Artificial intelligence and data science became front-page news in 2023, largely due to the surge in generative AI technologies, which captivated widespread attention. Looking ahead to the rest of 2024, the big question is: what new developments will AI and data science bring, and how will they continue to change the business landscape and transform our everyday lives? In this blog, we look at four key trends in AI and data science that are likely to have the biggest impact this year.

  1. Automation and Efficiency: The Business Game-Changer

Over the past few years, there has been a significant shift in how businesses approach artificial intelligence (AI) and machine learning (ML) technologies. Initially, these technologies were primarily explored in experimental projects, where businesses sought expertise to understand and leverage their potential. These early projects were often limited in scope, serving more as proofs of concept rather than core elements of business operations. Today, AI and ML have become crucial for many organizations.

Most of the AI and ML projects we do for our clients today focus on full operationalization, leading to increased automation and efficiency. This includes various technologies, including  computer vision, AI operations monitoring, predictive analytics and optimization. In essence, we help our clients harness the power of AI and data science, making their operations smarter, more efficient and better equipped to meet the challenges of today’s fast-paced business world.

2. Advancements in Natural Language Processing (NLP): Leveraging the information captured in written and spoken language

You might be familiar with the term Natural Language Processing (NLP). It is a branch of artificial intelligence (AI) that allows computers to understand human language, whether it’s in writing, speech, or even a quick scribble. For many years, it has been a valuable information source in business, we have been using NLP techniques to extract information from written or spoken text. Until a few years ago, most applications of NLP in business processes involved custom-trained models on self-owned and labeled datasets.

One example that created high business value, is the following case where our Data Scientists trained and implemented an NLP model to sort emails into categories: sorting emails. Custom models work well for specific use cases, but need a lot of attention and are very sensitive to the quality of the training data. Challenges particularly arise in areas like text generation and translation where lots of data and complex models are required – many businesses do neither have the necessary amounts of data, nor the funds to invest in training these complex models.


By loading the video, you agree to YouTube's privacy policy.
Learn more

Load video

Fast forward to now, we have these amazing pre-trained models like BERT, GPT-4, LLama, and Gemini! Instead of starting from scratch, Data Scientists can start with a general model that already works really well and just add some custom data and logic on top. This makes it much faster, cheaper and easier to start implementing NLP solutions, opening up more opportunities. Summarizing articles can be done efficiently. Creating a chatbot that mimics human conversation is more straightforward than ever. Additionally, translating your website into various languages is achievable with minimal effort. These new tools are not only cost-effective, but they’re also user-friendly.

3. Responsible AI: The Importance of Transparency

The more influential AI becomes, the more it is necessary to ensure that AI systems are fair, transparent and accountable. Take the financial services industry as an example: explainability has always been an important topic because of strict regulations. This aspect is particularly vital in sectors such as financial services and healthcare, where decisions can have major implications for people’s financial stability and quality of life. For this reason, in these sectors Data Scientists often apply simpler AI models, such as logistic or linear regression. These models are easier to understand than more complex models that may perform slightly better but are less transparent. Especially with the rise of complex LLMs, which have billions of parameters and are trained on huge data sets, the emphasis on explainability is likely to increase. Making AI work well while making sure we understand how it works is a big challenge, especially in regulated or high-risk industries.

4. AI in Healthcare: Revolutionizing Patient Care

In recent years, the demand for predictive analytics in patient care and hospital operations in healthcare has increased significantly. For example, we worked on a project for Zuyderland, a hospital in the southern part of the Netherlands. During this project, we helped them to gain insight into the predicted number of incoming patients and the simulated movement of patients from one department to another. This resulted in solid predictions of the number of beds needed throughout the hospital, which was useful for tracking long-term capacity planning. The project shows how artificial intelligence (AI) can revolutionize hospital management. By streamlining hospital operations, accelerating medical research and delivering more personalized and precise patient care, AI is poised to significantly change the healthcare landscape.

Moreover, there are numerous promising developments in drug discovery, personalized medicine and diagnostics, all powered by artificial intelligence (AI). Given these developments, it is reasonable to expect AI to have a major impact on healthcare in the coming years.

Closing Thoughts

AI and data science are now a big part of our daily lives. They help businesses work better and focus more on their customers, thanks to smarter automation and improved language processing. We also need to make sure that AI is used in a way that is ethical and fair for everyone. In healthcare, AI is doing more than just making things run smoothly; it’s really changing how we care for patients and the results we can achieve. It will also dramatically change the way we work and it is crucial this year to train people how to use AI effectively.

By keeping up with these trends, we’re not just staying ahead in our field. We’re also able to provide real business value using AI and data science. This means working together, being open to change, and exploring all the exciting opportunities these technologies offer!

This blog is written by Maartje Keulen, Head of Data Science & AI at Mediaan Conclusion.