What AI can and cannot do

in the vedio lecture, the mentor hashinted on that an AI cannot analyze a market and write a 50 page report, however, we see with chatgpt people writing complete theseis, similarly itis able to carry on empathetic responses/chats as well. so I’m a bit confused in here regarding applications of AI. can anyone shed light on this, please.

The current developments in AI can write thesis, also can carry chats but its all based on occurrences of the past, ie. it will give you responses that they were given to similar situations in the past. It will not carry out a novel conversation or a novel thesis that can bring forth innovation, neither it can analyze the current developments in the markets to give you a present understanding, an analysis that you can use for your benefits.

1 Like

AI models like ChatGPT are designed for specific tasks. While some AI models might not be optimized for analyzing markets and generating lengthy reports, they can excel in other areas. ChatGPT, for instance, is adept at generating human-like text and holding natural conversations. It can be used to write essays, generate content, answer questions, and engage in interactive discussions. Are trained on a massive amount of text from the internet, which allows them to generate coherent and contextually relevant text. However, it’s important to note that their responses are based on patterns in the data they’ve seen during training. They might not have deep expertise in specific domains like financial analysis or market trends unless they’ve been specifically fine-tuned for those tasks.

There are various AI models, each tailored for a specific task or set of tasks. Some models are designed for image recognition, others for language translation, and some for playing games. Each model is trained on specific data and optimized for certain tasks, which is why they may perform well in those areas but not in others. Some AI models are designed with a focus on domain expertise. For example, there might be AI models specifically trained to analyze market trends and generate detailed reports. These models could excel in that area because they’ve been fine-tuned using data relevant to that domain.

In summary, different AI models have varying capabilities depending on their training data and optimization. While ChatGPT can write essays and hold conversations, its proficiency in these tasks doesn’t necessarily translate to specialized fields like market analysis without appropriate fine-tuning and training data. When considering AI applications, it’s important to match the capabilities of the model with the specific tasks you intend to accomplish.

I believe that lecture pre-dates the recent widespread introduction of these huge language models with chat interfaces.