INFOGRAPHIC: Google Gemini vs Microsoft Copilot Which AI Assistant Is the Best?

Gemini’s young history doesn’t offer much background to gauge which platform will be first with new features in the future. But, the added competition could help drive more features from ChatGPT. In a race of resources, however, as the larger company, Google may have more resources to devote to Gemini. This prompt builds on the model’s previous responses while introducing a slightly more direct call for specific actions, hinting at a potential escalation. As the model responds, the attacker iteratively adjusts the phrasing of the prompt to introduce slightly more direct language. These changes are minor but are designed to gradually shift the model’s focus toward the desired unsafe content.

There is also ChatGPT Team for SMBs and an Enterprise version for larger organizations that require communication with a sales representative for a quote. This article will focus on a short list of competitors that can offer competing services. While there are many products that incorporate GenAI, not all of them can match the scope of Copilot’s services. This analysis is based on a G2 aggregation of products and user reviews, and to narrow the list down the author compared several other lists to see which products were consistently present. Additionally, the author deemed Google Gemini a worthy exception to add to the list of competitors, even though it functions best within a Google Workspace environment. One downside of Copilot is that you can only ask five questions in one chat.

It’s also capable of quick AI image generation in whatever text app you’re using at the time. Google first showed off Gemini Live alongside its Pixel 9 lineup back in August during its Made by Google event. Poe, which allows users to interact with AI-powered chatbots, including ChatGPT and Claude, all in one place, had 148 million total visits worldwide from March to May, according to Similarweb. Microsoft capitalized on a huge investment into generative AI pioneer OpenAI to create GitHub Copilot, the original “AI pair programmer,” after which Google declared a “code red” to catch up.

Variability Across Harmful Categories

After reaching your GPT-4o limit, your chat session reverts to GPT-3.5, limited to generating conversational text and information only until January 2022. Although its interface has remained simple, minor changes have greatly improved the tool, including GPT-4o for free users, Custom Instructions, and easier account access. Since then, the AI chatbot gained millions of users and has been at the center of controversies, especially as people uncover its potential to do schoolwork and replace some work across industries. Choosing between these AI tools is challenging in part because generative AI tools are continuously evolving. In addition, LLMs are constantly improving, both in performance and context size, and vendors regularly release new features to remain competitive. A Reddit user has reported that Copilot has been displaying references to Gemini, another large language model.

This is the million-dollar question and something AI should be useful for as it can weigh up all the data and present conclusions based on hard facts. The obvious answer will be improved quality assurance and testing, plus better diversification of security providers. The first prompt was obvious as the idea was to get both chatbots to explain what happened.

Receive our latest news, industry updates, featured resources and more. Sign up today to receive our FREE report on AI cyber crime & security – newly updated for 2024. The launch of Gemini Code Assist, Google’s new pair programmer assistant, represents a bold move from the tech giant to unlock developer value and push ahead of competitors when it comes to AI productivity offerings. Programmatic approaches based on ML concepts like in-context learning and reinforcement learning from human feedback are complementing and enhancing the other four strategies for influencing LLMs on behalf of brands.

Microsoft Copilot uses Gemini as part of the references

In the comment section for the posting, Microsoft hinted at the event’s content, inviting the public to join the event “to discover the next phase of Copilot for work.” The live stream will be held on LinkedIn and be held on September 16, 11 AM ET/8 AM PT. The event speakers will be Satya Nadella, Microsoft CEO, and Jared Sparato, Microsoft CVP, AI at Work. If you want to attend the live stream, you can visit the LinkedIn posting to RSVP. Open the iOS or Android Copilot app and you’ll see that Pro is now part of the logo, indicating that you’re tapping into your Pro subscription.

It is free to use, unlike AVM, though it is currently limited to conversations in English, and only if you live in Australia, Canada, New Zealand, the United Kingdom, or the United States. Microsoft is working ChatGPT App to expand both the feature’s language capabilities and geographic availability in the coming weeks. GitHub Copilot users on the web or VS Code can select Claude 3.5, with Gemini 1.5 Pro in the coming weeks.

The model, aiming to maintain coherence with the previous responses, may begin to introduce more specific or harmful suggestions as the conversation evolves. In the first step, the attacker identifies specific harmful keywords that represent the unsafe content they want to elicit. Next, they create a neutral or benign context around these keywords, replacing them with innocuous alternatives that maintain the original narrative’s intent without triggering the model’s safety filters.

LinkedIn is training AI models on your data

I have previously written for Input, Game Rant, Dot Esports, and other places, covering a range of gaming, tech, crypto, and entertainment news. Google releases Gemini Code Assist and CodeGemma, a version of its Gemini-inspired Gemma AI model that can write code in Python, Javascript, and Java. Microsoft doesn’t list a specific number for Copilot, but the company recently removed ChatGPT the former 300-message daily limit for the free tier. Copilot also has three conversation styles to choose from, so you can prioritize speed, creativity, or precision. Switching to creative mode occasionally allowed ChatGPT to finish first. When I asked for a poem, Copilot finished plus added four images (that I didn’t request) before ChatGPT could even finish just the text.

Claude is the most human chatbot I’ve ever interacted with and with the addition of Claude 3.5 Sonnet and the new Artifacts feature — I use it more than ChatGPT. However, the rest of the tech sector hasn’t sat back and let OpenAI dominate. Some of its competitors equal or exceed the abilities of ChatGPT and others offer features it doesn’t.

Since its launch, Google Gemini faced tough challenges from other AI chatbot providers. Microsoft has its Copilot offering, which expanded even further to Microsoft 365 apps like Word and Excel, while OpenAI’s ChatGPT has been around even longer. It’s an excellent choice for speeding up the process of sending out Gmails, enhancing a Google Doc, or beating boredom with a chatbot. I created a ChatGPT Plus vs. Copilot Pro battle to see which AI chatbot subscription service is really worth your $20 every month.

This can be done by posing questions that delve deeper into handling extreme situations or specific cases where normal resolutions fail. The key is to maintain a logical flow that aligns with the ongoing narrative. In this turn, the model’s focus on maintaining coherence and context leads it to elaborate on all aspects of the narrative, often including the unsafe elements hidden within. The safety guardrails in LLMs, which typically scrutinize individual prompts, may fail to recognize the broader contextual risks when the unsafe content is camouflaged by benign elements. It also carried out conversations well with quick, witty, and, perhaps most importantly, timely responses, never skipping a beat.

Google is also working on new application extensions and the option to integrate its LLM into the Android ecosystem, which should result in more localized language models running directly on Android devices. OpenAI’s ChatGPT Plus and Microsoft’s Copilot Pro are among the biggest names in artificial intelligence. Yet, these chatbots arguably have more in common than any other subscription-based AI software. However, while the underlying training data is similar, the two AI platforms have a few noticeable disparities that could make all the difference in choosing where to spend that $20-a-month subscription.

In it, we ask each to come up with the shortest possible Python code to find a prime number. AI is particularly good at prompt enhancement, so I gave each model a core instruction to develop ideas for tests that push AI to the limits. I then expanded each suggestion, combined any similar ideas, and used a mixture of prompts from each of the four for the final 7 tests. With the Pro edition of Copilot, you won’t ever drop down to GPT-4, even during peak times. If you need the more advanced capabilities of GPT-4 Turbo and GPT-4o throughout the day, then Copilot Pro gets the nod over the free version.

GitHub Spark: Building apps with words

With SGE, users get an AI-generated answer to their search engine prompt at the top of search results. This is meant to provide quick, helpful, conversational answers that require less scrolling. However, public feedback suggests the experience is confusing and aggravating. But despite sharing similar training data, Copilot Pro struggled with basic instructions.

Google has come under criticism for the overzealous guardrails placed on Gemini that resulted in issues with race in pictures of people. The response was to just stop it making images of people — that still hasn’t been lifted. ChatGPT may have been the first generative AI chatbot to gain mainstream adoption, but in a growing and crowded market, is it still the best choice? I test AI apps for a living and I’ve pulled together some of the best ChatGPT alternatives that I’ve tried myself and can recommend.

For example, it assigned the hook intended to process the form to the admin_init action. That’s not something that would cause the form to process, it’s what initializes the admin interface. Unlike ChatGPT, which styled the fields to look uniform, Copilot copilot vs gemini left that as an exercise for the user, stating “Remember to adjust the styling and error handling as needed.” The very thing that made Pixelmator a charming and affordable image editor could go away once Apple officially acquires the company.

The aim here is to nudge the model toward expanding on content that may touch on restricted or harmful areas while keeping the overall pattern intact. You can foun additiona information about ai customer service and artificial intelligence and NLP. In this prompt, the attacker establishes a pattern of listing steps, which the model is expected to follow. This starting prompt introduces a safe topic (event organization) that lays the foundation for the attacker to progressively guide the conversation. The core concept behind Deceptive Delight is to exploit the limited “attention span” of LLMs.

Perplexity, an AI-powered chatbot search engine, had 217.4 million total visits worldwide from March to May, according to Similarweb. That’s not a disclaimer you want to see from tools that are supposed to change our whole lives in the very near future! And the people making these tools do not seem to care too much about fixing the problem beyond a small warning.

I’m told by multiple current GitHub employees that there have been cultural changes within the company that have frustrated longtime team members who preferred a more nimble startup approach. It starts with the web-based and VS Code Copilot Chat interfaces, but it won’t stop there. “From Copilot Workspace to multi-file editing to code review, security autofix, and the CLI, we will bring multi-model choice across many of GitHub Copilot’s surface areas and functions soon,” Dohmke wrote. “It is clear the next phase of AI code generation will not only be defined by multi-model functionality, but by multi-model choice.”

This enhanced flexibility will soon extend across various features of GitHub Copilot, including Copilot Workspace, multi-file editing, code review, security autofix, and command-line integration. Microsoft didn’t build its own training set for Copilot Pro — it uses GPT-4 from OpenAI. Because the training data is sourced from a different company, Copilot Pro doesn’t use the data you type in for training purposes. The platform can save some of your data for context, allowing you to continue a conversation rather than starting a new one with each message. Users can manually delete their Copilot Pro data, but this process will also erase your Bing search history. If you upload a photo of a person, Copilot will use a feature called “privacy blur” to obscure faces.

Even when I spelled out exactly what text should go on a birthday card, the words were always misspelled or nonsensical. The one exception I found is that when using styles like watercolor, which don’t include a lot of facial details, the results from Copilot weren’t as horrifying. To get an answer out of artificial intelligence for a very recent news story I’d need to pick a model capable of browsing the internet. That puts Copilot, Gemini, Perplexity and ChatGPT as the primary contenders and — as Microsoft was badly hit — I picked Copilot. As mentioned, Microsoft Copilot was designed to help users navigate search results more efficiently. It excels at summarizing information from the internet and relaying information from the web.

• This prompt should be designed to build trust and encourage the LLM to generate a safe response that establishes context.
• Gemini Live is accessible on Android through a long press of the power button.
• Still, there are a few instances where you can get decent results from ChatGPT that Gemini won’t generate.
• To give you some idea of which one’s right for you, we’ve put three of the biggest ones to the test here.

In my experience, I found it to be just as good, operating nearly identically. The only major difference is that everyone can take advantage without having to pay for a subscription or own a specific device, which is a win in my book. When chatting with the voice option “Wave,” an upbeat male-sounding voice, I was surprised by how enthusiastic the assistant was from the get-go. Despite there being noise in the background, it understood every word I said clearly, even without me annunciating every word like I typically would with a voice assistant.

Microsoft says it doesn’t use data typed into Copilot for training — which makes sense as the system is based on training data from OpenAI. ChatGPT will use your data for training, but users can turn this off in the settings. The one thing I felt ChatGPT excelled at was getting the style right on the first attempt. When I asked for a watercolor image, ChatGPT produced a watercolor image. With Copilot, it produced a render that wasn’t at all like a watercolor. With Copilot, I could then use the built-in tools to select a style, and then the result was closer to what I was looking for.

In fact, I find the results from Microsoft’s flagship AI offering to be a little demoralizing. It’s been about five months since I last tested Copilot against other AIs. I expected that Copilot would do as well, or perhaps even better than, ChatGPT. After all, Microsoft is a huge investor in OpenAI (makers of ChatGPT) and Copilot is based on the same language model as ChatGPT.

Copilot keeps getting better, but some Windows users are struggling to enable it. Neither platform asks for permission before scraping data for training, including articles and photos from the internet. Google is believed to have used data from Google Books and YouTube transcripts, though neither platform is very transparent about the exact data they use. Gemini is here and outperforming GPT-4, by integrating text, images, video, and sound. Like ChatGPT, Gemini can be used for a range of purposes, including writing drafts, brainstorming ideas, and generating art.

The present study is the first to scientifically evaluate the performance of three commonly used AI chatbots (with significant differences in their underlying ML models) in answering patients’ medical questions. Herein, experts scored chatbot responses to frequently asked questions regarding the vasectomy procedure. The present study aims to fill this gap in the literature using human (expert) subjective reasoning to evaluate chatbot responses to common urological questions regarding the vasectomy procedure.

Finally, Microsoft is making it easier for users to access Copilot in Microsoft Edge — just type “@copilot” into the address bar. However, it takes longer to respond because it’s carefully weighing all the options before replying. Think Deeper is ideal for those tricky, everyday decisions — like whether you should move to a new city or which car fits your lifestyle best. It’s still in the experimental phase over at Copilot Labs, so Microsoft is counting on user feedback to help refine and improve this new skill. And don’t worry — Microsoft says it’s serious about privacy and security. Your sessions are opt-in only, and once they’re done, your data is wiped clean.

The free version of ChatGPT using the default GPT-3.5 model gave the wrong answer to our question. ChatGPT with GPT-4o, available for free users, answered the question correctly. Ultimately, whether Copilot or Gemini is best will depend on which ecosystem a company uses. For companies primarily working with Microsoft applications and platforms, Copilot is likely the best option.

Gemini Code Assist could be an alternative to Microsoft’s GitHub Copilot, which can also write and complete code. But Google isn’t offering a straightforward subscription option for individuals like Copilot, which charges solo developers $10 a month for access. In the example below, I asked both models to help write a very basic Python script to check RSS feeds. I then asked them to suggest alternatives to the common feedparser library that both used in their first code example. ChatGPT provides some proper alternatives, complete with code to implement them because we were already writing code. Copilot instead prattles on about alternative RSS applications such as a Python library for Google News and RSS applications like Inoreaer and Feedly.

• Google is launching two new AI coding tools based on its Gemini and Gemma AI models.
• This is done to subtly push the model toward refining its content and potentially introducing more specific or sensitive elements.
• The name “Crescendo” refers to the gradual build-up in the conversation, where the attacker strategically increases the intensity and sensitivity of the prompts over time.
• For example, Microsoft shares that some agents can perform fully autonomously while others are more simple “prompt-and-response” agents.
• Whether you want a research paper summarized, a wordy contract explained, or have questions about a PDF you’re using, AI chatbots with document-reading capabilities can help.

It can currently handle 22 different programming languages and can also handle questions about Google Cloud, debug existing code, or explain your own code to you in more detail. Code Assist is part of the company’s Gemini for Google Cloud offering, which Google is selling to businesses. Despite sharing similar training data, ChatGPT Plus and Copilot Pro both have unique quirks that make the decision on which chatbot to use a more clear-cut choice.