top of page
  • Writer's pictureSarah Ruivivar

OpenAI's GPT-5: The Dawn of Superintelligence?

OpenAI's CEO, Sam Altman, recently dropped a bombshell in an interview with the Financial Times: the company has started building its next-generation AI model, ChatGPT-5, and it could possess superintelligence. But before we all start preparing for a Skynet-style takeover, let's delve into what this really means.

Firstly, the creation of this model isn't a solo endeavour. OpenAI is in a fruitful partnership with Microsoft, a collaboration that Altman describes as "working really well". Microsoft has already invested over $10 billion into OpenAI, and Altman anticipates more funding from the tech giant and other investors.

Why the hefty price tag? Building a major AI model like ChatGPT is a mammoth task, requiring billions of dollars, vast computer resources, and extensive fine-tuning and safety testing. But the potential payoff is huge: the creation of an AI that surpasses human knowledge and reasoning.

This concept, known as artificial general intelligence (AGI), is the next frontier in AI development. AGI goes beyond regurgitating information to expressing original, complex ideas. The prospect of such powerful AI has sparked debates about how to prevent these superintelligent systems from going rogue.

Microsoft's investment isn't just altruistic; the company has pivoted its entire business model around AI, integrating its Copilot system, built on GPT-4, into Windows and Microsoft 365. Altman's vision, however, extends beyond business applications. He sees the true product as intelligence itself.

The development of GPT-5 will require unprecedented processing power and data, sourced from public online data and purchased company data. Altman hasn't given a timeline for GPT-5's completion, but it's expected to take months, if not years, to train the model.

So, while the dawn of superintelligence is an exciting prospect, there's no need to start building your bunker just yet. The future of AI is coming, but it's still on the horizon.

Made with TRUST_AI - see the Charter:

12 views0 comments


bottom of page