Thanks to ChatGPT an AI race has been started. Every company is rushing to create a new artificial intelligence service.
Currently, there are three big conversational AI services available for everyone. ChatGPT, CoPilot and Gemini. ChatGPT is by far the most famous of the three.
Created by a company called OpenAI. CoPilot is an AI service created by Microsoft and Gemini is a similar AI service created by Google.
Interestingly, Microsoft’s CoPilot is using ChaTGPT in the backend. These are absolutely free for anyone to use and they can be pretty handy for us in a lot of ways.
I personally use them to plan a trip, get cooking suggestions, write a big document easily, or even summarize a huge document into bullet points.
You can use them like Google search but 100 times more intelligent. You can have a proper conversation and use them for daily tasks like projects and ideas.
But which of the three is the best? What can each AI model do the best and worst? Let’s put all three of them to the test today in this epic ChatGPT Vs CoPilot Vs Gemini battle.
Personally, I was surprised by the results, to be honest. I will compare the three AI Services ask the same questions and evaluate them depending on the speed, accuracy, intelligence, and usefulness of the results.
We will also check if they are able to maintain context between questions properly.
It’s going to be a super insightful article. So let’s get started.
Before we start asking questions to each of them, remember that I’m using the free version of all three services. There is a paid version for ChatGPT, CoPilot, and Gemini but I’m going to be testing only the free versions, which even you can use.
Okay, how can we use these AI Services? All three AI Services – ChatGPT, CoPilot, and Gemini have a website where you can ask questions and get answers. Here are the websites of each of them.
ChatGPT: https://chat.openai.com/
CoPilot: https://copilot.microsoft.com/
Gemini: https://gemini.google.com/
These are the websites to start accessing each AI model. ChatGPT and CoPilot also have a smartphone app, which is a huge advantage to use them whenever we need it.
Google is also working on an app and there should be one for Gemini soon or maybe there is one already when you’re reading this article.
Let’s log into each of them and ask some questions to see how good they are.
In terms of interface, ChatGPT is very clean and minimal. I like Google’s Gemini as well, it has a very clean look. CoPilot’s interface is very cluttered in my opinion.
I’m starting with a very simple general knowledge question, which even Google answers easily – what is the largest building in the world? I’m using the Chrome browser for all three.
Before I hit enter, I’m going to copy-paste the question to all three AI models and see how quickly each of them answers. Wow, ChatGPT is fast. It finished even before I could hit enter in other windows.
Google’s Gemini is also fast. However, CoPilot is painfully slow. I think there is some issue here. So to make the comparison fair, I’m going to use the Microsoft Edge browser for CoPilot because I’ve seen a much better result with Edge because CoPilot is integrated with Edge. So let’s try that.
Still, charity is the fastest. Now, CoPilot seems much faster followed by Google. Fast is one thing, let’s see how accurate the answers are.
ChatGPT’s data is a little old as you can see in the image below. It’s not fully updated but it gave the answer right away accurately. It’s a very short and precise answer, nothing extra.
CoPilot and Gemini, on the other hand, gave a better answer in my opinion. It gave three different options based on what the largest meant and answered with a summary and links.
Even though ChatGPT was fastest others were comprehensive. Okay, CoPilot takes this one because it was comprehensive and slightly faster than Gemini.
Let’s ask a follow-up question in the same context: what are the new seven wonders of the World?
ChatGPT is again the fastest and gave a very crisp, bullet-point accurate answer. CoPilot, although it started answering first, kept typing and typing. Gemini gave the whole answer in one go, so effectively, Gemini was faster than CoPilot.
In this case. I also like how CoPilot and Gemini were able to give images as well, and CoPilot gave some facts as well.
In this case, my pick will be Gemini for being faster than CoPilot and for the way it presented it nicely. It may be subjective in your case if you like the CoPilot better.
So far, even though ChatGPT is fast and precise, Google and Microsoft are beating it to the punch by giving some extra information and presenting it nicely.
Now, let’s ask a question that is not readily available on Google: if I live in New York City, what is the best and fastest way to go to New Delhi?
I think we are starting to see the limitations of ChatGPT; it’s still the fastest, but its data is limited. It just told us to take a flight. CoPilot and Gemini are much better here.
Gemini straight away started giving results from Google Flights, which is nice, but it was not an exact response to the question I asked.
CoPilot, on the other hand, first gave me the answer I wanted and then gave me the flight options using Bing Flights. It was slow, but in the end, appropriate data is more important than a few extra seconds. CoPilot got this one better, followed by a close second by Gemini.
Now, let’s check how updated these AI models are. I’m going to ask: what is the latest phone from Samsung?
I think we should eliminate ChatGPT here, as it’s really lagging due to its old data. CoPilot, on the other hand, was very accurate, and it even gave the right specs for the phone.
Everything was accurate, and the features of the phone were also precise. I didn’t like the ads at the bottom, though.
Gemini, on the other hand, started rough. It said, as of today, 2023 s24 launched in 2024. Other than that, it gave a nice table of features and the right phones. The facts were slightly off, though, so it’s starting to get questionable.
No ads, so that’s a plus. CoPilot again wins this round. CoPilot is using the same model, GPT-4, just like the ChatGPT, but it has got more recent data using Bing, so it’s able to give more accurate recent results. So, is CoPilot the best AI? We are not done yet.
Also Read: How Are AI Avatars Transforming Ecommerce?
Okay, let’s use it as an assistant. Let’s see if it can plan a trip for me.
The question was a bit ambitious, but I think Gemini messed up this time. It did not give me a proper breakdown. It did not tell me how much to drive in a day, where to stop, etc.
ChatGPT was great; it looked like a proper plan and it assumed I needed to drive all seven days. CoPilot was even better, giving me a proper day-to-day plan, how much I needed to drive, where to stay, what to explore, etc., in a very readable view.
Let’s add a follow-up question; let’s make it three days. Okay, here are some mixed results.
All three gave a nice breakdown this time, but ChatGPT and CoPilot’s breakdowns were inaccurate. They were not split equally and not the shortest. Google gave a good split this time and, in fact, links to Google Maps directions.
This shows that all three can mess up and be good at the same time. You have to validate the results for yourself and use them as a guideline.
Okay, one more. Let’s ask for some car recommendations.
Based on the input, ChatGPT and Gemini stuck strictly to the requirements. CoPilot took the liberty to suggest cars beyond $50,000.
But if you ask me, Gemini gave the most expected results than the other two. The cars it recommended felt more agreeable to my requirements. ChatGPT was close, but Google did a great job here.
Okay, these are all text-based, right? AI should do more than that.
Let’s ask them to generate an image according to my instructions.
I’m going to give this prompt and see what each of them gives us. ChatGPT just bailed, as it cannot generate an image on the free version. CoPilot and Gemini gave us some amazing images here.
Seems like a tie between both, right? But if you look at the variations and look into the details, CoPilot did a much better job of getting the details so realistic here. If you look at the Gemini’s image, you may not be able to see a lot many details.
I gave another prompt. Google’s image looks fine from a distance, but if you zoom in, the details are in there. CoPilot is easily better so far for image generation.
Okay, one last test before we summarize the results. I will give it a website and ask it to summarize the results. ChatGPT was not able to read the website.
CoPilot seems like it could, but it did not do anything; it just gave me a summary of the website. Gemini did a little better. It did summarize what the website was, but it also read through the web page and gave a one-liner of the news. I verified the news page, and it did pick up the latest news from the web page. So well done, Gemini.
CoPilot does have a trick, though. Since it’s integrated into the Edge browser, you can simply open the web page on Edge, then click the CoPilot to ask to summarize the current page. This seems to be giving the best results, even better than Gemini.
Okay, there is a lot more that we can test with each of them: Fitness recommendations, integrations with AI like maps and flight bookings, software code generation, generating music, and much, much more. But based on the results you just saw, can you pick a clear winner?
One thing is clear for me: ChatGPT is very limited due to its old data set. It also gives text-only results, no images, no links, and no additional info. But it’s the fastest, so if you want any information without recent data or additional info, ChatGPT can give you quick and precise results.
Between Google’s Gemini and Microsoft’s CoPilot, it’s a mix-up. I like Gemini’s interface, while I preferred CoPilot’s summarizing of data. Sometimes CoPilot had accurate data, while other times Gemini was accurate, but clearly both are different.
In most cases, I think I’m going to stick with Microsoft’s CoPilot because its image generation and data set are far superior, and even if sometimes it doesn’t give the most precise result if I ask again with a different prompt, it does give me a much better answer next time. Another huge advantage with CoPilot is that it’s integrated closely into the browser windows and all other Microsoft apps, so if you’re into that, it’s going to be a great advantage.
All these AIs are in their infancy, and they have a long way to go to blindly trust them. Did you try these out? Which of these three did you find accurate and useful? Let us know.
Adverts help keep our content free. But we have detected that you are using extensions to block ads. Please support us by allowing ads for your visit. Thank you!
