“The real profit impact of AI is yet to come.”

Values ​​associated with artificial intelligence (AI) drove the market rally during the first part of 2023. For example, the DWS fund focused on this topic, DWS Invested Artificial Intelligence, returned more than 30% in a year (7.25% analyzed in three years). , according to Morningstar). Its director, Tobias Rommel, believes that the potential for artificial intelligence will increase in the coming years while the impact of this technology on the benefits of companies in the sector is already being appreciated.

Are we facing a new tech bubble, or is intelligence here to stay?

The trend of increasingly intelligent machines has been the norm for the past 20 years. Now, artificial intelligence (AI) already has many practical applications. For example, in 2017, Google DeepMind forcibly erupted creating an anticipation that started a new cycle. This is one of the points that led us to launch the DWS Invested Fund for Artificial Intelligence in 2018. It is an idea that was reinforced in December last year with the launch of JetGPT, which in just two months reached 100 million users. No application has ever seen such a rapid increase in the number of users and this clearly leads to an emergence Artificial intelligence to the next level.

Should we distinguish between producing AI components and developing AI applications when looking for market opportunities?

I will do three main pillars. This is how we handle our fund. One is history [los datos y tratamiento de los mismos] and companies that help collect the data or that own the data. The second is the computing power, which is the semiconductor, needed to make AI models faster and more powerful. graphics processing units [unidad de procesamiento gráfico] Nvidia processes information a million times faster than 10 years ago. Data and computing power are the two most important drivers of AI, and third, we have use cases where companies apply AI in a different way to drive revenue and profit growth.

And which of them can boost the economy in general?

It is clearly the semiconductor branch that has the potential to transform other industries, such as agriculture or autonomous tractors. Healthcare is a very exciting field for the application of AI in the future.

Now the main protagonist of the story is Nvidia. Is it really an industry leader?

Yes, in terms of ways to train AI or AI algorithms. graphics processing units, What is nvidia doingproven effective because unlike CPUs [unidad central de procesamiento] Traditional ones, for example from Intel, can only do one thing at a time. GPUs, by design, were historically intended for computer graphics processing, so they could compute many things at the same time, such as compute all the pixels on a computer at once. This is very effective.

And there are no other players in the same market?

This Nvidia GPU is the first of its kind on the market. There is another elusive leader, AMD. What influences their growth is that when you invest in data centers, you invest in these types of companies. Before, semiconductors were bought that relied more on the CPU part and less on the GPU. Now with artificial intelligence, GPUs are becoming more relevant. It is likely that soon there will be more significant players like Google or Microsoft, trying to develop their own solutions.

Are there competitors in artificial intelligence in Europe?

I think Europe does not have much weight in this sector at the moment. If you look at our portfolio you’ll see it’s barely 10%. It’s a small part, but it’s growing. We are a global fund and we don’t want to invest in US companies only. We have an important role to play in Asia as well, because I think the AI ​​race is basically a battle between the US and China. And there will be winners in China as well because they are so far ahead in terms of technology. However, Europe is lagging behind because there are few companies that are leaders in what they do. It also affects the ecosystem, since we don’t have anything like Silicon Valley where you have the best universities and the best scientists…

What is the basis for searching for investment opportunities in companies related to artificial intelligence?

We want to invest in companies that have outstanding AI product offerings. It’s not just a bank that has a centralized communication service chatbot, because probably all banks will have one in a couple of years. It’s not something that drives performance or improves your bottom line in any significant way. So we look at the AI ​​products around the company model and also what kind of growth potential the company offers with their AI products or services. Another element is whether they really make a difference with their AI strategy and how it affects the company’s market valuation. It’s an analysis process that we do internally because companies don’t report a percentage of profits related to AI.

Are you planning to reduce exposure to US stocks?

American companies get the most attention. If there are companies like the ones in the US in Europe, the outlook is much better. It’s been a great year for these technologies. We didn’t, frankly, expect the evolution to be so significant. There’s even an impact on earnings, as we’ve seen in the last weeks business results.

What are the changes in expectations as a result of these results?

Some have already raised their earnings guidance for the year because of AI’s intangible gains. But I think it’s too early to see the real impact, because integrating AI tools into your products takes time for your customers to adapt and adjust pricing. This process may take up to two years. However, we are long term investors, so we invest with a 5 or 10 year perspective. I don’t expect a lot of changes from month to month in our portfolio. We focus more on fundamental analysis and what we think about the potential of the company in question.

I mentioned earlier China. There is currently no best relationship between the country and the United States, so could this relationship be a problem for the industry?

Yes, strategic protection of semiconductors is something that all governments have done. Also with the pandemic, when we saw the problems in the supply chain. We are now trying to bring more semiconductor manufacturing into our region, and the US is restricting access to some critical technologies from going to China. But it has not reached the point where China’s ambition or progress is being constrained in any meaningful way. But with the limitations, it’s clear that China’s semiconductor production will suffer for years to come.

Do you see any other risks of this new technology?

For me the question is How will we organize artificial intelligence? Historically, when you had a new technology or a new product, very serious processes were run and someone had approved the product for AI. We didn’t have that, we just launched LLMs, screening outlets… Sometimes we find out what they can do after they’re approved. For example, doing software programming and then we use it for those applications. So I think the important question is around regulation. How can Europe, the United States and China develop a common or similar regulation on a global scale? I think it is very difficult to do this. This is something we need to monitor, but it’s still too early to tell how businesses will be affected.

What other industries could benefit the most?

The potential use of artificial intelligence is enormous. For example, the use of autonomous vehicles. I’ve seen tests of self-driving vehicles in China or the US where improvements are evident in recent years. We already have self-driving tractors, in the countryside it is easier than on city streets where there is no traffic and traffic lights … Everything related to health is another point where artificial intelligence can help, also in reducing potential costs. It can be used to ingest and process diagnostic cancer images to improve treatment or, at some point, to better identify cancer from a human.