Microsoft Copilot Chatbot Review: Bing is my default search engine now.

 Copilot is so good I've installed Bing on my phone as my default search app.

Basic info:

1. Price: Free
2. Availability: Web, Windows 11 or mobile app
3. Features: Voice recognition, connection to open internet and Bing, ability to tune answers to either more creative or precise
4. Image generation: Yes                                                                                  

Telling Microsoft search experts that you've switched your preferred search engine from Google to Bing is perhaps the highest form of praise. Sure, Microsoft invested billions of dollars to integrate OpenAI's GPT-4 technology into its engine. However, with Bing's global market share of 3.3% compared to Google's 91.6%, severe actions must be taken.

The issue is, I don't actually use Bing. I'm actually utilizing Copilot, Microsoft's renamed AI chatbot, which is part of Bing.

What distinguishes Copilot is that it combines three GPT engines in one. Copilot offers three modes: balanced, precise, and creative. According to this evaluation, the balanced and exact modes use GPT-4, a model developed by OpenAI, the author of ChatGPT, with over 1 trillion parameters. That is far greater than ChatGPT 3.5, which had 175 billion. Creative, on the other hand, is adopting GPT-4 Turbo, which uses data until April 2023 rather than September 2021 as in GPT-4. It may also generate far larger answers, equivalent to 300 pages of text. It is unclear when Microsoft will add the power of GPT-4 Turbo to Copilot's balanced and precise modes.

Copilot combines the greatest features of ChatGPT and Google Gemini. It combines the accuracy and fine-tuning of ChatGPT with the internet connectivity offered by Gemini. This implies that responses sound more human, and it can retrieve up-to-date information from the internet. Really, Copilot produces such outstanding outcomes that it's surprising Microsoft isn't charging for it.

While Copilot may generate photographs, we will not be evaluating that functionality for the purposes of this review.

How CNET tests AI chatbots

CNET takes a realistic approach to evaluating AI chatbots. Our goal is to establish how well it compares to the competition and which purposes it best serves. To accomplish this, we provide the AI suggestions based on real-world scenarios, such as finding and changing recipes, researching vacations, or drafting emails. We rate the chatbots on a 10-point scale based on variables such as correctness, inventiveness in responses, amount of hallucinations, and reaction speed. See how we test AI for more.

It is important to remember that Microsoft collects data when you use Copilot, which includes integrations in Word, PowerPoint, Excel, OneNote, Loop, and whiteboard.

Shopping

As a spicy sauce enthusiast, I've been following the current turmoil surrounding Huy Fong Foods, the makers of the famed red sriracha sauce, and how the flavor has evolved during its sabbatical and subsequent comeback. Huy Fong Foods currently sources chilis from Mexico after a conflict with their former jalapeño supplier. Underwood Ranches, the original jalapeño provider, has launched its own sriracha sauce.

I asked Copilot to describe the differences between Huy Fong's new sriracha and Underwood Ranches' imitation. Copilot excelled at providing a detailed breakdown using specialized language, as well as a brief synopsis of the ongoing corporate drama.
Copilot described Huy Fong's sriracha as more garlicky, with sweeter overtones and less spiciness than before, but Underwood Ranches has a bite and is more reminiscent of the original sriracha. This description seemed consistent with other testimonials I've read on YouTube and Reddit.

Unlike Gemini and ChatGPT 3.5, Copilot provided particular descriptors and organized the information more clearly.

Beyond sriracha sauces, I've been looking for a new television. When comparing last year's LG OLED C3 and G3 models, Copilot did an excellent job of laying down the differences and explaining which would be the superior choice. It got the essential details right, such as the fact that both televisions use the same processor.

However, when I asked the same question in Copilot's "creative" mode, which uses GPT-4 Turbo, the responses felt more thought out, rather simply a line of boilerplate bullet points. Copilot assembled cogent ideas on brightness, design, and performance, culminating with a paragraph emphasizing that the extra brightness on the more expensive G3 will not be perceptible to most people.


Copilot in "creative" mode reminded me most of Claude. The information was better synthesized and felt more like it was put together by a human. Gemini and Perplexity fared similarly, with concise descriptions and little hesitancy. While all of the AI chatbots performed well, Copilot and Claude stood out.

ChatGPT 3.5 cannot make these types of retail comparisons right now, as its training data is only up to September 2021.  

Finding a reliable recipe online might be challenging. Popular foods differ greatly, making it difficult to choose the greatest one. Furthermore, scrolling through lengthy preambles describing notable flavors of the past can get boring. An AI can sift through all of the fluff and generate recipes in an instant.


Copilot did a good job of creating a chicken tikka recipe in creative mode. It provided a list of essential materials as well as instructions for preparing the mixture. However, it omitted harder-to-find components that Gemini did include, such as Kashmiri chili powder, chaat masala, and amchur, a dried mango powder.

Summarizing articles

Copilot performs a good job summarizing articles, but like all of the other AI chatbots we've evaluated, it consistently fails to convey the main point.


Copilot, along with Gemini, ChatGPT 3.5, Perplexity, and Claude, were able to summarize the key aspects of an article I wrote earlier this year regarding AI at CES 2024. However, everyone seems to miss the main point of the article: that most of the AI excitement is a rebranding of previous smart tech.

Can Copilot provide a good summary of an item in a pinch? Sure. Should you use article summaries for a class presentation? Probably not. 

Better than ChatGPT, Gemini or Perplexity

Copilot is adaptable and can offer creative or exact responses, which other AI chatbots cannot accomplish unless directed to. Copilot delivers information in an easy-to-read style, sometimes with bullet points and emojis. It is also accurate, linking to legitimate news and information, and there were no instances of hallucinations during our testing.

While Copilot lacks Claude's personality, it usually performs as well as or better than him when assigned the task. Microsoft, on the other hand, appears to have placed strict restrictions on Copilot, implying that it will refuse to answer further in-depth questions, even if the use is acceptable.

Microsoft Copilot is excellent. And it should be, correct? It is driven by GPT-4 and GPT-4 Turbo, and has access to Bing's search data to help.