So which AI is right for me? Well, that’s a bit like asking: Which car should I buy?
The short answer: The one you want.
The long answer begins with the question: Well, what do you want to use it for?
AI is not the magic wand with all the answers – as I have discovered the hard way. It can’t think for you, no matter what the promotions and marketing messages say. And no, it can’t reason for you either. In fact, I am not even sure it can reason for itself!
Background Questions before you Select an AI platform
But AI platforms can be great support systems and choosing a platform becomes easier once you figure out the answers to the following basic questions:
- Is cost a factor?
- Do you have some background in machine learning or AI?
- Does popularity or user stats matter?
- Why are your friends and family using the platforms they are using?
- Is it for general use or a specific purpose (research, study support, self-learning, assessments)?
On the cost front, all the platforms I have tried have a free model (you have to sign up with an email of course). The free or basic versions have some limitations but they are pretty good to get started and evaluate the platforms. If you have some understanding of machine learning or artificial intelligence in general, it can help in prompting, but you don’t need any specialised background to start using a generative AI platform. Popularity or market share by number of users matters mainly if your decision is driven by peer or market opinions, in which case one might just pick the most popular platform irrespective of use or cost.
Practically, however, it usually comes down to the purpose for which you will use the AI platform. And this is where it is helpful to have some information about different platforms: what are they built for, what are their strengths, what do they not have, what can they do, what can’t they do, how well do they do what they say they do, and so on.
Popular Generative AI Platforms (Partial, non-exhaustive list)
So which are the generative AI platforms out there? ChatGPT, Quillbot, Gemini, CoPilot, Grok, Claude, Perplexity, Jasper, Poe, to name a few. But all of them are not comparable, apples-to-apples so to say. I would say what most of us look for in the beginning are called generative AI chatbots or conversational AI platforms. Following is a summary of a few such platforms. The statistics quoted are as of May 2025 and they are from StatCounter Global Stats. I haven’t used Copilot or DeepSeek for too long therefore haven’t provided a remark. I do use a few others such as Grok and Stability but have not included them in the summary table since I haven’t yet pegged their utility.
Platform | Market | Strength | Content | Remarks |
ChatGPT | 84% | Conversational, general purpose assistant. | Text, Images, Audio | Very easy to use, image generation has improved a lot. |
Perplexity | 12% | Research-oriented assistant | Text, Images, Audio | Provides multiple models to choose from (ChatGPT, Claude, etc). |
Gemini | 2.3% | Multimodal, integrated with Google ecosystem | Text, Images, Audio, Video | Works well if one is primarily in Google ecosystem |
DeepSeek | 0.9% | General purpose | Text | NA |
Claude | 0.3% | Ethical, reasoning conversational | Text | Good fit for conversation-based learning, guided self-study |
Copilot | 0.2% | Productivity integration with Microsoft 365 | Text, images | NA |
Selecting One (or 2 or 3) AI platforms for yourself
So how do you choose from a list of AI platform? I have been using Generative AI platforms for more than two years now and have found it helpful to ask questions like: Am I going to be generating a lot of images or videos? Will I be doing a lot of text content structuring or text content generation? Do I need audio to text generation? And so on. With respect to media types, generative AI can broadly be looked at as being platforms for: Text Generation, Image Generation, Video Generation, Audio Generation, Code Generation, 3D Environment Generation and Multimodal (text, audio, video etc combined).
If I am not using image generation and mainly working with text content, Claude works the best (in my experience) for self-study or brainstorming or research type of tasks. This is because Claude has the ability to engage in meaningful conversation and also question itself if prompted. I use Claude the most often for brainstorming and research.
ChatGPT is the easiest to use, and has the best balance of all the platforms I have used. Which means that the conversational ability is good, image generation is more than acceptable for informal / social media / blog kind of posts, and research/information retrieval / structuring is comparable to other platforms if not better. Having text, audio and image generation within the same platform makes it very convenient. If you have a paid subscription you also get access to video generation from Sora.
If you are going to need to check responses from multiple platforms, Perplexity is the one. It offers choice between ChatGPT, Claude, DeepSeek, among others. Perplexity doesn’t work for me because it doesn’t seem to add anything of value to the models (I already have subscriptions to them individually). Gemini is good when I am using Google ecosystem since everything is seamless, but outside of that ecosystem, I switch to ChatGPT. My experience is similar with Copilot: it’s convenient when I am on a Windows laptop with Office 365, but most of the time I am on a MacBook and so I stick to Claude and ChatGPT for general purposes.
I do use several other AI platform such as MidJourney, Jasper, Leonardo, Otter, etc but those are not general purpose platforms, so will leave that discussion for a different day. For the time being, if you are trying to figure you which model works for you, I would say sign up for a free basic account for different platforms, use them, evaluate their responses from your own perspective and then pick one or two or more! Also observe how much continuous use you do on a platform and when the free limits run out. Typically, the limits are reset after a few hours. If you can wait and it does not obstruct your work, the free model should suffice and you can upgrade later as and when you require it.
Bonus Tip
Here’s a practical tip: all AI platforms are competing with each other to add features, improve capability etc. So keep checking your platforms every few months (specially after they release a new version/model) since they get better and better.
Study with Goseeko: https://www.goseeko.com/landing/international-students/
UAE Students Register here: https://www.goseeko.com/landing/middle-east/