Ark Investment: Google Nano Banana Pro is quite impressive, but its adoption rate is still not as high as ChatGPT.

Ark Investment (ARK Invest) latest discussions indicate that while Google's Gemini 3 has successfully returned to the top model ranks, the groundbreaking aspect of this update is not its language capabilities, but rather the new image and information generation model “Nano Banana Pro.” The Ark team believes that this update from Google signifies a resurgence of its AI technology, but it still faces challenges in user adoption and commercialization strategies.

Gemini 3 responds to doubts, Google emphasizes that the law of AI expansion has not failed.

ARK pointed out that Google's latest release, Gemini 3, has excelled in multiple benchmark tests, thereby countering external doubts about the “diminishing returns of large models.” The Google engineering team stated that scaling up still brings visible benefits, demonstrating a significant increase in training capacity.

In terms of practical experience, Ark AI Research Director Downing stated that Gemini 3 has been able to interchangeably handle most tasks with ChatGPT (5.1 Thinking), with each having its advantages. He believes that Google's integration of memory and personalization features into Gemini is key to this product's significant improvement in maturity.

Nano Banana Pro is the real breakthrough.

Compared to language models, ARK believes that the performance of Nano Banana Pro is more disruptive. The model can condense large amounts of text into structured information such as images, presentations, flowcharts, etc., and it can even correctly output text from images, which is a threshold that most models have struggled to overcome in the past.

ARK pointed out that this capability has high application value for content creation, marketing materials, and visual information work, indicating that Google is clearly leading in image generation and visual understanding.

AI architecture integrates language and images, Google has a first-mover advantage.

The chief futurist of Ark, Winton, believes that the future architecture of AI will integrate language reasoning, image generation, and long-term memory systems, and that Google has made arrangements in all three technologies, including the previously released Titans memory architecture.

The Gemini 3 and Nano Banana Pro are seen as key elements for Google's advancement towards the next generation of AI architecture.

Google and OpenAI are in direct confrontation, with memory functions as the new battleground.

ARK points out that cross-chat memory is becoming the key to retaining users on the platform. After the launch of ChatGPT's cross-chat memory feature, it can remember user preferences and backgrounds, significantly enhancing user engagement.

Google has synchronized a similar design in Gemini this time, indicating that both parties view memory as the next competitive focus. However, AR believes that the challenges of long conversation forgetfulness and the inconvenience of historical content transfer are still issues that the entire industry needs to overcome together.

TPU faces off against GB200, both models chose NVIDIA as the winner.

It is worth noting that when Ark separately asked Gemini 3 and ChatGPT about the “performance/power consumption differences between Google TPU v7 and NVIDIA GB200,” both provided the same conclusion, which is that NVIDIA still has the advantage in terms of performance/power consumption ratio.

ARK points out that although Google has a cost advantage in capital expenditure due to its self-developed TPU, AI training and inference are quickly becoming limited by “electricity.” As electricity becomes a new bottleneck for generative AI, the performance/power consumption ratio will directly determine the amount of tokens a model can output per watt of power, which will also affect overall operational efficiency and revenue limits.

( The technical differences between NVIDIA GPU, Google TPU, and Amazon AWS self-developed AI chips, and the future market trends )

If YouTube Premium is tied to Gemini, Google may rewrite the game.

According to Ark's analysis, if Google merges YouTube Premium and Gemini Pro into a single subscription plan at $19.99 per month, it will create significant pressure on the market. (

AI companies like OpenAI rely on subscription revenue to survive and find it difficult to endure price wars, while Google has multiple sources of income from search, advertising, and cloud services, allowing it to significantly enhance its consumer competitiveness through its content ecosystem.

Google's biggest weakness lies in user adoption, which is not as high as that of ChatGPT.

Despite the recognition of Google's technological progress this time, AR also admitted that its biggest weakness still lies on the user side. AR cited data on app usage hours in the United States, indicating that ChatGPT leads overwhelmingly with 99%, while Gemini only accounts for 1%.

Even when expanded to the global market, Gemini's usage remains significantly behind, not only failing to catch up with ChatGPT but also being clearly outpaced by xAI's Grok in terms of usage minutes. ARK pointed out that this indicates that competition among AI platforms is not solely based on the model capabilities themselves; product entry points, marketing efforts, and user habits are the key variables determining long-term adoption rates.

) Test: Gemini 3 Nano Banana Pro automatically generates humorous illustrations and comics after thinking, turning Trump back into a young man (

This article Ark Investment: Google Nano Banana Pro is quite outstanding, but its adoption rate is still not as good as ChatGPT first appeared in Chain News ABMedia.