Just as we were getting used to the wonderful features of GPT-4, new rumors surfaced about ChatGPT's next generation neural network GPT-5

This version of the Generative Pre-trained Transformer is said to be coming this winter and will take OpenAI technology to the next level

According to Siqi Chen, CEO of Runway, an AI video expert, GPT-5 has the potential to help chatbots achieve so-called artificial intelligence (AGI) In other words, an AI that is already brusquely chatty could have human-level capabilities to understand and perform tasks As Chen points out, even though there is some debate as to whether GPT-5 will reach the AGI level, by definition it will

It would also not be a leap to suggest that it could pass the Turing test

ChatGPT is already smarter than many of us on Monday morning, but the ChatGPT model trained on GPT-5 will need to meet several more criteria to achieve artificial intelligence However, it is difficult to know exactly what to expect, as it is clear that none have ever reached AGI

To qualify as an AGI, ChatGPT must demonstrate the same level of learning and understanding ability as humans Currently, ChatGPT needs its own specific prompts to work optimally, but AGI will show initiative and be able to consider context as well

Using early AGI tools, AI experts like Yohei Nakajima of venture capital firm Untapped are providing the world with several examples of AGI-level AI They instructed the AI to set the goal of starting and growing a business and to figure out what its first task would be The AI then searched the Internet for "how to start a business" and, having the ability to notice that the results were too broad, declared "no good search results found"

Eventually, it narrowed the search to "how to start a small business" and learned the requirements for that task From there, AI set itself the task of creating a business plan, developing a marketing strategy, etc, and dove deep into each element

On paper, AGI sounds like a Terminator-esque sci-fi nightmare (Google Bird already has many sci-fi opinions), but it will help to resolve the ambiguity that has hamstrung chatbots like the new Bing and Google Bird Of course, there will be legitimate concerns about the potential for disinformation and bias to influence the public's use of AI, but this is an exciting development nonetheless