AI will accelerate growth and difference between high- and low-quality growth companies
Insights | 12 February 2024
Financial impact of the artificial intelligence (AI) is still underestimated. According to the Goldman Sachs reports, in major developed markets AI could increase productivity growth by 1.5 percentage points annually. At the same time there will be higher dispersion between high- and low-quality growth companies.
Rapid development of AI technologies and products has made it possible for all companies to accelerate their growth with AI and make sure they are in the group of high-quality growth companies.
Technology and Market
AI is not new technology. However, during the last few years it has become more relevant as the amount of reachable relevant data and computing power has increased. AI discussion has been more around the data, which has also been called the new oil. Entry of Nvidia into the sphere of global dominating tech companies and, for example, the great book of Chris Miller Chip War: The Fight for the World’s Most Critical Technology have helped us to understand the importance of computing power as well. Around these building blocks of AI – data and computing power – we see lot of interesting growth and investment opportunities.
From our venture capital funds in Butterfly Ventures we have invested into the Nordic and Baltic based AI startups during the last 10 years. In our portfolio we have great AI companies in the areas of MedTech, process industry, video, computer vision and robotics. Many of these areas are the ones where AI will have a big effect. Some of our portfolio companies have built tools for developing more efficient AI and helping their customers to easily deploy AI. At the moment such tools are seizing the momentum and getting lot of attention from investors. In the future we expect to see more interesting AI cases in robotics.
When evaluating AI companies, understanding their target market is essential. Experiences gained from our AI investments have directed us to
- demand well thought HW interface strategy of MedTech AI cases to enable go-to-market in a reasonable time window,
- understand the niche-like and specific market of video and machine vision cases, as the pressure from cloud giants and their foundation models in this area increases all the time, and
- appreciate democratization of AI through no-code and easy-to-deploy solutions, as the AI is needed in many industries where customers do not have time and resources to hire own data scientists, business analysts and developers.
From the AI technology perspective, and before the current hype in generative AI (GenAI) and large language models (LLMs), main evaluation points were access to the relevant data, machine learning models (MLs) and technology required to teach these MLs. Rapid rise of GenAI, that can create images, videos, audio, text etc., and LLMs, like ChatGPT or Bard, has changed the focus of technology assessment a bit and we also need to understand the interaction of a new AI solution with the foundation models.
At the moment we see startups building their solutions on top of LLMs and foundation models. For a Deep Tech investor like us these companies are not always most likely investment targets as their solutions may resemble more features than products.
Deployment of AI
While Deep Tech investors focus on emerging technologies based on science and protected by IPRs, more interesting part of the AI is its effect to all businesses from grocery stores to energy production.
During the last years, service companies providing AI development services have assisted their clients in finding and annotating relevant data and building AI solutions. These kinds of services are quickly becoming less valuable as the amount of ready-made AI technologies and solutions makes it easier for companies to utilize AI without extensive development projects. Microsoft owned Copilot is available for Microsoft 365 users and ChatGPT provides new features and functionalities to make development of specific AI solutions easier.
Now we are at the stage where companies can really focus on evaluating AI’s benefits for their business instead of developing their own AI systems and technologies. For this, companies will need support in understanding what AI solutions are already available and how to efficiently use these in their business. This will create a need of new kind of AI services compared to previous data and development services. Service providers, on the other hand, need to be better equipped to really understand the business of their clients.
Writer: Ville Heikkinen
Writer is the Co-founding Partner of Butterfly Ventures and Advisor at HT Growth Partners