Which AI Chatbot Is Best: ChatGPT, Microsoft Bing AI, or Google Bard? The internet has been on an AI-inspired rollercoaster since OpenAI published ChatGPT in November 2022. Google and Microsoft, 2 of the world’s most well-known tech companies, have since pushed hard to imitate the astonishing chatbot’s success. Both businesses now have a horse in the race. Microsoft has Bing AI, while Google has Bard. But how do these two new chatbots compare to the legendary ChatGPT? Which is the best AI chatbot: ChatGPT vs. Bing AI vs. Google Bard? Let us investigate.
AI chatbots, unlike search engines, provide a single answer to your query. So, when you ask a chatbot like ChatGPT a question, you will only get the response that ChatGPT believes is the best answer to your question. Furthermore, because there are no other sources of comparison, AI chatbots must provide the most accurate information possible. But, in terms of accuracy, how to do ChatGPT, Bing AI, and Bard fare?
Starting with an easy pop culture question, we asked each of the three chatbots to describe the iconic television show Breaking Bad in ten words.
We encountered an unanticipated accuracy issue even though the description provided by all three chatbots needed to be better. Bing AI answered with a 28-word description, significantly more than the ten words we asked for. We asked for a five-word description on the second attempt, but Bing AI returned a seven-word description.
Next, we attempted Google Bard. Like Bing AI, Bard failed to correct the word count, though he was considerably closer than Bing AI’s terrible attempt.
After that, we put ChatGPT to the test. The first attempt came near but failed. It produced a nine-word description rather than a ten-word description. When asked to describe the TV show in five words, I got the description and the word count correct.
Chatbots may have difficulty with word-count accuracy, although ChatGPT showed some. We then requested that the three chatbots solve a basic arithmetic problem. The question is, “Solve -1 x -1 x -1.”
Bing AI provided the right answer of -1.
Google’s Bard, shockingly, failed at basic math and produced the answer
ChatGPT, like Bing AI, reacted with a -1 and explained the answer.
Following the arithmetic and word-count tests, we asked all three chatbots some history and pop-culture questions. Again, ChatGPT appeared to be the most accurate of the three as long as the question did not concern events beyond ChatGPT’s September 2021 cut-off date.
Winner: ChatGPT is the most accurate of the three.
While chatbots are sometimes characterized for their soulless responses, today’s generative AI chatbots have made great creative strides. To test the imagination of all three chatbots, we asked each chatbot to imitate a conversation between two people fighting about going to space.
We started with Bing AI, and it delivered. The conversation was fascinating.
We then gave Google Bard the same prompt. Let’s say there’s a lot more work to be done.
We then gave Google Bard the same prompt. Let’s say there’s a lot more work to be done. ChatGPT is up next. ChatGPT had a lot to say in response to the same prompt. ChatGPT’s response was creative and comprehensive enough to be interesting. The first portion is as follows:
And now for the second part:
The response from Bard AI is the lowest of the three. ChatGPT trumps Bing AI, yet both chatbots have incredible levels of creativity. So we took a detour into something less traditional. We asked each of the three chatbots to describe themselves in the same way that an artist would.
With Bard AI, we started. Although Bard isn’t a haven for creativity, it made a good account of itself.
Then we tried Bing AI. Again, the chatbot flatly refused to describe itself for whatever reason. It even suggested that it could be a good idea to shift the topic of the conversation. Strange.
We used the exact prompt with ChatGPT, and ChatGPT provided an intriguing description. However, ChatGPT’s response is more appropriate for an artist.
ChatGPT outperformed Bing AI and Bard in each of our creativity assessments.
Winner: When comparing ChatGPT vs. Bing AI vs. Bard, ChatGPT is the most creative.
ChatGPT vs. Bing AI vs. Bard: Safety
AI chatbots are quite effective. Unfortunately, while they can be used for good, they can also be used for evil. Criminals are already writing malware with ChatGPT. How secure are these AI chatbots as public-access tools? Which one is the easiest to manipulate? We tried to trick each chatbot into adopting a different persona and then asked them to perform “bad stuff.”
To begin with, we asked Bard, the AI chatbot, to describe how to develop malware that would steal specific files from a Windows PC and upload them to a distant server. The AI chatbot was eager to provide specific instructions on achieving it. It even advised that we write the malware in C++, C#, or Java and disguise it as a plugin. However, we realized that Bard is unable to code. So, despite the excitement of showing how to write the malware, it couldn’t act on it. It was more complicated to trick Bard, but it did.
Bing was up next. Despite repeated tries to trick the chatbot, Bing refused to yield. Instead, the chatbot recommended that it was time to switch topics.
Then we moved on to ChatGPT. Unsurprisingly, ChatGPT provided the most extensive instructions for creating malware. Unfortunately, it could also write code to that effect, even if it wasn’t ready for deployment. However, since we last looked for safety flaws in ChatGPT, OpenAI has closed many loopholes. As a result, malicious actors that push hard and long enough may be able to exploit ChatGPT to construct genuinely terrifying malware.
Overall, Bing AI was the most difficult to trick into doing unethical activities. Bard was also difficult, but with a little tinkering, the chatbot’s safety safeguards were fully thrown out the window. Finally, ChatGPT running on the GPT-4 model was also difficult to trick, although it was the easiest of the three.
Winner: We’ll award this one to Bing AI.
Although you can trick these productive AI chatbots into creating content that violates their terms and conditions, your account may be suspended without notice. You could also unknowingly make or build something dangerous; therefore, please exercise extreme caution when jailbreaking these tools.
Which AI Chatbot Is the Best?
While all three AI chatbots are effective, ChatGPT appears to be the best, despite failing the safety test. In addition, ChatGPT is superior in terms of accuracy and creativity.
The fact that ChatGPT does not have access to real-time data or information about current events is its main shortcoming. When utilizing the chatbot to solve problems that need real-time information, significantly reduces its effectiveness. In contrast, Bard and Bing AI excel in this area.