Google’s Gemini Versus OpenAI’s ChatGPT: A Comparison of Two Generative AI Tools

HOUSEKEEPING NOTICE: Yes, I know this blogpost is somewhat off-topic for a blog about “news and views on social VR, virtual worlds, and the metaverse,” but hey, my blog, my rules. 😉

Also, this is the first of what I suspect will be many blogposts about the Apple Vision Pro spatial computing headset, which I am eager to get my hands on as soon as I can! Accordingly, I have created a new blogpost category, called Apple Vision Pro. I have a second blog post on that project, which I hope to publish very soon. (This device is not yet available in Canada, and we still don’t know when it will go on sale here.)

With my recent focus on Second Life these past few months, you can expect to see a return to more varied topics on the RyanSchultz.com blog from here on in, including but not limited to the latest crop of new VR headsets and metaverse platforms. For example, you might not know that I have received a Meta Quest 3 VR headset as a test/evaluation unit for the virtual reality lab project at my university, and I have been working on a review of that device, but I’m afraid it’s not done yet!

Thank you for your continued patience as I try to find some sort of balance between my full-time paying job (which pays the bills!), this blog (which does not!), and the rest of my life (Auntie Ryan is 60 now, and she needs her beauty sleep… 😜).

Yesterday, Google announced that their generative AI tool, called Google Bard, had been renamed to Gemini, and was now finally available in Canada (The Globe and Mail newspaper provides some background on why there was a delay). So, I signed up, and immediately put it to task, asking both it and OpenAI’s ChatGPT the same question, to compare responses and see how their answers would differ.

The cost of both these generative AI services is approximately the same, at US$20 per month. I subscribed to Gemini Advanced, which gives users access to Ultra 1.0, Google’s largest and most advanced AI model, as one of the benefits of a new Google One AI Premium plan (I was already a Google One user for cloud storage, so this was an upgrade). I get to try out Gemini Advanced for two months on a free trial period, before my credit card is charged.

For my first search, I used both ChatGPT and Gemini to answer a question that I have been thinking about over the past few days:

I am planning to purchase a MacBook Pro laptop to replace my Windows desktop computer. I plan on using the MacBook Pro laptop alone, and with an Apple Vision Pro spatial computing headset when they become available for sale in Canada. What things so I need to keep in mind when I am moving from a Windows environment to a Mac environment?

(Please note, this is a question about my work setup, not my home setup…I’m planning on keeping my high-end Windows 11 PC and Valve Index VR headset!)

OpenAI’s ChatGPT

I wanted to use the latest and most advanced model of ChatGPT, ChatGPT4, which includes web browsing, advanced data analysis, and integration with the DALL-E image generation app. Unfortunately, today ChatGPT4 was not cooperating, forcing me to verify my humanness every time I submitted my query, then giving me the following error message when it did go through:

I got this message every time I tried to use ChatGPT4, perhaps because the system is overloaded?
Then, I kept getting the following error: “Our systems have detected unusual activity from your system. Please try again later.”

So, after several attempts, I gave up, and switched to ChatGPT3.5, which bills itself as “our fastest model, great for most everyday tasks.” (I should note that I use ChatGPT with the free Superpower ChatGPT plug-in for Firefox, which allows me create folders, save chats, etc.)

Here’s the response from ChatGPT3.5:

Moving from a Windows environment to a Mac environment can be a smooth transition, but there are several things to keep in mind to ensure a seamless switch:

  1. User Interface Differences: macOS and Windows have different user interfaces and keyboard shortcuts. Spend some time familiarizing yourself with macOS, including the menu bar, Finder, and the Dock.
  2. Software Compatibility: Check if the software you currently use on Windows has macOS equivalents or if there are alternative options available. Many popular software applications have versions for both platforms, but some may be exclusive to Windows.
  3. File Compatibility: Ensure that your important files are compatible with macOS. Most common file types like documents, images, and videos are interchangeable, but specialized formats or proprietary software files may require conversion or specific software on macOS.
  4. Cloud Services Integration: If you use cloud services like Google Drive, Dropbox, or OneDrive, make sure they have macOS applications or web interfaces. This ensures seamless access to your files across platforms.
  5. Learning macOS Basics: Take advantage of online tutorials, guides, and resources to learn macOS basics, such as navigating the file system, using built-in apps like Safari, Mail, and Calendar, and understanding system preferences.
  6. Security Differences: While both macOS and Windows have built-in security features, they operate differently. Familiarize yourself with macOS security settings and best practices to keep your system secure.
  7. Backup Strategy: Set up a backup strategy using Time Machine or a third-party backup solution to ensure your data is protected. macOS has built-in backup software called Time Machine, which makes it easy to back up and restore your files.
  8. Hardware Compatibility: Ensure that any external devices you use, such as printers, scanners, or external hard drives, are compatible with macOS. Most modern peripherals have macOS drivers available, but it’s always good to check before making the switch.
  9. Apple Ecosystem Integration: Since you plan on using an Apple Vision Pro spatial computing headset, explore how it integrates with macOS and any specific requirements or software needed for compatibility.
  10. Community Support: Join online forums, communities, or user groups dedicated to macOS to get help, tips, and advice from experienced users.

By considering these factors and taking the time to adjust to macOS, you can smoothly transition from a Windows environment to a Mac environment and make the most of your new MacBook Pro.

Now, compare this with the response I received from Gemini:

Absolutely! Switching from Windows to macOS can be exciting, and even smoother with the potential use of an Apple Vision Pro headset. Here’s a breakdown of key things to consider for a successful transition:

The Basics

  • Learning Curve: Expect a bit of a learning curve. The interface and overall philosophy differ slightly between macOS and Windows.

Be patient, give yourself time to adjust, and utilize resources like Apple Support for guidance. Keyboard Shortcuts: Get acquainted with Mac shortcuts. The Command key (⌘) replaces the Windows Ctrl key for many core shortcuts (e.g., Copy = ⌘+C, Paste = ⌘+V). “Right-Click”: Macs have a few ‘right-click’ methods: Ctrl-click, two-finger click on a trackpad, or clicking the right-side of an Apple mouse. Finder: This is the Mac equivalent of Windows Explorer. It’s your primary tool for navigating files and folders.

Ecosystem & Software

  • Apple Ecosystem: Macs shine within the Apple ecosystem (iPhone, iPad, etc.). Features like AirDrop (seamless file transfer) and Handoff (continue tasks across devices) integrate beautifully.
  • Software Compatibility: Most major software has Mac versions (Office, Adobe Suite, etc.). But always double-check if you rely on specialized Windows apps – you might need to find Mac alternatives or explore remote desktop solutions.
  • iCloud: Apple’s cloud storage is deeply integrated with macOS. It’s great for backups, syncing files across devices, and photo/document storage.

Hardware & Connectivity

  • Ports: MacBook Pros primarily feature Thunderbolt/USB-C ports. Get USB-C adapters if you have peripherals that use older USB-A or HDMI connections.
  • Touchscreen: While Macs don’t generally have touchscreens, many gestures become intuitive on the trackpad. The Apple Vision Pro may introduce a significantly different touchscreen-based interaction model with macOS.

Preparing for Apple Vision Pro

  • System Requirements: Be sure to check the requirements for running Apple Vision Pro. It’ll likely need a powerful MacBook Pro model for optimal performance.
  • Spatial Computing Workflow: Concepts like using apps “around you” in space or hand gestures will be part of spatial computing. Begin thinking about how these interactions will change your usual workflow.
  • Developer Resources: If you’re interested in developing for Apple Vision Pro, keep an eye out for Apple’s SDKs (Software Development Kits) and development resources.

Useful Resources

Additional Tips

  • Data Migration: Apple’s Migration Assistant can help transfer files and settings from your PC.

Trial Period: If possible, try a friend’s Mac or spend some time with one at an Apple Store to get a feel for the environment before purchase.

Let me know if you’d like a deeper dive into any of these areas or have more specific questions!

One difference which I noticed immediately is that Gemini integrated Google Search results into its answer. When it did this, it highlighted the text, and provided a link to click through to the Google Search result referred to (see the red arrows in the diagram below):

Interestingly, in one case, there was a link, but no Google Search result! (And, since the Apple Vision Pro has its own CPU and GPU, I would think that you could use any MacBook Pro model without issue.)

And, as you can see in the full response I quoted from Gemini, it even went so far as to give me a direct link to an Apple support document on how to transition from Windows to macOS! (Plus, another Google Search result, a second Apple support document talking about Apple’s Migration Assistant, a potentially useful service which I did not know about.)

Comparing the ChatGPT3.5 and Gemini responses, I found the former to be much more general and less helpful overall, than the latter. This reference librarian also dislikes how ChatGPT just spits out an answer, without telling me the sources of the information.

Also, Gemini had a feature I had never seen before in ChatGPT: the ability to see successive interim drafts of its answer! For a follow-up question, I noticed a Show drafts link which, when clicked on, allowing you to switch between three drafts of the answer to see how they differ from each other.

Anyway, for this question at least, I am going to award Gemini the win (which might not be fair, as I wasn’t able to use ChatGPT4, which might have been a fairer head-to-head comparison). Then again, if ChatGPT is struggling with performance issues due to too many users, maybe that’s a serious problem that OpenAI (backed by Microsoft) needs to address. It’s definitely a factor to take into consideration when comparing the two services (and yes, eventually Gemini, too, could become overloaded).

I will continue to kick the tires on both Gemini and ChatGPT3.5/ChatGPT4 over the next few months, and then I will probably decide to cancel one of the two subscriptions, since I honestly cannot see myself shelling out US$40 a month for two services that essentially duplicate each other. Stay tuned!

Liked it? Then please consider supporting Ryan Schultz on Patreon! Even as little as US$1 a month unlocks exclusive patron benefits. Thank you!
Become a patron at Patreon!

2 thoughts on “Google’s Gemini Versus OpenAI’s ChatGPT: A Comparison of Two Generative AI Tools”

  1. Even if you say that your blogpost is somewhat off-topic, I liked it anyway. 😊
    You can also have a look at Copilot. It, too, uses GPT-4 (the large language model) like ChatGPT Plus (the application, so Copilot is another application), but it’s free (there is a paid version, too). Copilot has 3 modalities, though, and it’s not GPT-4 always. For that kind of questions, looking for factual information, I’d use the Copilot’s Precise mode: that one doesn’t hallucinate so much and touches grass by searching the web, then it provides the links of its sources. Copilot’s Creative mode instead, as the name suggests, is to create stories etc., so of course it will invent/hallucinate anything (BTW Gemini Advanced seems excelling at creative writing), but it can suggest you ideas and other things. Also Creative is less robotic and it can be cute to chat with it, if you treat it as a sort of friend. I’m looking at different systems depending on what I need. For factual info and links to sources, Perplexity.ai is worth a look too.

Comments are closed.