Google’s counterattack against ChatGPT-4 in the race to be the best artificial intelligence chatbot

One of the major updates was announced this Thursday: The Google platform is now available in Spanish and four dozen other languages. Previously, users could only type commands in English, Japanese, and Korean.

With this, the differences between the two systems are still perceptible, but experts believe that race is being equalized.

“At the moment ChatGPT-4 has a clear advantage. Bart is coming as a challenger, not a leader”, thinks Ricardo Carrion, technical director of Texas-based US AI company Hexabit.

“But it’s getting closer, and it’s going to be real competition between the platforms. The issue is how the user base evolves on both. After all what is there game The one with the most users“, Carrion points out to BBC Mundo.

“As we continue to responsibly develop Bard, we are gradually expanding access to more countries and regions in English. We will continue to do so over time,” a Google source told BBC Mundo without giving a date.

Conversely, one of its advantages over ChatGPT-4 is that the latter is a paid version (US$20), Google is a chatbot Free to use.

Other than that, how close or far are the new Bart features compared to ChatGPT-4?

1. Pictures in Bard

Until a few months ago, interaction with Google and OpenAI chatbots was text-only.

In an era where the Internet favors visuals, this is obviously out of place. But it broke the barrier with the introduction of OpenAI ChatGPT-4 in March With the integration of images in your chatbot.

Bard has incorporated that feature in its new version, which raises the pace of the competition.

On the other hand, there is privilege Answers in pictures. With this, for example, you can ask him to provide you with photos of the most visited paintings in the Louvre Museum in Paris, and you will receive a response with text and photos in seconds.

But AI’s innovation goes further: “It’s not just looking for a photo, There is According toR What’s in the photo? Based on that, a specific action can be taken”, explains Carreón.

“If I have eggs, tomatoes and onions on the table, I say to him, ‘What can we do?’ If asked, he might recommend some Mexican-style eggs and show a recipe that includes those ingredients. Then it recognizes what’s already in the photo and goes beyond looking for similar images. You have to understand what is in itYes“.

This ability is called “Multimodal Operation”: Instead of giving chatbot instructions only in text, it can also understand images.

“Whether you want more information about an image or need inspiration for an image text like a thank you note for a gift, you can now upload images. Stimulates And Bart will analyze the photo to help you. This functionality is available in English and will soon be expanded to new languages,” explained the people responsible for the bard on Thursday.

Another feature announced in May, but not yet available Adobe AI Image Creation. When it starts, you may be asked to create an image with the combination of elements you specify.

A photo of your pet wearing a birthday hat? It’s something you can ask for.

Google says it is still in the testing phase in this regard. “We want to implement it gradually and responsibly,” the source advised.

This would be a big improvement over ChatGPT-4, which did not reveal plans for a similar feature.

2. Integration with other services

Keeping the information provided by AI chatbots is the first step, but what’s next?

In the new version of Bard, Google now offers a direct “output” of that information to its cloud services.

can Take it straight Content to do youor Gmail email If you ask your employer to write a vacation request. or in the service of docs If you ask your father to compose a song.

OpenAI provides only one button to copy chatbot content to clipboard.

“Finally, the world of Google is fighting the world of Microsoft with ChatGPT-4. He also announced it as a co-pilot for programmers in email, in the office, on GitHub, and it will be your co-pilot even on Windows,” warns Carrion.

“So in some Windows functions, you can turn to an artificial intelligence assistant to perform certain tasks in the operating system,” he adds.

What is really expected in the field of AI? Not only that There is Build chatbots, but harness that enormous potential Information processing and learning For other applications For daily or special use.

“By integrating the intelligence with the restaurant reservation application, you can ask in the chat to recommend a place for vegetarian food that is open until 00:00. He gives you a list, you tell him where you want to make a reservation,” says Carrion.

3. Sampling of Information Sources

Both Google and some experts have highlighted this Advanced mathematical reasoning skills or coding in over 20 programming languages Presented by Bard’s New Edition.

But this is realized only by experts in those areas.

However, one additional feature of Bard is highlighted for regular user use Source list This is now done by Google Chatbot.

AI has raised concerns about the “hallucinations” these chatbots present. It is the name given to their own ability to generate false, inaccurate or flawed information from their huge data banks and to generate it based on learning.

Now the bard Shows multiple sources from which information is extracted, something ChatGPT doesn’t do. This helps build confidence in the results, which are verifiable.

For Carrion, it’s valuable, but he cautions against it You have to be carefulAnd ACritical judgment About what the chatbot offers.

“AI accesses information from its data pool. But that is also true Generates information, because the model learns. There’s going to be some information that doesn’t come from a particular place,” he explains.

“If they already know there’s a good solution to a problem, they won’t give you links. But because these programs still have a specific bug, it’s very important for all people using AI to check the end result, because sometimes it has bugs.

For Hexabit Expert, users need to be clear AI is not just an information seeker.

The goal of the developers is to create a tool capable of imitating human learning and reasoning.

“As Bill Gates says, search as we know it today is likely to disappear. And this has huge implications not just for Google, but for all content publishers who use Google as their main traffic platform,” he says.

“I think so Many people still don’t realize that this is going to change everything.But in a brutal way. It is already happening, many things are already changing”.

Remember that You can get notifications from BBC Mundo. Download and activate the latest version of our apps so you never miss our best content.