“Since at least 2016, AI has been identified as a major driver of growth and investment opportunities in the US market,” said Michel Burgon, senior equity portfolio manager at Indos SFund Solutions in an interview on August 12, 2025. For example, Indossue has been in charge of NVIDIA's stock since 2014.
Bulgon is responsible for the performance of American opportunities, large funds, and two core funds, small and medium and mid-US. Over the past two years, he also managed Global Trends, a fund focusing on current major growth trends, including AI.
Understand AI and its background
Bourgon emphasized that the current perception that AI is only about large-scale language models (LLMs) like CHATGPT is incomplete. Developed by Openai and The Current Star in the AI Space, ChatGpt was originally an internal tool for creating artificial general information (AGI), a more advanced AI that allows you to get initiatives.
Given useful features beyond the development of AGI, it was proposed to make ChatGpt available to more public people to test their applications. The release of ChatGpt in November 2022 sparked immense interest and led to a major wave of capital expenditures (CAPEX) in data centers and semiconductors. However, “the number of applications must increase significantly to make AI successful,” emphasized Brugon.
Main Beneficiaries and Strong Stock Market Success Areas
“In semiconductors, or semiconductors, we hold a dominant and almost challenging leadership position in the GPU and ecosystem (CUDA software),” commented Brugon. Nvidia's ecosystem has reportedly made it extremely difficult for competitors such as AMD to replace GPUs.
Bulgon even highlighted Nvidia's CEO's opposition to US sanctions on China, encouraging China to develop its own AI ecosystem and fears it will challenge Nvidia's control. For example, Huawei is open-sourced GPU software to promote its ecosystem. “When the standard is standardized, it applies to everyone.”
We are primarily in the sci-fi area as to the extent that we can really disable these systems.
“The cloud providers are primarily made up of Amazon, Alphabet, Microsoft and Oracle,” Bourgon said. The cloud industry is likened to “real estate today” due to the combination of key investment cycles (CAPEX) and repeated revenues from client subscriptions. Even after the initial investment, clients continue to pay for data storage and services such as rentals. To make his point, he said that office construction is declining, while cloud infrastructure is on the rise.
The future of AI and its applications
The final stages of AI include application development. This includes “AGIs that can take initiatives, and act as intelligent agents that are potentially beyond the engine, like their colleagues, beyond the engine that asks questions.” Bourgon acknowledged concerns about loss of control or governance issues regarding the creation of exponential applications generated by AGIS, but noted that lawmakers are implementing safeguards, including requiring AI providers to be able to disrupt AI systems.
AI offers greater hope than risk
The current regulations are minimal in the US, but European laws are even stricter. “We are mostly in the sci-fi realm of the extent to which we can truly disable these systems,” Brugon smiled.
Energy and Environmental Challenges
This is a powerful, long-term theme that is intertwined with AI and environmental goals. Because AI requires a large amount of energy, Bourgon observed that companies like Microsoft and Amazon directly order small modular reactors (SMRs) to generate electricity in order to increase energy efficiency. This sector, including companies involved in electricity infrastructure (such as bird construction), offers long-term investment opportunities with an investment cycle of 10-20 years, independent of any particular AI player.
While AI consumes a considerable amount of electricity, Bourgon believes that AI also offers immeasurable possibilities for addressing environmental challenges such as energy efficiency innovation and CO2 emissions. “AI offers greater hope than risk,” he believes it explains why a large part of the US population views energy regulations as unnecessary, as many believe, “AI can save the world.”
Regulation scrutiny
US regulators such as the Federal Trade Commission and the FTC are becoming increasingly challenging to even reconsider previously approved acquisitions (such as Meta's WhatsApp and Instagram). There are rumours that Google may be forced to sell YouTube because of its dominant position. Bulgon believes that the fundamental question for the US is whether they can afford to weaken their tech superstars in the face of competition with China, and that they may not impose similar restrictions. This is a balance between consumer protection and global hegemony.
War of Talent
To avoid antitrust regulations that limit acquisitions, companies like Meta are offering unprecedented amounts ($100 million welcome bonus and annual salary, as well as up to $1 billion of Mira Murati (formerly located in Openai) to demonstrate their “invaluable” expertise in the field.
Nasdaq achieves the record after the record: Will the remaining value be played?
“Magnificent 7” generally looks expensive, but Alphabet (Google) is thought to be P/E with a P/E of 20, lower than the overall S&P. Despite the concerns about AD-based search models (per pale) being heavily confused by AI summary, Alphabet has the maximum volume of collected data, far surpassing the second meta, and categorizes user habits into 39 categories.
Most analysts estimate that even if these companies split up (as explained above), their parts totals are more valuable than their current valuation.
The ability to collect and utilize data is an important competitive advantage for AI companies. In contrast, Apple is perceived as less persuasive in AI because its internal ethical policies collect little or no user data, as “similar to the EU with data regulation” depends on other people's models. This policy has built strong consumer trust and brand loyalty (like Apple to a luxury brand like Hermès), but limits AI capabilities. Despite the agreement between Apple, Openai and Alphabet's Gemini, Bourgon has noted rumors suggesting it could win Prplexity, a response engine powered by free AI that provides real-time answers to questions.
