GPT – 5 Pro impresses with its complex and corporate response to prompts. The jewel of the Crown of the deployment of the GPT-5 this month even made the CEO of Openai Sam Altman nervous with some of his answers. But you should not confuse brilliant algorithmic models with a real independent thought, according to Dr. Ben Goertzel, who helped popularize the term artificial general intelligence (AG) in the early 2000s.
Now, the CEO of the artificial superintelligence Alliance and Trueagi Inc., and the founder of Singularitynet, Goertzel wrote a test renting GPT – 5 pro as “a remarkable technical achievement” that he finds useful for formatting research documents, analyzing mathematical frameworks and improving his own prose. But, it does not confuse the capacities of the model with the real human style brain.
“These models, as impressive as they are, totally lack the creative and inventive spark that characterizes human intelligence at best,” said Goertzel. “More fundamentally, they” literally do not know what they are talking about “. Their knowledge is not based on experience or observation, it is the correspondence of models to an extraordinarily sophisticated level, but the correspondence of models nevertheless.”
No matter how fast the model performance is fast, it is ultimately superficial. You can be dazzled by the show, but nothing happens under statistical inference. People who see a blurred line between GPT – 5 pro and AG are not surprising, it hastened to add, because it can imitate logic, extend reasoning and resemble a process of reflection, but that is nothing like a human or animal brain. Recording associations from training is not the same as relying on memory, experience or a vision of future objectives.
“This distinction is not semantic nitpicking. The real act requires knowledge of the grounding in an external and internal experience,” wrote Goertzel. “In terms of these fundamental aspects of open cognition, today’s LLMs are far below a one -year human child, despite their incredible intellectual installation.”
The future of AG
The GPT – 5 Pro and its brothers and sisters are built on an increasingly tense premise that the scaling of languages of languages will inevitably act. He also suggested that the current LLM approach is merged to a business model that limits innovation. OPENAI, he notes, simultaneously tries to build act and sell evolving chatbot services to billions of users. The AG label, he warns, is thrown too freely. While GPT-5 PRO and other tools are undeniably powerful, to qualify them as mind is, in his opinion, premature and possibly misleading.
“GPT5-Pro deserves recognition as a remarkable realization in AI engineering. For researchers and professionals needing sophisticated technical assistance, he is currently unrivaled,” wrote Goertzel. “But we must not confuse the gradual improvements in the correspondence of natural language motifs on a large scale for progress towards authentic general artificial intelligence.”
Goertzel’s description of a real act is a model that constantly learns new things, whatever a user who interacts with it. The continuous evolution of a spirit, human experience, goes far beyond the specific training and deployment of an AI model. GPT – 5 Pro is frozen when it is deployed; A pot -sealed pot.
Goertzel’s work would crush this pot and disseminate intelligence through decentralized systems. Finally, he hopes to produce an intelligence that does not imitate the functioning of the brain, but works like one, with internal models of the world and beliefs that he would update over time.
“The path to AGE will not be found by simply expanding current approaches. It requires fundamental innovations in the way we waste knowledge, allow continuous learning and incorporating different cognitive capacities,” concludes Goertzel. “GPT-5 and his successors will probably play important support roles in future AG systems, but the main role requires more innovative actors, we are still creating.”
