What is AI and why is there so much interest in it now?
Jonathan McMullan, Global Sector Specialist, Public Equity: “Artificial intelligence (AI) essentially means any technology that enables computers to complete tasks that normally require human intelligence. The concept of AI has been around for a long time. It can be traced all the way back to 1950, when Alan Turing introduced the Turing test for machine intelligence, and like many conceptual technologies, its progress has had its ups and downs over the years. was characterized by alternating periods of heightened excitement followed by periods of disillusionment.
“We already encounter a lot of AI in our daily lives, often without even realizing it. , Generative AI, especially ChatGPT, which was announced late last year, has captivated people’s imaginations.”
“What really sets generative AI apart from other technologies that have been hyped in recent years, like cryptocurrencies or the metaverse, is its tangible, practical nature. It’s not just an abstract idea. It’s becoming part of everyday workflows, and you don’t have to stretch your imagination too far to realize its revolutionary potential.”
Michael White, Global Sector Specialist, Public Equity: “ChatGPT’s success has been amazing. It was the fastest platform in history to reach 100 million users and is now around 170 million. The speed of reaching this milestone suggests that a social habit is forming around the use of text-based generative AI, and that this practice will continue.”
What’s behind the new wave of generative AI apps?
Paddy Flood, Global Sector Specialist, Public Equity: “There are multiple factors in the emergence of generative AI. These include:
- New (ish) architecture: There are various architectural approaches to AI, but in 2017 Google introduced a new architecture based on Transformers. This architecture is an essential building block for the large-scale language models (LLMs) we see today, and among other things, the model contextualizes the entire question (rather than looking at words and phrases in isolation) to It means you can train fast.
- Enhanced computing power: Semiconductors have become smaller and more powerful, allowing them to perform tasks faster and more efficiently. In addition to this, the prevalence of cloud computing has enabled a company to outsource his IT infrastructure to third parties. Without it, companies around the world would have had to invest in expensive AI-related infrastructure, potentially delaying the adoption of generative AI.
- Data: Improved availability and usability of data, which is a key part of LLM, is another reason. The world continues to generate large amounts of data, and advances such as cloud computing are making it easier to access and store data.
- AI at the edge: Finally, we also have the technology to deploy AI at the edge. This means AI computations are done on the device where the data is created, not in a far-off data center. This is very important for applications such as autonomous driving where data instructions need to be executed immediately without delay or latency.
What companies are operating in the generative AI space?
Ankur Dubey, Private Equity Investment Director: “We need to understand the ‘technology stack’, the set of technologies required to build a generative AI application. The stack has four layers.
- The compute layer is the base of the stack. Generative AI systems require large amounts of computing power and storage capacity to train and run models. Hardware (semiconductor chips) provides computing power, while cloud platforms such as Amazon Web Services, Microsoft Azure, and Google Cloud Platform provide services such as virtual machines and storage.
- Then comes the base model layer. A foundation model is a system with a wide range of capabilities that can be adapted for different, more specific purposes. This is perhaps the most important layer of the generative AI stack. These underlying models are large-scale statistical models built using advanced machine learning algorithms to generate human-like responses from large amounts of data based on training. The underlying model is divided into a closed source model and an open source model. Closed source software is proprietary and can only be modified by the company that owns it. Open source, on the other hand, means that the source code is publicly available and can be modified by programmers.
- infrastructure layer. These are tools/infrastructure companies for apps that don’t use their own base model. Such apps require infrastructure companies to be able to take full advantage of the technology available at the foundational level. Apps with their own model (such as ChatGPT) do not need to rely on third parties for infrastructure or base model layers.
- Finally, the top of the stack is the application layer. This is software that allows users to interact with the underlying AI technology. This may include in-house built solutions like his ChatGPT product from OpenAI, or his Schroders in-house AI product named “Genie”.
Which companies will profit most from generative AI?
Uncle Davey: “There is still no consensus on which of these layers will create the most value. Ultimately, the technology is still in its early stages. As an example of NVIDIA, we can agree that the stock is up about 190% year-to-date (Factset, as of June 30) – indicates that the market agrees.
“That said, I doubt whether the cutting-edge technology that NVIDIA is designing today has the potential to become commoditized over time.”
Michael White: “For now, the compute tier ‘picks’ look like winners thanks to their existing dominant positions. As the use cases for generative AI grow, so will the demand for chips. NVIDIA is an expert with overwhelming market share in GPUs (Graphic Processing Units), which are essential for AI processing.
“On the cloud side, the cloud computing market is an oligopoly. At least for now, big players like Amazon Web Services, Microsoft Azure, and Google Cloud Platform have invested heavily in infrastructure in recent years to establish customer relationships. and may maintain its dominance.
“But we must remember that new technologies enable new ways of doing things and create entirely new businesses. It was allowed to thrive because it offered a product that was better than traditional pay TV.
“Similarly, Uber is a company whose business model is built on smartphones and mobile internet. Sure, this exciting new technology seems to offer new ways of doing things, but perhaps those businesses are still emerging. It’s too early to say, and this is what we’re looking for.”
Mike McLean, Senior Investment Director, Private Equity: “Looking outside the tech industry itself, one possibility is that data-rich companies, say companies that have a large amount of their own user-generated content, will see the value of that data in training AI models. It could be a worthwhile company for a reason alone.
“From the venture capital side, as the graph below shows, the flow of money into AI companies has surged in recent years. Investments in the AI space fell last year, reflecting a weaker venture market more generally.
“The bottom line is that AI is becoming an increasingly important factor in the types of companies being created in today’s market.”
