META ACQUIRES MANUS
We begin with one of the most striking acquisitions of recent years. Meta announced that it is acquiring the startup Manus for a price exceeding 2 billion dollars. And here is the part that really stands out. Manus was founded just 9 months ago. I will say it again. Nine months. That alone is astonishing.
Manus is not just another chatbot. It is an AI platform that operates as an autonomous agent. This means it does not simply answer questions or write emails, but understands complex tasks, makes decisions, uses tools, and completes projects, for example building an entire website or researching and writing a full report, with minimal user intervention.
The company was founded by Butterfly Effect Technology in China, but very quickly moved its headquarters to Singapore in order to avoid the impact of US sanctions on Chinese AI companies. This relocation also allowed it to attract investments, enabling it to rapidly reach an annual revenue run rate of 125 million dollars. And now, it becomes the new jewel in Meta’s crown, as the company continues to invest systematically in upgrading its recommendation systems and next generation immersive hardware.
NVIDIA’S HIDDEN DEAL
Now let us move to Nvidia. Shortly before Christmas, it announced an equally impressive move: the acquisition of Groq’s core technologies and key personnel for 20 billion dollars. A deal that many analysts are already calling an “acqui-hire”, meaning an acquisition primarily for talent and technology.
Groq is known for its LPU chips, Language Processing Units, which are designed specifically for inference, the stage where large language models are applied in real world use. These chips promise low latency, high energy efficiency, and exceptional speed.
With this acquisition, Nvidia strengthens its position not only in model training, where it already dominates, but also in inference. This means it adds another powerful weapon to its ecosystem, which already includes GPUs, software, infrastructure, and now LPUs as well. Analysts emphasize that this move is comparable in importance to the acquisition of Mellanox in 2020, which became a cornerstone of Nvidia’s networking capabilities.
And yes, 20 billion dollars sounds like a lot. But for a company with such a strong cash position and cash flow, it is essentially small change. A strategic bet to remain dominant in the future of AI.
THE FORECASTS
And then we have the analysts. Let us take a look at which companies are expected to lead in 2026 according to Wedbush and Cantor Fitzgerald.
Wedbush believes that Microsoft will enter a full phase of AI driven growth, with Azure and enterprise AI solutions reaching their peak during the year. For Palantir, it predicts that the company could reach a market capitalization of 1 trillion dollars, driven by the strengthening of partnerships with both governments and the private sector.
Tesla, on the other hand, is expected to surge due to Full Self Driving and the Cybercab service. The most optimistic scenario forecasts a market cap of 3 trillion dollars by the end of 2026. CrowdStrike continues to expand, with artificial intelligence becoming even more deeply embedded in cybersecurity. And Apple, which until recently appeared hesitant about AI, has now begun to reap benefits. Analysts forecast added value of 75 to 100 dollars per share from leveraging its massive base of 2.4 billion iOS devices.
From the Cantor Fitzgerald perspective, the spotlight once again falls on Nvidia and Broadcom. Nvidia, with its new Blackwell and Vera Rubin chips, is entering a “supercycle”, meaning a period of accelerated revenue and profit growth. Broadcom also benefits from the rise of data centers and AI infrastructure, mainly due to its role in the supply chain.
