Skip to main content
AI and machine learning July 17, 2023

Beyond ChatGPT: Al, startups, and future opportunities

By Thorsten Lambertus
AI generated image of technology growing from earth and trees
Just three years ago, Al was the #1 buzzword in the startup ecosystem and every investor chasing Al startups.

Yet, as reported in the Financial Times in 2019, the London-based venture capital firm MMC Ventures found no evidence that Al was an important part of the products offered by 40 percent of Europe's 2,830 Al startups. 

A lot has happened since then. Al has emerged as one of the most promising and exciting fields in technology, with the potential to revolutionize the way businesses operate across industries. Al startups are at the forefront of this innovation, developing cutting-edge solutions that leverage machine learning, natural language processing, computer vision, and other Al techniques. 

The state of Al startups today  

In recent years, the Al startup ecosystem has experienced explosive growth and record levels of investment. According to the 2022 State of Al report by CB Insights, in 2021 alone, Al startups raised over $70 billion in funding, a significant increase from the $31 billion raised in 2019. Despite investments decreasing in 2022 to still $45 billion, the long-term uptrend will presumably continue, as will the variety of applications.  

While the growth of the Al startup ecosystem is exciting, there are also several challenges for startups looking to develop and deploy Al solutions. Perhaps the biggest challenge is the shortage of skilled Al talent, which has made it difficult for startups to find and hire the technical expertise they need to develop and scale their solutions. Besides talent shortages, Al startups also face other challenges, including data quality and availability, regulatory and ethical considerations, and the need to build and maintain customer trust. Many Al startups are working with sensitive data, such as personal information or financial data, which requires careful handling and compliance with data privacy laws and regulations. 

Al potential in healthcare and biotech 

Despite these challenges, we would rather look at the opportunities enabled by Al. Since there are too many to cover in an article like this and given that we at ESMT Berlin run a Creative Destruction Lab – a world-class mentoring program for young startups in health – we will put a focus on Al applications in healthcare.  

Al has enormous potential in healthcare and biotech: 

  • Diagnosis and treatment: Al can analyze vast amounts of patient data, including medical records, imaging, and genomic information, to improve diagnosis and personalize treatment plans. 

  • Drug discovery: Al can accelerate the drug discovery process by predicting how molecules will interact with the body, identifying new targets for drug development, and designing new compounds with specific properties. 

  • Clinical trials: Al can optimize clinical trials by identifying eligible patients, predicting outcomes, and monitoring safety and efficacy in real time. 

  • Medical imaging: Al can improve the accuracy of medical imaging, such as X-rays, CT scans, and MRls, and help doctors identify and classify abnormalities. 

  • Precision medicine: Al can match patients with treatments that are tailored to their specific genetic, environmental, and lifestyle factors. 

  • Remote monitoring: Al-powered wearable devices and sensors can collect real-time data on patient health, enabling early detection and intervention. 

  • Healthcare operations: Al can help streamline healthcare operations by optimizing staffing levels, scheduling appointments, and managing the supply chain. 

Real-world examples in healthcare Al  

Innovative startups and established companies alike are using Al to revolutionize medical research and to achieve better patient outcomes via healthcare delivery. To make this tangible, here are some real-life examples of how Al is being used in the medical sector, including BioNTech's acquisition of lnstaDeep, Kheiron Medical Technologies' Al-powered breast cancer screening tool, and PeakProfiling's use of vocal biomarkers to detect adult ADHD. 

Viruses, vaccines, and variants  

BioNTech, the German biotech company behind the mRNA COVID-19 vaccine, recently made headlines with their acquisition of lnstaDeep, an Al startup, for over $680 million in cash and stock. lnstaDeep, founded in 2015 and with offices in London and Tunis, has been dedicated to bringing state-of-the-art decision-making Al systems into enterprises. The company has worked with a diverse range of clients - from teams on the earlier side of Al deployment, such as Deutsche Bahn (Europe's largest rail operator), to companies with leading Al teams, such as Google and NVIDIA. According to Reuters, lnstaDeep and BioNTech had already collaborated on projects, including those focused on evaluating the consequences of new coronavirus variants. With this acquisition of lnstaDeep, BioNTech is doubling down on its commitment to integrating Al systems into all aspects of its biotech work.  

Early detection in cancer treatment  

Kheiron Medical Technologies is a UK-based medical imaging startup and an alumnus of the Creative Destruction Lab Toronto. The company's mission is to help healthcare professionals detect cancer earlier and more accurately, ultimately leading to better patient outcomes. Their flagship product, Mia (Mammography Intelligent Assessment), is an Al-powered breast cancer screening tool. Mia uses machine learning algorithms to analyze mammograms and assist radiologists in detecting breast cancer at an earlier stage, improving the accuracy of diagnoses, and reducing the need for unnecessary biopsies.  

Detecting adult ADHD  

PeakProfiling merges expertise in voice and sound analytics, derived from musicology, with the latest advancements in Al. The sound analytics startup was founded in Berlin as a spin-off from the mathematics and musicology faculties at Humboldt University. PeakProfiling has successfully developed algorithms to identify adult ADHD using vocal biomarkers, as part of a clinical study conducted in collaboration with Charite Berlin and Forschungszentrum Julich. Despite the challenge of detecting ADHD in adults and the lack of objective biomarkers, the algorithms achieved high success rates comparable to those of single-rating doctors. This study, which involved over 1,000 recordings from nearly 700 participants, is one of the largest purely clinical studies in the field to date. 


Al has already changed the world and has the potential to improve many aspects of our lives. Beyond healthcare, industries as diverse as automation, education, and customer service have realized important gains in streamlined processes, error reduction, pattern identification, and customer relationship management.  

However, it is important to ensure that the development and implementation of Al is done responsibly, ethically, and with consideration for potential negative consequences. Even in healthcare, the data used to train Al algorithms can be inaccurate or incomplete, such that data on diverse demographic groups are underrepresented or that the results are biased in a way that leads to poor patient outcomes.  

In this respect, Al has also transformed governance and consumer advocacy. Regulatory and political bodies have used Al to analyze vast quantities of data in areas such as environmental sustainability and cybersecurity to detect threats and create policies that can improve system- and data security.  

With Al startups in the healthcare ecosystem, we are realizing what we could only imagine in healthcare delivery. There are still many challenges that need to be addressed to ensure fair and ethical treatment for all, but, as these examples show, there are increasingly more opportunities to leverage Al to create medical products and services that can truly transform the healthcare industry for the better.

Add new comment