Artificial intelligence is progressing at a breakneck pace, and GPT-5 looks set to mark a new milestone in this field. Mira Murat, Chief Technology Officer at OpenAI, shared some exciting previews of what we can expect from the next iteration of the popular language model.
A quantum leap in the cognitive capabilities of machines
In a recent interview, Murati described the transition from GPT-4 to GPT-5 as an individual's leap from high school age to college. If GPT-3 had an intelligence comparable to that of a child and GPT-4 that of a bright high school student, GPT-5 promises to achieve an intelligence level equal to that of a PhD, at least for specific tasks.
This represents an extraordinary progress in the field of artificial intelligence, confirming the rumors that have long been circulating about the potential of GPT-5. Kevin scott, Microsoft's CTO, had already hinted at next-generation AI systems capable of passing doctoral exams thanks to better memory and reasoning capabilities.
Timing and expectations for the launch of GPT-5
Murati also provided some indications on the release timing of GPT-5. Contrary to initial speculation predicting a launch by the end of last year, or even this summer, it looks like we'll have to wait until the end of 2025 or the beginning of 2026 to see this new generation of AI in action.
While the wait may seem long, it is important to consider the magnitude of progress that GPT-5 promises to represent. As Murati points out, these systems have already reached a level of human intelligence in specific tasks, but in many other areas there is still much work to be done.
The potential impact of GPT-5 on society and technology
The advent of GPT-5 could have significant repercussions across multiple industries. From healthcare to finance, education to law, an AI with university-level cognitive capabilities could revolutionize the way we work and live. Obviously, the ethical and social challenges that such advanced AI could pose also rise to the level. Issues such as data privacy, transparency and algorithmic ethics (to use the term recently used by Jorge Bergoglio, “algorethics”) and the impact on the labor market will require careful consideration and regulation.
In this sense, the time that separates us from GPT-5 is an extraordinary opportunity to fix things, and it's not that much. Because, as highlighted by Mira Murati, progress in this field proceeds at a rapid pace. In recent years we have witnessed an unprecedented acceleration in the development of increasingly powerful language models, capable of understanding and generating natural language with astonishing fluency and precision.
If we look at the trajectory of improvement, systems like GPT-3 had perhaps child-level intelligence. Then, systems like GPT-4 are more like smart high school intelligence. And then, over the next two years, we're aiming for doctoral-level intelligence for specific tasks. Things are changing and improving rather rapidly.
Mira Murat, CTO of OpenAI
The future of AI
GPT-5 will be an unthinkable step forward in the field of artificial intelligence, and "in cahoots" with theembodiment it will set new frontiers in terms of the cognitive capabilities of machines. We must, I will never tire of writing, develop robust regulatory frameworks to ensure that AI is used responsibly and transparently, protecting citizens' rights and promoting the common good.
The "university career" of the next AI has just begun: let's prepare well for the next exam.
