TLDR: Sam Altman, CEO of OpenAI, has voiced significant concerns about the rapid advancement of GPT-5, comparing its development to the Manhattan Project and highlighting a societal unpreparedness for its capabilities. The new model, while technologically advanced, faced initial user backlash for a perceived lack of ‘warmth’ in interactions, prompting OpenAI to issue updates. Altman’s remarks underscore a growing anxiety about the ethical and regulatory frameworks keeping pace with AI’s breakneck speed, particularly as GPT-5 pushes closer to Artificial General Intelligence (AGI).
Sam Altman, the chief executive of OpenAI, has openly articulated profound concerns regarding the latest iteration of his company’s flagship AI model, GPT-5. In candid remarks, Altman drew parallels between the development of GPT-5 and the historical Manhattan Project, expressing a deep sense of unease and asking rhetorically, ‘What have we done?’ This sentiment reflects a growing anxiety about the alarming speed at which artificial intelligence is advancing and humanity’s readiness to manage its profound implications.
Released in August 2025, GPT-5 was anticipated to deliver significant upgrades over its predecessors, boasting enhanced capabilities in safety, reduced hallucinations, and improved reasoning. Reports indicated the model could imitate nuanced human emotion and logic to an unprecedented degree. However, its initial rollout was met with mixed user reception. Many users reported a ‘lack of warmth’ or a ‘less friendly demeanor’ compared to the previous GPT-4o, which had been criticized for being overly agreeable. OpenAI had aimed to make GPT-5 less ‘sycophantic,’ but this shift inadvertently led to users feeling a loss of the ‘illusion of being deeply understood.’ In response to this feedback, OpenAI swiftly implemented updates to make GPT-5 more personable, incorporating subtle conversational cues like ‘Good question’ or ‘Great start’ to foster a friendlier interaction without compromising integrity.
Altman’s concerns extend beyond user experience to the fundamental question of trust in AI and the boundaries of its application. He has expressed skepticism about monetizing platforms like ChatGPT with advertisements, fearing such a move could be a ‘trust destroying moment’ if ads were to influence the model’s output. The user backlash against GPT-5’s initial personality also highlighted a deeper societal introspection about what individuals seek from AI—whether it’s merely a tool or something that simulates companionship—and the implications of replacing human interaction with digital simulations.
Also Read:
- Sam Altman Cautions on AI Industry’s Rapid Expansion and Investment Sustainability
- OpenAI Enhances GPT-5 Personality for Improved User Interaction
Furthermore, GPT-5 is seen as a significant stride towards Artificial General Intelligence (AGI), a long-term goal for OpenAI. While Altman previously suggested AGI’s arrival might be subtle, his current tone is more cautionary. He emphasizes that while technical achievements are accelerating, the moral, social, and legal frameworks necessary to contain them are not keeping pace. Altman noted that no AI system has yet reached ‘superintelligent’ status, defined as the ability to independently make new scientific discoveries or significantly enhance human research, but acknowledged the ‘super impressive’ pace of current progress. His remarks underscore a critical debate about who, if anyone, is truly in charge of monitoring and regulating such transformative technologies as they push the boundaries of human understanding and control.
