ChatGPT had the edge for a long time - now it could face the same fate as Firefox

Google introduces Gemini - ChatGPT had the edge for a long time - now it could face the same fate as Firefox

The tech industry has been talking about AI for years, but it was only ChatGPT that brought it into the mainstream. As the first AI conversation partner, the bot was a huge success - which even caught the inventors at OpenAI out of the blue(find out more here). But competitors are slowly following suit. With Project Gemini, Google yesterday announced a competitor that should be able to do much more. And it will be installed on billions of devices in no time. Difficult times are likely to come for OpenAI.

Gemini is not one product, but many

Gemini is not a specific program, but an AI model like the GPT-4 on which ChatGPT is based. Google claims that Gemini performs better in almost every respect compared to its well-known competitor. It is said to be better at reading, calculating, science and programming. GPT-4 is only slightly ahead in a test of everyday logic after a long list of comparisons.

The AI is able to understand and simultaneously process images, text, video and audio. In a demonstration, for example, it was able to read, understand and correct a child's handwritten physics homework. In the MMLU comparative test, which assesses problem-solving skills and general knowledge, Gemini is even expected to perform better than a human for the first time.

Google wants to get back to the top

Gemini is an important step for Google. The company has been regarded as one of the major pioneers in the AI field for years. Whether snapping photos, live subtitles on YouTube or the smart Google Assistant: AI has long been deeply integrated into the company's products. The fact that OpenAI is seen by the mainstream as the inventor of AI came as something of a shock to the company.

When CEO Sundar Pichai tried out the bot, he was reportedly surprised. Not because the program could do so much - because Google had long since developed similarly powerful AI - but because the bot inspired so many people despite its many flaws, reports the New York Times. Google followed suit a short time later with the rather ridiculed chat assistant Bard, and Gemini is now finally set to close the gap.

Fighting the supremacy of AI

This is not good news for OpenAI. One fact in particular is likely to cause the company a headache: Gemini is not being released as a single program, but is intended to improve Google's entire product range, so to speak. ChatGPT competitor Bard is one of the first to receive Gemini support; the model powers the English-language version of the chatbot in 170 countries. Google's Pixel series smartphones also have Gemini built in: The Nano variant, which does not require an internet connection, can now create a live transcript of what is being said during voice recordings. In the long term, Gemini is likely to be rolled out to almost all Google services.

This is a huge challenge for competitors. After all, why go to the trouble of opening your own website or app when the AI pre-installed in the device or browser works just as well?

Is OpenAI learning from other people's mistakes?

Other companies that actually dominated a market before Google had the same problem. For years, the Mozilla Foundation's Firefox browser was one of the best and most popular options for surfing the web. Then came Google with its Chrome browser. It inexorably pushed its way to the top. The smartphone revolution in particular gave the Google browser a huge boost: with a market share of almost 70 percent, Google's Android system is the most important smartphone operating system in the world - and Chrome is the default choice.

OpenAI can learn how not to do it from the mistakes of its partner Microsoft: the Windows company was also overrun by Chrome in the browser war - even though it actually supplied its own browser with Windows. However, Internet Explorer's best days were long behind it, and the more modern browsers Firefox and then Chrome were mercilessly overtaken. By the time Microsoft relaunched Edge, users had long since become accustomed to Chrome.

The race has been wide open since ChatGPT

But the game is not yet over for OpenAI. Gemini is not where it should be. Google's self-confident comparisons all refer to the strongest model, Gemini Ultra. The Gemini Nano and Gemini Pro variants that have now been rolled out are nowhere near as powerful. And unlike the most popular Google services such as Maps, Search or Chrome, Gemini Pro is not free. Instead, the company only wants to offer its premium AI to paying subscribers - just as OpenAI does with GPT-4.

For OpenAI, it is now important to play its cards right. Here, too, it can learn from Microsoft. Unlike with the browser, the company has managed to successfully hold its own against Google's attacks with one of its most important products: Microsoft's office software suite Office continues to be the standard - even though Google also offers free Office programs.

Instead of trying to maintain a monopoly as it did with Windows, Microsoft made itself the de facto standard - and offered many of its Office programs in a slimmed-down version free of charge for every operating system. In a mixture of satisfaction and habit, users also use Microsoft's programs at home. And then expect them at work too. There, at the latest, they also pay for the subscription.

However, if OpenAI wants to achieve this status with ChatGPT, it will have to work hard. And win over users until Google can convince them with a better product that is also pre-installed.

Firefox is still around, by the way. Its market share is 3.4 percent.

Sources: Google announcement, Technology Review, New York Times, Stat-Counter.

Read also:

Google, seeking to regain its prominence in the AI field, announced Project Gemini, an AI model aimed at surpassing ChatGPT in various aspects, such as reading, calculating, science, and programming.
The New York Times reported that Google CEO Sundar Pichai was surprised by ChatGPT's popularity and impact, not because of its capabilities, but because of how inspiring it was to people, despite its flaws.
Google claims that Gemini can simultaneously process images, text, video, and audio, as demonstrated by its ability to correct a child's handwritten physics homework.
As a competitor to ChatGPT, Gemini poses a significant threat to OpenAI, as Google plans to integrate it into its various services, such as Bard and Google Assistant, potentially providing users with an all-in-one AI solution straight from their devices.
Similar challenges have been faced by other companies that once dominated their markets, such as Firefox, which struggled against Google Chrome, eventually being overtaken by the latter's widespread prevalence on Android devices.
Microsoft, another competitor, also faced difficulties with its Internet Explorer browser, ultimately losing ground to Chrome and other modern browsers, but managed to maintain its dominance with its Office software suite. OpenAI can learn from these experiences and continue to innovate in the face of competition from Google's Gemini.

Source: www.stern.de