Navigating the AI Zoo

In a world where 'AI said this' is a common refrain, have you ever wondered which AI? When we say something is 'GPT-powered', it's almost like saying a car has an engine – but what kind? Let's talk about the incredible sprawl of LLMs with their distinctive capabilities, the need for everyday users to recognize different AIs like they do brands of cars, and why it's crucial to understand that there is no one-size-fits-all AI that knows everything.

Who is this one, all-knowing AI that people keep referring to? You've probably heard phrases like, 'According to AI, this is what the future holds,' or 'AI suggests this is the best course of action', as if there's a single, omniscient digital oracle. In reality, the AI landscape is full of chatbots and tools like ChatGPT and Bart, which, in turn, utilize a variety of models ranging from GPT-3 to GPT-4, Google Gemini to Anthropic Claude. Each of these models (shall we compare them to car engines?) has its own unique strengths and weaknesses, and the tools and chat bots that make use of them are tailored for specific applications and contexts. And they also evolve over time. They learn, adapt, or even get censored – you name it. Essentially, let's acknowledge that there isn't a single AI with its own will and consciousness, at least not to my knowledge. But I'm sure this is common sense, so let's move on.

However, it's worth noting that my experience with ChatGPT might differ from yours, and my satisfaction with the results may not be the same as yours. Imagine you're using ChatGPT, and the results are just not what you expected. You've heard stories of people achieving incredible outputs with ChatGPT, and you can't help but wonder, what are you missing? The answer might be simpler than you think, and it's twofold: the engine under the hood and the skill of the driver.

Mastering AI: Prompt Engineering & Language Skills

You see, those miracle-makers might be using a paid version of ChatGPT, powered by the more robust GPT-4 engine, while your free version is cruising on the somewhat less capable GPT-3.5-turbo. It's like comparing a high-performance sports car to a reliable family sedan. Both can get you from point A to point B, but the journey and the capabilities can be vastly different.

But there's another layer to this - learning to 'drive' these AI models. Just as a powerful car requires a skilled driver to truly utilize its potential, effective use of GPT models often involves a bit of know-how, particularly in the art of prompt engineering. It's not just about the AI you're using; it's also about how you use it. There's a growing number of language models to choose from, but don't worry too much, especially if you're not a technical person. Just take a look at the AI-powered tools available and try to find out what "engine" that "car" uses. As you navigate this AI zoo, it's important to not only understand the different 'species' of AI, but also learn how to interact with them effectively. It's all about learning to coexist and collaborate with them. If you've read one of my books, you'll find plenty of examples and in-depth discussions on this topic.

Understanding Language Models: The Key to Success

Do you need to be a data scientist or software engineer to understand how language models work, like a mechanic who can take apart an engine and put it back together? You don't have to, and this is where my car analogy falls short. You just need to develop an understanding of how those language models behave and react to language. If you want to make the most out of language models, the best advice I can give you is: learn how to work with... language! Sure, taking technical courses on scaling large language models in the cloud is useful, and learning Python and APIs helps you build smart, interconnected applications for users. But at the core, it is the understanding of language, improving your writing skills, and grasping linguistic nuances that will massively improve the results you get from language models.

The fantastic part is that people with writing skills and imagination clearly have an edge here – what a time to be alive!

Please check out my book "Neurons & Nonsense – How AI rewired my creative DNA", where you can read about my personal journey of human-AI collaboration and how it entirely changed the way I think, create, and explore ideas.

https://www.electricminds.net/book