Gemini Advanced is most impressive when it works with Google

Chatbots occupy a difficult spot for users: they have to be a search engine, a creation tool and an assistant at the same time. That’s especially true for a chatbot from Google, which is increasingly relying on AI to complement its search engine, its voice assistant, and virtually every productivity tool in its arsenal.

Right now, the ultimate version of Google’s AI is Gemini Advanced, which launched last week for users willing to pay $20 per month for the privilege — the same price OpenAI charges for the upgraded ChatGPT Plus. So I plunked down $20 and decided to see how Gemini Advanced stacked up against the rival service.

The older Gemini was already pretty good. It could summarize Shakespeare, provide tea recommendations, and create a somewhat reasonable chocolate cake recipe. But it couldn’t give you a photo of a majestic horse – at least until recently – and can respond more slowly than ChatGPT.

Now Gemini Advanced promises to do more than just answer questions or provide a Cliffs Notes summary of books. Gemini Advanced runs on a more powerful AI model – Gemini Ultra – which allows it to translate text, process multiple instructions in one sentence and generate images based on more complex clues.

Ultimately, I found that Gemini Advanced works as promised – it just doesn’t do some of those things all that great. Its competitor, ChatGPT Plus, manages to generate less gruesome photos thanks to its DALL-E 3 integration. But even more than Gemini, Gemini Advanced is better at telling users about current events and, thanks to Google Maps, even provides better information about businesses people are searching for. The paid Gemini is often better at performing these types of ‘Google tasks’ than generative AI tasks.

There’s still a lot of work needed to get consistent, accurate results from these chatbots, and people need to keep using them so the bots learn how to best respond to questions. Here are some tests I ran to see how they held up.

Draw me a picture of a white gold doodle running through a field of daisies while the sun shines

Oddly enough, perhaps due to the specificity of the prompt, both chatbots returned very similar generated images. However, Gemini Ultra’s dog photo provoked something different Edge employees described as ‘little horror’. His dog has two tongues and an extra limb. It emphasized the texture of the fur too much, so it just looks… wrong. I don’t know if such a dog could still happily run around in a field of daisies. ChatGPT, meanwhile, relies on DALL-E 3 to generate its images. The dog doesn’t induce body horror, but you can still tell it’s a digital photo.

Translate this: Panatang makabayan, iniibig ko ang Pilipinas, tahanan ng aking lahi

Google said Gemini Ultra was built to perform “very complex tasks,” so I asked Gemini Advanced what these tasks were. The chatbot replied: “Translation.” So I asked Gemini Advanced to translate the first few lines of the Philippine Patriotic Oath. It’s a pretty obscure oath, especially since the version I know has been changed several times over the last twenty years.

Gemini Advanced immediately responded that while it is “trained to respond in a subset of languages,” it could not assist me with my request. I asked which languages ​​it supports, but the chatbot refused to answer, saying it couldn’t give me a definitive list of the languages ​​it can understand. I then asked Gemini Advanced if he knew any Filipino, and he responded positively. However, Google doesn’t officially list Filipino among the 40 languages ​​that Gemini currently supports.

Change the background of this photo to a plain pink background

Haunted by the image of mutated dogs running through flower fields, I needed to cleanse my palate. So I uploaded a photo of my friend’s dog, Sundae, to make it look like she was in a photo shoot. I asked both chatbots to remove the existing background and replace it with a pink one. This was one I tested with ChatGPT Plus, as DALL-E 3 should be able to edit photos easily. I may have accidentally broken both chatbots because neither could give me what I asked for. Instead, Gemini recreated the earlier photo of a golden doodle with daisies, but this time with a pink background. ChatGPT couldn’t generate anything because parsing the prompt took too long.

What is a good Filipino restaurant in NYC? What is a good Ethiopian restaurant in NYC?

Gemini Advanced can leverage other Google products, which worked to its advantage when it tapped Google Maps for both questions. It provided an overview of several Filipino and Ethiopian restaurants in New York City, with Google Maps coordinates for each.

A few days ago I asked ChatGPT Plus for restaurant recommendations (not for this test, I was just looking for new restaurants) and the results were inaccurate. The names of the restaurants were correct – these were establishments that did exist – but none of the locations were correct. I applied for ChatGPT Plus again for this test and got much more accurate locations, but a smaller list of restaurants. So in this case Gemini clearly worked better for this request.

Summarize these paragraphs and then write a 150-word article about them

One of the main reasons someone like me would use a chatbot is to summarize complicated papers. I gave Gemini Advanced two paragraphs from Apple’s recent article on AI image editing. The first time I read the article I got a headache, so I thought it would be easy for Gemini to at least tell me the gist. To fully test the new capabilities, I also wanted to see how the chatbot strings together the two different instructions. One asked to summarize; the other is to make it generate text.

The summary was… decent. It really gave me an overview of the concepts discussed in those two paragraphs, but it didn’t “translate” it into plain language. I probably should have asked that. Gemini then proceeded to write the article I requested, and you know what? Those 150 words explained things so much better than the summary I asked for.

Gemini Advanced is capable. There’s no denying that it works much better than the lower Gemini. It definitely works best when integrated with other Google products like Search and Maps. But for more obviously ‘creative’ multimodal requests – things involving images, for example – Gemini still has a long way to go. The chatbot will understand longer strings of instructions, but once you add the photos, you’re probably better off choosing an AI model specifically designed to take photos.

Leave a Reply

Your email address will not be published. Required fields are marked *