Last week was marked by one dominant topic in the world of artificial intelligence – a quick and slightly unexpected ascent of Google in a ceaseless AI race.
The search giant, known for its Gemini AI chatbot, rekindled the race with a new version called Gemini 3, which it described as its "most intelligent AI model." Boasting its "state-of-the-art in reasoning," the company also said that the Gemini app surpasses 650 million monthly users and that more than 70% of its Cloud customers use its AI.
According to the figures shared by Google, Gemini 3 Pro excels in various benchmarks, including Global PIQA, a new multilingual benchmark covering 100 languages for evaluating physical commonsense reasoning in culturally relevant contexts, in which it scored 93.4%.
Gemini’s performance, however, still lags behind that of ChatGPT, when considering market share and monthly visits, data from various tracking sites, including Statista and Demandsage, shows.
Also, last week, users were stunned by the performance of its new image generator, Nano Banana Pro, which demonstrated top-notch ability to generate all sorts of graphic content and realistic images in a matter of seconds. Some of those who trialed it compared it with artificial general intelligence (AGI), which is often described as human-level intelligence, that would match or surpass human capabilities across virtually all cognitive tasks and is seen as the ultimate goal of some tech companies and startups.
Consequently, fueled by a sharpened focus on AI tools throughout the year and with a renewed investor attention since last week, Google's parent company, Alphabet, was on track to hit a historic $4 trillion market valuation on Tuesday, a new milestone for the tech giant.
Even Salesforce CEO Marc Benioff made headlines just days ago, after he publicly praised Gemini 3, saying he was ditching ChatGPT. "I've used ChatGPT every day for 3 years. Just spent 2 hours on Gemini 3. I’m not going back. The leap is insane — reasoning, speed, images, video... everything is sharper and faster. It feels like the world just changed, again," he wrote on X.
However, Google was not the only one with a new card up its sleeve, as Anthropic, developer of the Claude chatbot, also recently debuted Opus 4.5, which it touted to be "the best model in the world for coding, agents and computer use."
This comes as Anthropic's valuation came to the range of $350 billion following a recent announcement that Microsoft will invest up to $5 billion, while Nvidia will invest up to $10 billion into the startup, according to CNBC. This makes the company, led by Dario Amodei, a former OpenAI researcher, one of the hottest players in the field.
Notably, the release of new Claude and Gemini models shifted the attention from talks about the "AI bubble," which hammered stock markets for weeks ahead of Nvidia's latest quarterly earnings. However, as the chip maker and the world's most valuable public company reported robust profits once again, it instilled calm and a temporary relief, while markets now turn their eyes to the next meeting of the Federal Reserve (Fed).
And while it may look like OpenAI, maker of ChatGPT, slipped in the back plan, when taking into consideration its latest PR crisis and concerns over its profitability, its chief executive, Sam Altman, hinted at his vision of what its forthcoming AI device could look like.
"When I use current devices or most applications, I feel like I am walking through Times Square in New York and constantly just dealing with all the little indignities along the way, flashing lights in my face ... people bumping into me, like noise is going off, and it’s an unsettling thing," he said at an event over the weekend.
"Bright stuff flashing, notifications coming," he added, claiming that this is not "making any of our lives peaceful and calm and just letting us focus on our stuff."
Sitting next to Apple’s former chief designer Jony Ive, whose startup OpenAI acquired earlier this year, Altman talked about the idea of a new AI gadget that would have a "vibe" of "like sitting in the most beautiful cabin by a lake and in the mountains and sort of just enjoying the peace and calm."
His talk came as The New York Times (NYT) shared a new report, in which it attempts to shed light on what it described as "a product used by hundreds of millions of people" but which "inadvertently destabilizes some of their minds."
The report notes how users relied on ChatGPT, turning to it sometimes for many hours, and the chatbot, in return, used to give them all sorts of answers that sometimes just did not make sense.
"ChatGPT told a young mother in Maine that she could talk to spirits in another dimension. It told an accountant in Manhattan that he was in a computer-simulated reality like Neo in 'The Matrix,'" the report says.
This follows a lawsuit in which several families sued the company, claiming ChatGPT drove people to suicide and harmful delusions.
As generative and agentic, more complex AI advances and as ChatGPT "turns 3," reports like this and a growing chorus of experts often point to a need to reconsider safety.
