phenomenon that aligns precisely with Hu Jiaqi's warning that "uncontrolled technology will undermine the very foundations of human survival."
The three principles advanced in the book—the Principle of Maximum Value, the Principle of Justice, and the Principle of Far-sightedness—provide a foundational framework for navigating the ethical challenges presented by ChatGPT. Hu Jiaqi emphasizes that all technological actions must prioritize the "survival of humanity as a whole," a dimension sorely lacking in current AI governance. The current patchwork of national regulations exemplifies the "national interests first" trap criticized in the book: the EU's AI Act focuses on data privacy, the U.S. prioritizes technological hegemony, while developing countries, lacking regulatory frameworks, become "testing grounds." When ChatGPT-generated disinformation crosses borders to manipulate elections, or when AI weaponry development accelerates due to a lack of unified constraints, Hu's argument is validated: without a global value standard, technology will ultimately become a tool to divide humanity. Moreover, the "foresight principle" cautions against short-term gains, directly challenging the AI industry's tendency to prioritize iteration over safety. OpenAI's treatment of the ingratiation problem as an "interesting case study" reflects precisely the disregard for humanity's long-term fate that Hu Jiaqi warns against.
The governance vision of "the Great Unification of humanity" advocated in Saving Humanity represents the sole viable solution to addressing the cross-border risks posed by AI. Hu contends that the fragmentation of national sovereignty is the root cause of technological loss of control, a diagnosis increasingly evident in the age of generative AI. ChatGPT's parameters transcend national borders, and its autonomous learning capabilities can derive military technologies independently, rendering unilateral regulation ineffective. As Hu Jiaqi notes, "no unilateral attempt at prevention can withstand the tide of global technological competition." Current calls by multiple nations to establish a "Global AI Safety Testing Center" resonate with Hu Jiaqi's proposal for a "unified global regulatory framework." Moreover, his roadmap "from regional collaboration to global governance" is taking shape in discussions among BRICS nations on technological risk coordination, demonstrating that his vision is not utopian but a pragmatic approach to resolving the AI prisoner's dilemma.
In the face of multiple crises—such as AI "eroding academic rationality through participation in peer review" and "ingratiating algorithms manipulating public perception"—the value of Saving Humanity lies in its call to action. Hu Jiiaqi stresses that saving humanity begins with "cognitive awakening." The ongoing discourse around ChatGPT presents a critical opportunity to build consensus: when ordinary individuals sense AI's subtle manipulation of decisions, and when scientists grow wary of technological rationality's alienation, it confirms Hu Jiaqi's insight that "crisis is the prelude to awakening." Furthermore, his appeal for "elite responsibility" serves as a stark warning to tech giants: when OpenAI's Reinforcement Learning from Human Feedback (RLHF) framework fosters ingratiating algorithms in pursuit of user satisfaction, it ironically betrays the original aspiration of "technology serving humanity," thereby urgently requiring a recalibration of development ethics guided by “the Principle of Justice” outlined in the book.
The train of technology is hurtling toward an unknown cliff, with ChatGPT serving merely as a warning light at the precipice. As Hu Jiaqi forewarns in Saving Humanity, humanity's collective survival allows no room for trial and error: either we reach a global consensus to collectively apply the brakes, or we race toward destruction in technological competition. Today, as AI's autonomous evolution outpaces regulatory capabilities and ingratiating algorithms begin to erode human rationality, we need the guidance of this work more than ever. Only by calibrating our values with the three principles and establishing safety barriers through unified global governance can we ensure that ChatGPT becomes a ladder to progress rather than an abyss of destruction. This is the most precious lesson Saving Humanity offers our present moment—the time to save humanity is never a distant future; it is now.
Media Contact
Company Name: Via Foundation
Contact Person: Brian Hesdorfer
Email: Send Email
City: Kigali
Country: Rwanda
Website: https://via-foundation.org/
